researchgate sentiment analysis

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

A survey of sentiment analysis: approaches, datasets, and future research.

1. Introduction

• A comprehensive overview of the state-of-the-art studies on sentiment analysis, which are categorized as conventional machine learning, deep learning, and ensemble learning, with a focus on the preprocessing techniques, feature extraction methods, classification methods, and datasets used, as well as the experimental results.
• An in-depth discussion of the commonly used sentiment analysis datasets and their challenges, as well as a discussion about the limitations of the current works and the potential for future research in this field.

2. Sentiment Analysis Algorithms

2.1. machine learning approach, 2.2. deep learning approach, 3. ensemble learning approach, 4. sentiment analysis datasets, 4.1. internet movie database (imdb), 4.2. twitter us airline sentiment, 4.3. sentiment140, 4.4. semeval-2017 task 4, 5. limitations and future research prospects.

• Poorly Structured and Sarcastic Texts: Many sentiment analysis methods rely on structured and grammatically correct text, which can lead to inaccuracies in analyzing informal and poorly structured texts, such as social media posts, slang, and sarcastic comments. This is because the sentiments expressed in these types of texts can be subtle and require contextual understanding beyond surface-level analysis.
• Coarse-Grained Sentiment Analysis: Although positive, negative, and neutral classes are commonly used in sentiment analysis, they may not capture the full range of emotions and intensities that a person can express. Fine-grained sentiment analysis, which categorizes emotions into more specific categories such as happy, sad, angry, or surprised, can provide more nuanced insights into the sentiment expressed in a text.
• Lack of Cultural Awareness: Sentiment analysis models trained on data from a specific language or culture may not accurately capture the sentiments expressed in texts from other languages or cultures. This is because the use of language, idioms, and expressions can vary widely across cultures, and a sentiment analysis model trained on one culture may not be effective in analyzing sentiment in another culture.
• Dependence on Annotated Data: Sentiment analysis algorithms often rely on annotated data, where humans manually label the sentiment of a text. However, collecting and labeling a large dataset can be time-consuming and resource-intensive, which can limit the scope of analysis to a specific domain or language.
• Shortcomings of Word Embeddings: Word embeddings, which are a popular technique used in deep learning-based sentiment analysis, can be limited in capturing the complex relationships between words and their meanings in a text. This can result in a model that does not accurately represent the sentiment expressed in a text, leading to inaccuracies in analysis.
• Bias in Training Data: The training data used to train a sentiment analysis model can be biased, which can impact the model’s accuracy and generalization to new data. For example, a dataset that is predominantly composed of texts from one gender or race can lead to a model that is biased toward that group, resulting in inaccurate predictions for texts from other groups.
• Fine-Grained Sentiment Analysis: The current sentiment analysis models mainly classify the sentiment into three coarse classes: positive, negative, and neutral. However, there is a need to extend this to a fine-grained sentiment analysis, which consists of different emotional intensities, such as strongly positive, positive, neutral, negative, and strongly negative. Researchers can explore various deep learning architectures and techniques to perform fine-grained sentiment analysis. One such approach is to use hierarchical attention networks that can capture the sentiment expressed in different parts of a text at different levels of granularity.
• Sentiment Quantification: Sentiment quantification is an important application of sentiment analysis. It involves computing the polarity distributions based on the topics to aid in strategic decision making. Researchers can develop more advanced models that can accurately capture the sentiment distribution across different topics. One way to achieve this is to use topic modeling techniques to identify the underlying topics in a corpus of text and then use sentiment analysis to compute the sentiment distribution for each topic.
• Handling Ambiguous and Sarcastic Texts: Sentiment analysis models face challenges in accurately detecting sentiment in ambiguous and sarcastic texts. Researchers can explore the use of reinforcement learning techniques to train models that can handle ambiguous and sarcastic texts. This involves developing models that can learn from feedback and adapt their predictions accordingly.
• Cross-lingual Sentiment Analysis: Currently, sentiment analysis models are primarily trained on English text. However, there is a growing need for sentiment analysis models that can work across multiple languages. Cross-lingual sentiment analysis would help to better understand the sentiment expressed in different languages, making sentiment analysis accessible to a larger audience. Researchers can explore the use of transfer learning techniques to develop sentiment analysis models that can work across multiple languages. One approach is to pretrain models on large multilingual corpora and then fine-tune them for sentiment analysis tasks in specific languages.
• Sentiment Analysis in Social Media: Social media platforms generate huge amounts of data every day, making it difficult to manually process the data. Researchers can explore the use of domain-specific embeddings that are trained on social media text to improve the accuracy of sentiment analysis models. They can also develop models that can handle noisy or short social media text by incorporating contextual information and leveraging user interactions.

6. Conclusions

Author contributions, institutional review board statement, informed consent statement, data availability statement, conflicts of interest.

• Ligthart, A.; Catal, C.; Tekinerdogan, B. Systematic reviews in sentiment analysis: A tertiary study. Artif. Intell. Rev. 2021 , 54 , 4997–5053. [ Google Scholar ] [ CrossRef ]
• Dang, N.C.; Moreno-García, M.N.; De la Prieta, F. Sentiment analysis based on deep learning: A comparative study. Electronics 2020 , 9 , 483. [ Google Scholar ] [ CrossRef ] [ Green Version ]
• Chakriswaran, P.; Vincent, D.R.; Srinivasan, K.; Sharma, V.; Chang, C.Y.; Reina, D.G. Emotion AI-driven sentiment analysis: A survey, future research directions, and open issues. Appl. Sci. 2019 , 9 , 5462. [ Google Scholar ] [ CrossRef ] [ Green Version ]
• Jung, Y.G.; Kim, K.T.; Lee, B.; Youn, H.Y. Enhanced Naive Bayes classifier for real-time sentiment analysis with SparkR. In Proceedings of the 2016 IEEE International Conference on Information and Communication Technology Convergence (ICTC), Jeju Island, Republic of Korea, 19–21 October 2016; pp. 141–146. [ Google Scholar ]
• Athindran, N.S.; Manikandaraj, S.; Kamaleshwar, R. Comparative analysis of customer sentiments on competing brands using hybrid model approach. In Proceedings of the 2018 IEEE 3rd International Conference on Inventive Computation Technologies (ICICT), Coimbatore, India, 15–16 November 2018; pp. 348–353. [ Google Scholar ]
• Vanaja, S.; Belwal, M. Aspect-level sentiment analysis on e-commerce data. In Proceedings of the 2018 IEEE International Conference on Inventive Research in Computing Applications (ICIRCA), Coimbatore, India, 11–12 July 2018; pp. 1275–1279. [ Google Scholar ]
• Iqbal, N.; Chowdhury, A.M.; Ahsan, T. Enhancing the performance of sentiment analysis by using different feature combinations. In Proceedings of the 2018 IEEE International Conference on Computer, Communication, Chemical, Material and Electronic Engineering (IC4ME2), Rajshahi, Bangladesh, 8–9 February 2018; pp. 1–4. [ Google Scholar ]
• Rathi, M.; Malik, A.; Varshney, D.; Sharma, R.; Mendiratta, S. Sentiment analysis of tweets using machine learning approach. In Proceedings of the 2018 IEEE Eleventh International Conference on Contemporary Computing (IC3), Noida, India, 2–4 August 2018; pp. 1–3. [ Google Scholar ]
• Tariyal, A.; Goyal, S.; Tantububay, N. Sentiment Analysis of Tweets Using Various Machine Learning Techniques. In Proceedings of the 2018 IEEE International Conference on Advanced Computation and Telecommunication (ICACAT), Bhopal, India, 28–29 December 2018; pp. 1–5. [ Google Scholar ]
• Hemakala, T.; Santhoshkumar, S. Advanced classification method of twitter data using sentiment analysis for airline service. Int. J. Comput. Sci. Eng. 2018 , 6 , 331–335. [ Google Scholar ] [ CrossRef ]
• Rahat, A.M.; Kahir, A.; Masum, A.K.M. Comparison of Naive Bayes and SVM Algorithm based on sentiment analysis using review dataset. In Proceedings of the 2019 IEEE 8th International Conference System Modeling and Advancement in Research Trends (SMART), Moradabad, India, 22–23 November 2019; pp. 266–270. [ Google Scholar ]
• Makhmudah, U.; Bukhori, S.; Putra, J.A.; Yudha, B.A.B. Sentiment Analysis of Indonesian Homosexual Tweets Using Support Vector Machine Method. In Proceedings of the 2019 IEEE International Conference on Computer Science, Information Technology, and Electrical Engineering (ICOMITEE), Jember, Indonesia, 16–17 October 2019; pp. 183–186. [ Google Scholar ]
• Wongkar, M.; Angdresey, A. Sentiment analysis using Naive Bayes Algorithm of the data crawler: Twitter. In Proceedings of the 2019 IEEE Fourth International Conference on Informatics and Computing (ICIC), Semarang, Indonesia, 16–17 October 2019; pp. 1–5. [ Google Scholar ]
• Madhuri, D.K. A machine learning based framework for sentiment classification: Indian railways case study. Int. J. Innov. Technol. Explor. Eng. (IJITEE) 2019 , 8 , 441–445. [ Google Scholar ]
• Gupta, A.; Singh, A.; Pandita, I.; Parashar, H. Sentiment analysis of Twitter posts using machine learning algorithms. In Proceedings of the 2019 IEEE 6th International Conference on Computing for Sustainable Global Development (INDIACom), New Delhi, India, 13–15 March 2019; pp. 980–983. [ Google Scholar ]
• Prabhakar, E.; Santhosh, M.; Krishnan, A.H.; Kumar, T.; Sudhakar, R. Sentiment analysis of US Airline Twitter data using new AdaBoost approach. Int. J. Eng. Res. Technol. (IJERT) 2019 , 7 , 1–6. [ Google Scholar ]
• Hourrane, O.; Idrissi, N. Sentiment Classification on Movie Reviews and Twitter: An Experimental Study of Supervised Learning Models. In Proceedings of the 2019 IEEE 1st International Conference on Smart Systems and Data Science (ICSSD), Rabat, Morocco, 3–4 October 2019; pp. 1–6. [ Google Scholar ]
• AlSalman, H. An improved approach for sentiment analysis of arabic tweets in twitter social media. In Proceedings of the 2020 IEEE 3rd International Conference on Computer Applications & Information Security (ICCAIS), Riyadh, Saudi Arabia, 19–21 March 2020; pp. 1–4. [ Google Scholar ]
• Saad, A.I. Opinion Mining on US Airline Twitter Data Using Machine Learning Techniques. In Proceedings of the 2020 IEEE 16th International Computer Engineering Conference (ICENCO), Cairo, Egypt, 29–30 December 2020; pp. 59–63. [ Google Scholar ]
• Alzyout, M.; Bashabsheh, E.A.; Najadat, H.; Alaiad, A. Sentiment Analysis of Arabic Tweets about Violence Against Women using Machine Learning. In Proceedings of the 2021 IEEE 12th International Conference on Information and Communication Systems (ICICS), Valencia, Spain, 24–26 May 2021; pp. 171–176. [ Google Scholar ]
• Jemai, F.; Hayouni, M.; Baccar, S. Sentiment Analysis Using Machine Learning Algorithms. In Proceedings of the 2021 IEEE International Wireless Communications and Mobile Computing (IWCMC), Harbin, China, 28 June–2 July 2021; pp. 775–779. [ Google Scholar ]
• Ramadhani, A.M.; Goo, H.S. Twitter sentiment analysis using deep learning methods. In Proceedings of the 2017 IEEE 7th International Annual Engineering Seminar (InAES), Yogyakarta, Indonesia, 1–2 August 2017; pp. 1–4. [ Google Scholar ]
• Demirci, G.M.; Keskin, Ş.R.; Doğan, G. Sentiment analysis in Turkish with deep learning. In Proceedings of the 2019 IEEE International Conference on Big Data, Honolulu, HI, USA, 29–31 May 2019; pp. 2215–2221. [ Google Scholar ]
• Raza, G.M.; Butt, Z.S.; Latif, S.; Wahid, A. Sentiment Analysis on COVID Tweets: An Experimental Analysis on the Impact of Count Vectorizer and TF-IDF on Sentiment Predictions using Deep Learning Models. In Proceedings of the 2021 IEEE International Conference on Digital Futures and Transformative Technologies (ICoDT2), Islamabad, Pakistan, 20–21 May 2021; pp. 1–6. [ Google Scholar ]
• Dholpuria, T.; Rana, Y.; Agrawal, C. A sentiment analysis approach through deep learning for a movie review. In Proceedings of the 2018 IEEE 8th International Conference on Communication Systems and Network Technologies (CSNT), Bhopal, India, 24–26 November 2018; pp. 173–181. [ Google Scholar ]
• Harjule, P.; Gurjar, A.; Seth, H.; Thakur, P. Text classification on Twitter data. In Proceedings of the 2020 IEEE 3rd International Conference on Emerging Technologies in Computer Engineering: Machine Learning and Internet of Things (ICETCE), Jaipur, India, 7–8 February 2020; pp. 160–164. [ Google Scholar ]
• Uddin, A.H.; Bapery, D.; Arif, A.S.M. Depression Analysis from Social Media Data in Bangla Language using Long Short Term Memory (LSTM) Recurrent Neural Network Technique. In Proceedings of the 2019 IEEE International Conference on Computer, Communication, Chemical, Materials and Electronic Engineering (IC4ME2), Rajshahi, Bangladesh, 11–12 July 2019; pp. 1–4. [ Google Scholar ]
• Alahmary, R.M.; Al-Dossari, H.Z.; Emam, A.Z. Sentiment analysis of Saudi dialect using deep learning techniques. In Proceedings of the 2019 IEEE International Conference on Electronics, Information, and Communication (ICEIC), Auckland, New Zealand, 22–25 January 2019; pp. 1–6. [ Google Scholar ]
• Yang, Y. Convolutional neural networks with recurrent neural filters. arXiv 2018 , arXiv:1808.09315. [ Google Scholar ]
• Goularas, D.; Kamis, S. Evaluation of deep learning techniques in sentiment analysis from Twitter data. In Proceedings of the 2019 IEEE International Conference on Deep Learning and Machine Learning in Emerging Applications (Deep-ML), Istanbul, Turkey, 26–28 August 2019; pp. 12–17. [ Google Scholar ]
• Hossain, N.; Bhuiyan, M.R.; Tumpa, Z.N.; Hossain, S.A. Sentiment analysis of restaurant reviews using combined CNN-LSTM. In Proceedings of the 2020 IEEE 11th International Conference on Computing, Communication and Networking Technologies (ICCCNT), Kharagpur, India, 1–3 July 2020; pp. 1–5. [ Google Scholar ]
• Tyagi, V.; Kumar, A.; Das, S. Sentiment Analysis on Twitter Data Using Deep Learning approach. In Proceedings of the 2020 IEEE 2nd International Conference on Advances in Computing, Communication Control and Networking (ICACCCN), Greater Noida, India, 18–19 December 2020; pp. 187–190. [ Google Scholar ]
• Rhanoui, M.; Mikram, M.; Yousfi, S.; Barzali, S. A CNN-BiLSTM model for document-level sentiment analysis. Mach. Learn. Knowl. Extr. 2019 , 1 , 832–847. [ Google Scholar ] [ CrossRef ] [ Green Version ]
• Jang, B.; Kim, M.; Harerimana, G.; Kang, S.U.; Kim, J.W. Bi-LSTM model to increase accuracy in text classification: Combining Word2vec CNN and attention mechanism. Appl. Sci. 2020 , 10 , 5841. [ Google Scholar ] [ CrossRef ]
• Chundi, R.; Hulipalled, V.R.; Simha, J. SAEKCS: Sentiment analysis for English–Kannada code switchtext using deep learning techniques. In Proceedings of the 2020 IEEE International Conference on Smart Technologies in Computing, Electrical and Electronics (ICSTCEE), Bengaluru, India, 10–11 July 2020; pp. 327–331. [ Google Scholar ]
• Thinh, N.K.; Nga, C.H.; Lee, Y.S.; Wu, M.L.; Chang, P.C.; Wang, J.C. Sentiment Analysis Using Residual Learning with Simplified CNN Extractor. In Proceedings of the 2019 IEEE International Symposium on Multimedia (ISM), San Diego, CA, USA, 9–11 December 2019; pp. 335–3353. [ Google Scholar ]
• Janardhana, D.; Vijay, C.; Swamy, G.J.; Ganaraj, K. Feature Enhancement Based Text Sentiment Classification using Deep Learning Model. In Proceedings of the 2020 IEEE 5th International Conference on Computing, Communication and Security (ICCCS), Bihar, India, 14–16 October 2020; pp. 1–6. [ Google Scholar ]
• Chowdhury, S.; Rahman, M.L.; Ali, S.N.; Alam, M.J. A RNN Based Parallel Deep Learning Framework for Detecting Sentiment Polarity from Twitter Derived Textual Data. In Proceedings of the 2020 IEEE 11th International Conference on Electrical and Computer Engineering (ICECE), Dhaka, Bangladesh, 17–19 December 2020; pp. 9–12. [ Google Scholar ]
• Vimali, J.; Murugan, S. A Text Based Sentiment Analysis Model using Bi-directional LSTM Networks. In Proceedings of the 2021 IEEE 6th International Conference on Communication and Electronics Systems (ICCES), Coimbatore, India, 8–10 July 2021; pp. 1652–1658. [ Google Scholar ]
• Anbukkarasi, S.; Varadhaganapathy, S. Analyzing Sentiment in Tamil Tweets using Deep Neural Network. In Proceedings of the 2020 IEEE Fourth International Conference on Computing Methodologies and Communication (ICCMC), Erode, India, 11–13 March 2020; pp. 449–453. [ Google Scholar ]
• Kumar, D.A.; Chinnalagu, A. Sentiment and Emotion in Social Media COVID-19 Conversations: SAB-LSTM Approach. In Proceedings of the 2020 IEEE 9th International Conference System Modeling and Advancement in Research Trends (SMART), Moradabad, India, 4–5 December 2020; pp. 463–467. [ Google Scholar ]
• Hossen, M.S.; Jony, A.H.; Tabassum, T.; Islam, M.T.; Rahman, M.M.; Khatun, T. Hotel review analysis for the prediction of business using deep learning approach. In Proceedings of the 2021 IEEE International Conference on Artificial Intelligence and Smart Systems (ICAIS), Coimbatore, India, 25–27 March 2021; pp. 1489–1494. [ Google Scholar ]
• Younas, A.; Nasim, R.; Ali, S.; Wang, G.; Qi, F. Sentiment Analysis of Code-Mixed Roman Urdu-English Social Media Text using Deep Learning Approaches. In Proceedings of the 2020 IEEE 23rd International Conference on Computational Science and Engineering (CSE), Dubai, United Arab Emirates, 12–13 December 2020; pp. 66–71. [ Google Scholar ]
• Dhola, K.; Saradva, M. A Comparative Evaluation of Traditional Machine Learning and Deep Learning Classification Techniques for Sentiment Analysis. In Proceedings of the 2021 IEEE 11th International Conference on Cloud Computing, Data Science & Engineering (Confluence), Uttar Pradesh, India, 28–29 January 2021; pp. 932–936. [ Google Scholar ]
• Tan, K.L.; Lee, C.P.; Anbananthen, K.S.M.; Lim, K.M. RoBERTa-LSTM: A Hybrid Model for Sentiment Analysis with Transformer and Recurrent Neural Network. IEEE Access 2022 , 10 , 21517–21525. [ Google Scholar ] [ CrossRef ]
• Kokab, S.T.; Asghar, S.; Naz, S. Transformer-based deep learning models for the sentiment analysis of social media data. Array 2022 , 14 , 100157. [ Google Scholar ] [ CrossRef ]
• AlBadani, B.; Shi, R.; Dong, J.; Al-Sabri, R.; Moctard, O.B. Transformer-based graph convolutional network for sentiment analysis. Appl. Sci. 2022 , 12 , 1316. [ Google Scholar ] [ CrossRef ]
• Tiwari, D.; Nagpal, B. KEAHT: A knowledge-enriched attention-based hybrid transformer model for social sentiment analysis. New Gener. Comput. 2022 , 40 , 1165–1202. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Tesfagergish, S.G.; Kapočiūtė-Dzikienė, J.; Damaševičius, R. Zero-shot emotion detection for semi-supervised sentiment analysis using sentence transformers and ensemble learning. Appl. Sci. 2022 , 12 , 8662. [ Google Scholar ] [ CrossRef ]
• Maghsoudi, A.; Nowakowski, S.; Agrawal, R.; Sharafkhaneh, A.; Kunik, M.E.; Naik, A.D.; Xu, H.; Razjouyan, J. Sentiment Analysis of Insomnia-Related Tweets via a Combination of Transformers Using Dempster-Shafer Theory: Pre–and Peri–COVID-19 Pandemic Retrospective Study. J. Med Internet Res. 2022 , 24 , e41517. [ Google Scholar ] [ CrossRef ]
• Jing, H.; Yang, C. Chinese text sentiment analysis based on transformer model. In Proceedings of the 2022 IEEE 3rd International Conference on Electronic Communication and Artificial Intelligence (IWECAI), Sanya, China, 14–16 January 2022; pp. 185–189. [ Google Scholar ]
• Alrehili, A.; Albalawi, K. Sentiment analysis of customer reviews using ensemble method. In Proceedings of the 2019 IEEE International Conference on Computer and Information Sciences (ICCIS), Aljouf, Saudi Arabia, 3–4 April 2019; pp. 1–6. [ Google Scholar ]
• Bian, W.; Wang, C.; Ye, Z.; Yan, L. Emotional Text Analysis Based on Ensemble Learning of Three Different Classification Algorithms. In Proceedings of the 2019 IEEE 10th IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications (IDAACS), Metz, France, 18–21 September 2019; Volume 2, pp. 938–941. [ Google Scholar ]
• Gifari, M.K.; Lhaksmana, K.M.; Dwifebri, P.M. Sentiment Analysis on Movie Review using Ensemble Stacking Model. In Proceedings of the 2021 IEEE International Conference Advancement in Data Science, E-learning and Information Systems (ICADEIS), Bali, Indonesia, 13–14 October 2021; pp. 1–5. [ Google Scholar ]
• Parveen, R.; Shrivastava, N.; Tripathi, P. Sentiment Classification of Movie Reviews by Supervised Machine Learning Approaches Using Ensemble Learning & Voted Algorithm. In Proceedings of the IEEE 2nd International Conference on Data, Engineering and Applications (IDEA), Bhopal, India, 28–29 February 2020; pp. 1–6. [ Google Scholar ]
• Aziz, R.H.H.; Dimililer, N. Twitter Sentiment Analysis using an Ensemble Weighted Majority Vote Classifier. In Proceedings of the 2020 IEEE International Conference on Advanced Science and Engineering (ICOASE), Duhok, Iraq, 23–24 December 2020; pp. 103–109. [ Google Scholar ]
• Varshney, C.J.; Sharma, A.; Yadav, D.P. Sentiment analysis using ensemble classification technique. In Proceedings of the 2020 IEEE Students Conference on Engineering & Systems (SCES), Prayagraj, India, 10–12 July 2020; pp. 1–6. [ Google Scholar ]
• Athar, A.; Ali, S.; Sheeraz, M.M.; Bhattachariee, S.; Kim, H.C. Sentimental Analysis of Movie Reviews using Soft Voting Ensemble-based Machine Learning. In Proceedings of the 2021 IEEE Eighth International Conference on Social Network Analysis, Management and Security (SNAMS), Gandia, Spain, 6–9 December 2021; pp. 1–5. [ Google Scholar ]
• Nguyen, H.Q.; Nguyen, Q.U. An ensemble of shallow and deep learning algorithms for Vietnamese Sentiment Analysis. In Proceedings of the 2018 IEEE 5th NAFOSTED Conference on Information and Computer Science (NICS), Ho Chi Minh City, Vietnam, 23–24 November 2018; pp. 165–170. [ Google Scholar ]
• Kamruzzaman, M.; Hossain, M.; Imran, M.R.I.; Bakchy, S.C. A Comparative Analysis of Sentiment Classification Based on Deep and Traditional Ensemble Machine Learning Models. In Proceedings of the 2021 IEEE International Conference on Science & Contemporary Technologies (ICSCT), Dhaka, Bangladesh, 5–7 August 2021; pp. 1–5. [ Google Scholar ]
• Al Wazrah, A.; Alhumoud, S. Sentiment Analysis Using Stacked Gated Recurrent Unit for Arabic Tweets. IEEE Access 2021 , 9 , 137176–137187. [ Google Scholar ] [ CrossRef ]
• Tan, K.L.; Lee, C.P.; Lim, K.M.; Anbananthen, K.S.M. Sentiment Analysis with Ensemble Hybrid Deep Learning Model. IEEE Access 2022 , 10 , 103694–103704. [ Google Scholar ] [ CrossRef ]
• Maas, A.; Daly, R.E.; Pham, P.T.; Huang, D.; Ng, A.Y.; Potts, C. Learning word vectors for sentiment analysis. In Proceedings of the IEEE 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies, Portland, OR, USA, 19–24 June 2011; pp. 142–150. [ Google Scholar ]
• Go, A.; Bhayani, R.; Huang, L. Twitter sentiment classification using distant supervision. CS224N Proj. Rep. Stanf. 2009 , 1 , 2009. [ Google Scholar ]
• Rosenthal, S.; Farra, N.; Nakov, P. SemEval-2017 task 4: Sentiment analysis in Twitter. arXiv 2019 , arXiv:1912.00741. [ Google Scholar ]

Click here to enlarge figure

Share and Cite

Tan, K.L.; Lee, C.P.; Lim, K.M. A Survey of Sentiment Analysis: Approaches, Datasets, and Future Research. Appl. Sci. 2023 , 13 , 4550. https://doi.org/10.3390/app13074550

Tan KL, Lee CP, Lim KM. A Survey of Sentiment Analysis: Approaches, Datasets, and Future Research. Applied Sciences . 2023; 13(7):4550. https://doi.org/10.3390/app13074550

Tan, Kian Long, Chin Poo Lee, and Kian Ming Lim. 2023. "A Survey of Sentiment Analysis: Approaches, Datasets, and Future Research" Applied Sciences 13, no. 7: 4550. https://doi.org/10.3390/app13074550

Article Metrics

Article access statistics, further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals

• Methodology
• Open access
• Published: 16 June 2015

Sentiment analysis using product review data

• Xing Fang 1 &
• Justin Zhan 1  

Journal of Big Data volume  2 , Article number:  5 ( 2015 ) Cite this article

187k Accesses

402 Citations

14 Altmetric

Metrics details

Sentiment analysis or opinion mining is one of the major tasks of NLP (Natural Language Processing). Sentiment analysis has gain much attention in recent years. In this paper, we aim to tackle the problem of sentiment polarity categorization, which is one of the fundamental problems of sentiment analysis. A general process for sentiment polarity categorization is proposed with detailed process descriptions. Data used in this study are online product reviews collected from Amazon.com. Experiments for both sentence-level categorization and review-level categorization are performed with promising outcomes. At last, we also give insight into our future work on sentiment analysis.

Introduction

Sentiment is an attitude, thought, or judgment prompted by feeling. Sentiment analysis [ 1 - 8 ], which is also known as opinion mining, studies people’s sentiments towards certain entities. Internet is a resourceful place with respect to sentiment information. From a user’s perspective, people are able to post their own content through various social media, such as forums, micro-blogs, or online social networking sites. From a researcher’s perspective, many social media sites release their application programming interfaces (APIs), prompting data collection and analysis by researchers and developers. For instance, Twitter currently has three different versions of APIs available [ 9 ], namely the REST API, the Search API, and the Streaming API. With the REST API, developers are able to gather status data and user information; the Search API allows developers to query specific Twitter content, whereas the Streaming API is able to collect Twitter content in realtime. Moreover, developers can mix those APIs to create their own applications. Hence, sentiment analysis seems having a strong fundament with the support of massive online data.

However, those types of online data have several flaws that potentially hinder the process of sentiment analysis. The first flaw is that since people can freely post their own content, the quality of their opinions cannot be guaranteed. For example, instead of sharing topic-related opinions, online spammers post spam on forums. Some spam are meaningless at all, while others have irrelevant opinions also known as fake opinions [ 10 - 12 ]. The second flaw is that ground truth of such online data is not always available. A ground truth is more like a tag of a certain opinion, indicating whether the opinion is positive, negative, or neutral. The Stanford Sentiment 140 Tweet Corpus [ 13 ] is one of the datasets that has ground truth and is also public available. The corpus contains 1.6 million machine-tagged Twitter messages. Each message is tagged based on the emoticons (☺as positive, ☹ as negative) discovered inside the message.

Data used in this paper is a set of product reviews collected from Amazon [ 14 ], between February and April, 2014. The aforementioned flaws have been somewhat overcome in the following two ways: First, each product review receives inspections before it can be posted a . Second, each review must have a rating on it that can be used as the ground truth. The rating is based on a star-scaled system, where the highest rating has 5 stars and the lowest rating has only 1 star (Figure 1 ).

Rating System for Amazon.com.

This paper tackles a fundamental problem of sentiment analysis, namely sentiment polarity categorization [ 15 - 21 ]. Figure 2 is a flowchart that depicts our proposed process for categorization as well as the outline of this paper. Our contributions mainly fall into Phase 2 and 3. In Phase 2: 1) An algorithm is proposed and implemented for negation phrases identification; 2) A mathematical approach is proposed for sentiment score computation; 3) A feature vector generation method is presented for sentiment polarity categorization. In Phase 3: 1) Two sentiment polarity categorization experiments are respectively performed based on sentence level and review level; 2) Performance of three classification models are evaluated and compared based on their experimental results.

Sentiment Polarity Categorization Process.

The rest of this paper is organized as follows: In section ‘ Background and literature review ’, we provide a brief review towards some related work on sentiment analysis. Software package and classification models used in this study are presented in section ‘ Methods ’. Our detailed approaches for sentiment analysis are proposed in section ‘ Background and literature review ’. Experimental results are presented in section ‘ Results and discussion ’. Discussion and future work is presented in section ‘ Review-level categorization ’. Section ‘ Conclusion ’ concludes the paper.

Background and literature review

One fundamental problem in sentiment analysis is categorization of sentiment polarity [ 6 , 22 - 25 ]. Given a piece of written text, the problem is to categorize the text into one specific sentiment polarity, positive or negative (or neutral). Based on the scope of the text, there are three levels of sentiment polarity categorization, namely the document level, the sentence level, and the entity and aspect level [ 26 ]. The document level concerns whether a document, as a whole, expresses negative or positive sentiment, while the sentence level deals with each sentence’s sentiment categorization; The entity and aspect level then targets on what exactly people like or dislike from their opinions.

Since reviews of much work on sentiment analysis have already been included in [ 26 ], in this section, we will only review some previous work, upon which our research is essentially based. Hu and Liu [ 27 ] summarized a list of positive words and a list of negative words, respectively, based on customer reviews. The positive list contains 2006 words and the negative list has 4783 words. Both lists also include some misspelled words that are frequently present in social media content. Sentiment categorization is essentially a classification problem, where features that contain opinions or sentiment information should be identified before the classification. For feature selection, Pang and Lee [ 5 ] suggested to remove objective sentences by extracting subjective ones. They proposed a text-categorization technique that is able to identify subjective content using minimum cut. Gann et al. [ 28 ] selected 6,799 tokens based on Twitter data, where each token is assigned a sentiment score, namely TSI(Total Sentiment Index), featuring itself as a positive token or a negative token. Specifically, a TSI for a certain token is computed as:

where p is the number of times a token appears in positive tweets and n is the number of times a token appears in negative tweets. \(\frac {tp}{tn}\) is the ratio of total number of positive tweets over total number of negative tweets.

Research design and methdology

Data collection.

Data used in this paper is a set of product reviews collected from amazon.com. From February to April 2014, we collected, in total, over 5.1 millions of product reviews b in which the products belong to 4 major categories: beauty, book, electronic, and home (Figure 3 (a)). Those online reviews were posted by over 3.2 millions of reviewers (customers) towards 20,062 products. Each review includes the following information: 1) reviewer ID; 2) product ID; 3) rating; 4) time of the review; 5) helpfulness; 6) review text. Every rating is based on a 5-star scale(Figure 3 (b)), resulting all the ratings to be ranged from 1-star to 5-star with no existence of a half-star or a quarter-star.

Data collection (a) Data based on product categories (b) Data based on review categories.

Sentiment sentences extraction and POS tagging

It is suggested by Pang and Lee [ 5 ] that all objective content should be removed for sentiment analysis. Instead of removing objective content, in our study, all subjective content was extracted for future analysis. The subjective content consists of all sentiment sentences. A sentiment sentence is the one that contains, at least, one positive or negative word. All of the sentences were firstly tokenized into separated English words.

Every word of a sentence has its syntactic role that defines how the word is used. The syntactic roles are also known as the parts of speech. There are 8 parts of speech in English: the verb, the noun, the pronoun, the adjective, the adverb, the preposition, the conjunction, and the interjection. In natural language processing, part-of-speech (POS) taggers [ 29 - 31 ] have been developed to classify words based on their parts of speech. For sentiment analysis, a POS tagger is very useful because of the following two reasons: 1) Words like nouns and pronouns usually do not contain any sentiment. It is able to filter out such words with the help of a POS tagger; 2) A POS tagger can also be used to distinguish words that can be used in different parts of speech. For instance, as a verb, “enhanced" may conduct different amount of sentiment as being of an adjective. The POS tagger used for this research is a max-entropy POS tagger developed for the Penn Treebank Project [ 31 ]. The tagger is able to provide 46 different tags indicating that it can identify more detailed syntactic roles than only 8. As an example, Table 1 is a list of all tags for verbs that has been included in the POS tagger.

Each sentence was then tagged using the POS tagger. Given the enormous amount of sentences, a Python program that is able to run in parallel was written in order to improve the speed of tagging. As a result, there are over 25 million adjectives, over 22 million adverbs, and over 56 million verbs tagged out of all the sentiment sentences, because adjectives, adverbs, and verbs are words that mainly convey sentiment.

Negation phrases identification

Words such as adjectives and verbs are able to convey opposite sentiment with the help of negative prefixes. For instance, consider the following sentence that was found in an electronic device’s review: “The built in speaker also has its uses but so far nothing revolutionary." The word, “revolutionary" is a positive word according to the list in [ 27 ]. However, the phrase “nothing revolutionary" gives more or less negative feelings. Therefore, it is crucial to identify such phrases. In this work, there are two types of phrases have been identified, namely negation-of-adjective (NOA) and negation-of-verb (NOV).

Most common negative prefixes such as not, no, or nothing are treated as adverbs by the POS tagger. Hence, we propose Algorithm 1 for the phrases identification. The algorithm was able to identify 21,586 different phrases with total occurrence of over 0.68 million, each of which has a negative prefix. Table 2 lists top 5 NOA and NOV phrases based on occurrence, respectively.

Sentiment score computation for sentiment tokens

A sentiment token is a word or a phrase that conveys sentiment. Given those sentiment words proposed in [ 27 ], a word token consists of a positive (negative) word and its part-of-speech tag. In total, we selected 11,478 word tokens with each of them that occurs at least 30 times throughout the dataset. For phrase tokens, 3,023 phrases were selected of the 21,586 identified sentiment phrases, which each of the 3,023 phrases also has an occurrence that is no less than 30. Given a token t , the formula for t ’s sentiment score (SS) computation is given as:

O c c u r r e n c e i ( t ) is t ’s number of occurrence in i -star reviews, where i =1,...,5. According to Figure 3 , our dataset is not balanced indicating that different number of reviews were collected for each star level. Since 5-star reviews take a majority amount through the entire dataset, we hereby introduce a ratio, γ 5, i , which is defined as:

In equation 3 , the numerator is the number of 5-star reviews and the denominator is the number of i -star reviews, where i =1,...,5. Therefore, if the dataset were balanced, γ 5, i would be set to 1 for every i . Consequently, every sentiment score should fall into the interval of [1,5]. For positive word tokens, we expect that the median of their sentiment scores should exceed 3, which is the point of being neutral according to Figure 1 . For negative word tokens, it is to expect that the median should be less than 3.

As a result, the sentiment score information for positive word tokens is showing in Figure 4 (a). The histogram chart describes the distribution of scores while the box-plot chart shows that the median is above 3. Similarly, the box-plot chart in Figure 4 (b) shows that the median of sentiment scores for negative word tokens is lower than 3. In fact, both the mean and the median of positive word tokens do exceed 3, and both values are lower than 3, for negative word tokens (Table 3 ).

Sentiment score information for word tokens (a) Positive word tokens (b) Negative word tokens.

The ground truth labels

The process of sentiment polarity categorization is twofold: sentence-level categorization and review-level categorization. Given a sentence, the goal of sentence-level categorization is to classify it as positive or negative in terms of the sentiment that it conveys. Training data for this categorization process require ground truth tags, indicating the positiveness or negativeness of a given sentence. However, ground truth tagging becomes a really challenging problem, due to the amount of data that we have. Since manually tagging each sentence is infeasible, a machine tagging approach is then adopted as a solution. The approach implements a bag-of-word model that simply counts the appearance of positive or negative (word) tokens for every sentence. If there are more positive tokens than negative ones, the sentence will be tagged as positive, and vice versa. This approach is similar to the one used for tagging the Sentiment 140 Tweet Corpus. Training data for review-level categorization already have ground truth tags, which are the star-scaled ratings.

Feature vector formation

Sentiment tokens and sentiment scores are information extracted from the original dataset. They are also known as features, which will be used for sentiment categorization. In order to train the classifiers, each entry of training data needs to be transformed to a vector that contains those features, namely a feature vector. For the sentence-level (review-level) categorization, a feature vector is formed based on a sentence (review). One challenge is to control each vector’s dimensionality. The challenge is actually twofold: Firstly, a vector should not contain an abundant amount (thousands or hundreds) of features or values of a feature, because of the curse of dimensionality [ 32 ]; secondly, every vector should have the same number of dimensions, in order to fit the classifiers. This challenge particularly applies to sentiment tokens: On one hand, there are 11,478 word tokens as well as 3,023 phrase tokens; On the other hand, vectors cannot be formed by simply including the tokens appeared in a sentence (or a review), because different sentences (or reviews) tend to have different amount of tokens, leading to the consequence that the generated vectors are in different dimensions.

Since we only concern each sentiment token’s appearance inside a sentence or a review,to overcome the challenge, two binary strings are used to represent each token’s appearance. One string with 11,478 bits is used for word tokens, while the other one with a bit-length of 3,023 is applied for phrase tokens. For instance, if the i th word (phrase) token appears, the word (phrase) string’s i th bit will be flipped from “0" to “1". Finally, instead of directly saving the flipped strings into a feature vector, a hash value of each string is computed using Python’s built-in hash function and is saved. Hence, a sentence-level feature vector totally has four elements: two hash values computed based on the flipped binary strings, an averaged sentiment score, and a ground truth label. Comparatively, one more element is exclusively included in review-level vectors. Given a review, if there are m positive sentences and n negative sentences, the value of the element is computed as: −1× m +1× n .

Results and discussion

Evaluation methods.

Performance of each classification model is estimated base on its averaged F1-score ( 4 ):

where P i is the precision of the i th class, R i is the recall of the i th class, and n is the number of classes. P i and R i are evaluated using 10-fold cross validation. A 10-fold cross validation is applied as follows: A dataset is partitioned into 10 equal size subsets, each of which consists of 10 positive class vectors and 10 negative class vectors. Of the 10 subsets, a single subset is retained as the validation data for testing the classification model, and the remaining 9 subsets are used as training data. The cross-validation process is then repeated 10 times, with each of the 10 subsets used exactly once as the validation data. The 10 results from the folds are then averaged to produce a single estimation. Since training data are labeled under two classes (positive and negative) for the sentence-level categorization, ROC (Receiver Operating Characteristic) curves are also plotted for a better performance comparison.

Sentence-level categorization

Result on manually-labeled sentences.

200 feature vectors are formed based on the 200 manually-labeled sentences. As a result, the classification models show the same level of performance based on their F1-scores, where the three scores all take a same value of 0.85. With the help of the ROC curves (Figure 5 ), it is clear to see that all three models performed quite well for testing data that have high posterior probability. (A posterior probability of a testing data point, A , is estimated by the classification model as the probability that A will be classified as positive, denoted as P (+| A ).) As the probability getting lower, the Naïve Bayesain classifier outperforms the SVM classifier, with a larger area under curve. In general, the Random Forest model performs the best.

ROC curves based on the manually labeled set.

Result on machine-labeled sentences

2-million feature vectors (1 million with positive labels and 1 million with negative labels) are generated from 2-million machine-labeled sentences, known as the complete set. Four subsets are obtained from the complete set, with subset A contains 200 vectors, subset B contains 2,000 vectors, subset C contains 20,000 vectors, and subset D contains 200,000 vectors, respectively. The amount of vectors with positive labels equals the amount of vectors with negative labels for every subset. Performance of the classification models is then evaluated based on five different vector sets (four subsets and one complete set, Figure 6 ).

F1 scores of sentence-level categorization.

While the models are getting more training data, their F1 scores are all increasing. The SVM model takes the most significant enhancement from 0.61 to 0.94 as its training data increased from 180 to 1.8 million. The model outperforms the Naïve Bayesain model and becomes the 2nd best classifier, on subset C and the full set. The Random Forest model again performs the best for datasets on all scopes. Figure 7 shows the ROC curves plotted based on the result of the full set.

ROC curves based on the complete set.

Review-level categorization

3-million feature vectors are formed for the categorization. Vectors generated from reviews that have at least 4-star ratings are labeled as positive, while vectors labeled as negative are generated from 1-star and 2-star reviews. 3-star reviews are used to prepare neutral class vectors. As a result, this complete set of vectors are uniformly labeled into three classes, positive, neutral, and negative. In addition, three subsets are obtained from the complete set, with subset A contains 300 vectors, subset B contains 3,000 vectors, subset C contains 30,000 vectors, and subset D contains 300,000 vectors, respectively.

Figure 8 shows the F1 scores obtained on different sizes of vector sets. It can be clearly observed that both the SVM model and the Naïve Bayesain model are identical in terms of their performances. Both models are generally superior than the Random Forest model on all vector sets. However, neither of the models can reach the same level of performance when they are used for sentence-level categorization, due to their relative low performances on neutral class.

F1 scores of review-level categorization.

The experimental result is promising, both in terms of the sentence-level categorization and the review-level categorization. It was observed that the averaged sentiment score is a strong feature by itself, since it is able to achieve an F1 score over 0.8 for the sentence-level categorization with the complete set. For the review-level categorization with the complete set, the feature is capable of producing an F1 score that is over 0.73. However, there are still couple of limitations to this study. The first one is that the review-level categorization becomes difficult if we want to classify reviews to their specific star-scaled ratings. In other words, F1 scores obtained from such experiments are fairly low, with values lower than 0.5. The second limitation is that since our sentiment analysis scheme proposed in this study relies on the occurrence of sentiment tokens, the scheme may not work well for those reviews that purely contain implicit sentiments. An implicit sentiment is usually conveyed through some neutral words, making judgement of its sentiment polarity difficult. For example, sentence like “Item as described.", which frequently appears in positive reviews, consists of only neutral words.

With those limitations in mind, our future work is to focus on solving those issues. Specifically, more features will be extracted and grouped into feature vectors to improve review-level categorizations. For the issue of implicit sentiment analysis, our next step is to be able to detect the existence of such sentiment within the scope of a particular product. More future work includes testing our categorization scheme using other datasets.

Sentiment analysis or opinion mining is a field of study that analyzes people’s sentiments, attitudes, or emotions towards certain entities. This paper tackles a fundamental problem of sentiment analysis, sentiment polarity categorization. Online product reviews from Amazon.com are selected as data used for this study. A sentiment polarity categorization process (Figure 2 ) has been proposed along with detailed descriptions of each step. Experiments for both sentence-level categorization and review-level categorization have been performed.

Software used for this study is scikit-learn [ 33 ], an open source machine learning software package in Python. The classification models selected for categorization are: Naïve Bayesian, Random Forest, and Support Vector Machine [ 32 ].

Naïve Bayesian classifier

The Naïve Bayesian classifier works as follows: Suppose that there exist a set of training data, D , in which each tuple is represented by an n -dimensional feature vector, X = x 1 , x 2 ,.., x n , indicating n measurements made on the tuple from n attributes or features. Assume that there are m classes, C 1 , C 2 ,..., C m . Given a tuple X , the classifier will predict that X belongs to C i if and only if: P ( C i | X )> P ( C j | X ), where i , j ∈ [1, m ] a n d i ≠ j . P ( C i | X ) is computed as:

Random forest

The random forest classifier was chosen due to its superior performance over a single decision tree with respect to accuracy. It is essentially an ensemble method based on bagging. The classifier works as follows: Given D , the classifier firstly creates k bootstrap samples of D , with each of the samples denoting as D i . A D i has the same number of tuples as D that are sampled with replacement from D . By sampling with replacement, it means that some of the original tuples of D may not be included in D i , whereas others may occur more than once. The classifier then constructs a decision tree based on each D i . As a result, a “forest" that consists of k decision trees is formed. To classify an unknown tuple, X , each tree returns its class prediction counting as one vote. The final decision of X ’s class is assigned to the one that has the most votes.

The decision tree algorithm implemented in scikit-learn is CART (Classification and Regression Trees). CART uses Gini index for its tree induction. For D , the Gini index is computed as:

where p i is the probability that a tuple in D belongs to class C i . The Gini index measures the impurity of D . The lower the index value is, the better D was partitioned. For the detailed descriptions of CART, please see [ 32 ].

Support vector machine

Support vector machine (SVM) is a method for the classification of both linear and nonlinear data. If the data is linearly separable, the SVM searches for the linear optimal separating hyperplane (the linear kernel), which is a decision boundary that separates data of one class from another. Mathematically, a separating hyperplane can be written as: W · X + b =0, where W is a weight vector and W = w 1 , w 2,..., w n . X is a training tuple. b is a scalar. In order to optimize the hyperplane, the problem essentially transforms to the minimization of ∥ W ∥ , which is eventually computed as: \(\sum \limits _{i=1}^{n} \alpha _{i} y_{i} x_{i}\) , where α i are numeric parameters, and y i are labels based on support vectors, X i . That is: if y i =1 then \(\sum \limits _{i=1}^{n} w_{i}x_{i} \geq 1\) ; if y i =−1 then \(\sum \limits _{i=1}^{n} w_{i}x_{i} \geq -1\) .

If the data is linearly inseparable, the SVM uses nonlinear mapping to transform the data into a higher dimension. It then solve the problem by finding a linear hyperplane. Functions to perform such transformations are called kernel functions. The kernel function selected for our experiment is the Gaussian Radial Basis Function (RBF):

where X i are support vectors, X j are testing tuples, and γ is a free parameter that uses the default value from scikit-learn in our experiment. Figure 9 shows a classification example of SVM based on the linear kernel and the RBF kernel.

A Classification Example of SVM.

a Even though there are papers talking about spam on Amazon.com, we still contend that it is a relatively spam-free website in terms of reviews because of the enforcement of its review inspection process.

b The product review data used for this work can be downloaded at: http://www.itk.ilstu.edu/faculty/xfang13/amazon_data.htm .

Kim S-M, Hovy E (2004) Determining the sentiment of opinions In: Proceedings of the 20th international conference on Computational Linguistics, page 1367.. Association for Computational Linguistics, Stroudsburg, PA, USA.

Google Scholar  

Liu B (2010) Sentiment analysis and subjectivity In: Handbook of Natural Language Processing, Second Edition.. Taylor and Francis Group, Boca.

Liu B, Hu M, Cheng J (2005) Opinion observer: Analyzing and comparing opinions on the web In: Proceedings of the 14th International Conference on World Wide Web, WWW ’05, 342–351.. ACM, New York, NY, USA.

Chapter   Google Scholar  

Pak A, Paroubek P (2010) Twitter as a corpus for sentiment analysis and opinion mining In: Proceedings of the Seventh conference on International Language Resources and Evaluation.. European Languages Resources Association, Valletta, Malta.

Pang B, Lee L (2004) A sentimental education: Sentiment analysis using subjectivity summarization based on minimum cuts In: Proceedings of the 42Nd Annual Meeting on Association for Computational Linguistics, ACL ’04.. Association for Computational Linguistics, Stroudsburg, PA, USA.

Pang B, Lee L (2008) Opinion mining and sentiment analysis. Found Trends Inf Retr2(1-2): 1–135.

Article   Google Scholar  

Turney PD (2002) Thumbs up or thumbs down?: Semantic orientation applied to unsupervised classification of reviews In: Proceedings of the 40th Annual Meeting on Association for Computational Linguistics, ACL ’02, 417–424.. Association for Computational Linguistics, Stroudsburg, PA, USA.

Whitelaw C, Garg N, Argamon S (2005) Using appraisal groups for sentiment analysis In: Proceedings of the 14th ACM International Conference on Information and Knowledge Management, CIKM ’05, 625–631.. ACM, New York, NY, USA.

Twitter (2014) Twitter apis. https://dev.twitter.com/start .

Liu B (2014) The science of detecting fake reviews. http://content26.com/blog/bing-liu-the-science-of-detecting-fake-reviews/ .

Jindal N, Liu B (2008) Opinion spam and analysis In: Proceedings of the 2008 International Conference on, Web Search and Data Mining, WSDM ’08, 219–230.. ACM, New York, NY, USA.

Mukherjee A, Liu B, Glance N (2012) Spotting fake reviewer groups in consumer reviews In: Proceedings of the 21st, International Conference on World Wide Web, WWW ’12, 191–200.. ACM, New York, NY, USA.

Stanford (2014) Sentiment 140. http://www.sentiment140.com/ .

www.amazon.com.

Go A, Bhayani R, Huang L (2009) Twitter sentiment classification using distant supervision, 1–12.. CS224N Project Report, Stanford.

Lin Y, Zhang J, Wang X, Zhou A (2012) An information theoretic approach to sentiment polarity classification In: Proceedings of the 2Nd Joint WICOW/AIRWeb Workshop on Web Quality, WebQuality ’12, 35–40.. ACM, New York, NY, USA.

Sarvabhotla K, Pingali P, Varma V (2011) Sentiment classification: a lexical similarity based approach for extracting subjectivity in documents. Inf Retrieval14(3): 337–353.

Wilson T, Wiebe J, Hoffmann P (2005) Recognizing contextual polarity in phrase-level sentiment analysis In: Proceedings of the conference on human language technology and empirical methods in natural language processing, 347–354.. Association for Computational Linguistics, Stroudsburg, PA, USA.

Yu H, Hatzivassiloglou V (2003) Towards answering opinion questions: Separating facts from opinions and identifying the polarity of opinion sentences In: Proceedings of the 2003 conference on, Empirical methods in natural language processing, 129–136.. Association for Computational Linguistics, Stroudsburg, PA, USA.

Zhang Y, Xiang X, Yin C, Shang L (2013) Parallel sentiment polarity classification method with substring feature reduction In: Trends and Applications in Knowledge Discovery and Data Mining, volume 7867 of Lecture Notes in Computer Science, 121–132.. Springer Berlin Heidelberg, Heidelberg, Germany.

Zhou S, Chen Q, Wang X (2013) Active deep learning method for semi-supervised sentiment classification. Neurocomputing120(0): 536–546. Image Feature Detection and Description.

Chesley P, Vincent B, Xu L, Srihari RK (2006) Using verbs and adjectives to automatically classify blog sentiment. Training580(263): 233.

Choi Y, Cardie C (2009) Adapting a polarity lexicon using integer linear programming for domain-specific sentiment classification In: Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing: Volume 2 - Volume 2, EMNLP ’09, 590–598.. Association for Computational Linguistics, Stroudsburg, PA, USA.

Jiang L, Yu M, Zhou M, Liu X, Zhao T (2011) Target-dependent twitter sentiment classification In: Proceedings of the 49th, Annual Meeting of the Association for Computational Linguistics: Human Language Technologies-Volume 1, 151–160.. Association for Computational Linguistics, Stroudsburg, PA, USA.

Tan LK-W, Na J-C, Theng Y-L, Chang K (2011) Sentence-level sentiment polarity classification using a linguistic approach In: Digital Libraries: For Cultural Heritage, Knowledge Dissemination, and Future Creation, 77–87.. Springer, Heidelberg, Germany.

Liu B (2012) Sentiment Analysis and Opinion Mining. Synthesis Lectures on Human Language Technologies. Morgan & Claypool Publishers.

Hu M, Liu B (2004) Mining and summarizing customer reviews In: Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining, 168–177.. ACM, New York, NY, USA.

Gann W-JK, Day J, Zhou S (2014) Twitter analytics for insider trading fraud detection system In: Proceedings of the sencond ASE international conference on Big Data.. ASE.

Roth D, Zelenko D (1998) Part of speech tagging using a network of linear separators In: Coling-Acl, The 17th International Conference on Computational Linguistics, 1136–1142.

Kristina T (2003) Stanford log-linear part-of-speech tagger. http://nlp.stanford.edu/software/tagger.shtml .

Marcus M (1996) Upenn part of speech tagger. http://www.cis.upenn.edu/~treebank/home.html .

Han J, Kamber M, Pei J (2006) Data Mining: Concepts and Techniques, Second Edition (The Morgan Kaufmann Series in Data Management Systems), 2nd ed.. Morgan Kaufmann, San Francisco, CA, USA.

(2014) Scikit-learn. http://scikit-learn.org/stable/ .

Download references

Acknowledgements

This research was partially supported by the following grants: NSF No. 1137443, NSF No. 1247663, NSF No. 1238767, DoD No. W911NF-13-0130, DoD No. W911NF-14-1-0119, and the Data Science Fellowship Award by the National Consortium for Data Science.

Author information

Authors and affiliations.

Department of Computer Science, North Carolina A&T State University, Greensboro, NC, USA

Xing Fang & Justin Zhan

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Xing Fang .

Additional information

Competing interests.

The authors declare that they have no competing interests.

Authors’ contributions

XF performed the primary literature review, data collection, experiments, and also drafted the manuscript. JZ worked with XF to develop the articles framework and focus. All authors read and approved the final manuscript.

Authors’ information

Xing Fang is a Ph.D. candidate at the Department of Computer Science, North Carolina A&T State University. His research interests include social computing, machine learning, and natural language processing. Mr. Fang holds one Master’s degree in computer science from North Carolina A&T State University, and one Baccalaureate degree in electronic engineering from Northwestern Polytechnical University, Xi’an, China.

Dr. Justin Zhan is an associate professor at the Department of Computer Science, North Carolina A&T State University. He has previously been a faculty member at Carnegie Mellon University and National Center for the Protection of Financial Infrastructure in Dakota State University. His research interests include Big Data, Information Assurance, Social Computing, and Health Science.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License ( https://creativecommons.org/licenses/by/4.0 ), which permits use, duplication, adaptation, distribution, and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Reprints and permissions

About this article

Cite this article.

Fang, X., Zhan, J. Sentiment analysis using product review data. Journal of Big Data 2 , 5 (2015). https://doi.org/10.1186/s40537-015-0015-2

Download citation

Received : 12 January 2015

Accepted : 20 April 2015

Published : 16 June 2015

DOI : https://doi.org/10.1186/s40537-015-0015-2

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Sentiment analysis; Sentiment polarity categorization; Natural language processing; Product reviews

Applying Transformers and Aspect-based Sentiment Analysis approaches on Sarcasm Detection

Taha Shangipour ataei , Soroush Javdan , Behrouz Minaei-Bidgoli

Export citation

• Preformatted

Markdown (Informal)

[Applying Transformers and Aspect-based Sentiment Analysis approaches on Sarcasm Detection](https://aclanthology.org/2020.figlang-1.9) (Shangipour ataei et al., Fig-Lang 2020)

• Applying Transformers and Aspect-based Sentiment Analysis approaches on Sarcasm Detection (Shangipour ataei et al., Fig-Lang 2020)
• Taha Shangipour ataei, Soroush Javdan, and Behrouz Minaei-Bidgoli. 2020. Applying Transformers and Aspect-based Sentiment Analysis approaches on Sarcasm Detection . In Proceedings of the Second Workshop on Figurative Language Processing , pages 67–71, Online. Association for Computational Linguistics.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• JMIR Public Health Surveill
• v.4(2); Apr-Jun 2018

Sentiment Analysis of Health Care Tweets: Review of the Methods Used

Sunir gohil.

1 Imperial College London, Department of Surgery and Cancer, London, United Kingdom

Sabine Vuik

Twitter is a microblogging service where users can send and read short 140-character messages called “tweets.” There are several unstructured, free-text tweets relating to health care being shared on Twitter, which is becoming a popular area for health care research. Sentiment is a metric commonly used to investigate the positive or negative opinion within these messages. Exploring the methods used for sentiment analysis in Twitter health care research may allow us to better understand the options available for future research in this growing field.

The first objective of this study was to understand which tools would be available for sentiment analysis of Twitter health care research, by reviewing existing studies in this area and the methods they used. The second objective was to determine which method would work best in the health care settings, by analyzing how the methods were used to answer specific health care questions, their production, and how their accuracy was analyzed.

A review of the literature was conducted pertaining to Twitter and health care research, which used a quantitative method of sentiment analysis for the free-text messages (tweets). The study compared the types of tools used in each case and examined methods for tool production, tool training, and analysis of accuracy.

A total of 12 papers studying the quantitative measurement of sentiment in the health care setting were found. More than half of these studies produced tools specifically for their research, 4 used open source tools available freely, and 2 used commercially available software. Moreover, 4 out of the 12 tools were trained using a smaller sample of the study’s final data. The sentiment method was trained against, on an average, 0.45% (2816/627,024) of the total sample data. One of the 12 papers commented on the analysis of accuracy of the tool used.

Conclusions

Multiple methods are used for sentiment analysis of tweets in the health care setting. These range from self-produced basic categorizations to more complex and expensive commercial software. The open source and commercial methods are developed on product reviews and generic social media messages. None of these methods have been extensively tested against a corpus of health care messages to check their accuracy. This study suggests that there is a need for an accurate and tested tool for sentiment analysis of tweets trained using a health care setting–specific corpus of manually annotated tweets first.

Introduction

Today’s doctors and patients take to online platforms such as blogs, social media, and websites to convey opinions on health matters [ 1 ]. Infodemiology is “the science of distribution and determinants of information in an electronic medium, specifically the Internet, or in a population, with the ultimate aim to inform public health and public policy” [ 2 ]. Data can be collected and analyzed from social media such as Twitter in real time with the ability to survey public opinion (sentiment) toward a subject [ 3 ]. Bates and colleagues have described social media as a “perfect storm” in regard to patient-centered health care, which is a valuable source of data for the public and health organizations [ 4 ]. Twitter is one such place, being easy to use, cheap, and accessible. Twitter is a mobile microblogging and social networking service. There are currently 955 million registered Twitter users who can share messages that contain text, video, photos, or links to external sources. One-third of people with a social media profile use Twitter, with 75% accessing from a handheld device to convey an opinion [ 5 , 6 ].

Sentiment analysis allows the content of free-text natural language—that is, the words and symbols used in a message—to be examined for the intensity of positive and negative opinions and emotions. Sentiment analysis from social media is already a widely researched subject [ 7 ]. It is useful for business marketing to understand the public or consumer opinion toward their product [ 8 ]. Computerized software tools have been produced that automate the process of sentiment analysis, allowing large numbers of free-text comments to be processed into quantitative sentiment scores quickly, for example, positive or negative [ 7 ]. They are commonly based on text classifiers or machine learning processes. These tend to be commercially orientated, expensive, and focused on gathering opinion on a specific chosen product or service [ 9 ]. During the H1N1 outbreak, Chew et al conducted a content analysis of tweets [ 10 ]. In this study, they measured sentiment in a qualitative categorical way using content classifiers such as “humor” or “sarcasm.” Accurate and automated sentiment analysis is challenging due to the subjectivity, complexity, and creativity of the language used [ 11 ].

Sentiment analysis in the health care setting is not a new phenomenon. Using only manual annotation of health care tweets, it has been found that 40% of messages contain some form of sentiment (either positive or negative) [ 12 ]. A manual method has also been used in the analysis of suicide notes and discharge summaries, where Cherry et al attempt to automate the manual process using machine learning approaches [ 13 - 15 ]. It was found that the manual classification of emotional text was difficult and inconsistent [ 13 ]. Greater positive sentiment within discharge summaries was associated with significantly decreased risk of readmission [ 14 ]. A study was also conducted measuring the sentiment of comments on the main National Health Service (NHS) website (NHS choices) over a 2-year period [ 16 , 17 ]. They found a strong agreement between the quantitative online ratings of health care providers and analysis of sentiment using their automated method.

Sentiment analysis has made its way into the mainstream analysis of Twitter-based health care research. Twitter is a popular platform as it allows data to be collected easily using their application programming interface. The limitations of other social media platforms such as Facebook are they do not allow such easy access to their data due to their varying privacy policies. It is not as easy to collect data in an open and automated way with other such media. The opinion of a tweet is found within the text portion of the tweet. This is captured in an unstructured, nonstandardized, free-text form. Accurately measuring the sentiment of a health care tweet represents an opportunity for understanding both the patient’s and health care professional’s opinion on a health subject [ 16 ]. Kent et al found that up to 40% of health care tweets contain some form of sentiment [ 12 ]. A validated tool for sentiment analysis of health care messages on Twitter would allow for the assessment of opinion on a mass scale [ 17 ]. Sentiment analysis in the medical setting offers a unique challenge as terms can have varying usage and meanings, and requires complementary context-specific features with a domain-specific lexicon [ 18 ]. The language used to convey sentiment in medicine is likely to be different than that toward a product, as the boundary between “patient,” “consumer,” and “customer” is difficult to define and terms can have varying usage and meanings [ 11 , 19 ]. Therefore, the sentiments may be expressed differently in a health care context [ 18 ].

To date, there has been no study looking at all the methods used for sentiment analysis on Twitter in the health care setting. Currently available sentiment analysis tools have not been developed based on a health care setting. SentiStrength [ 20 ], a popular open source software was based on nonspecific messages sent via MySpace [ 21 ]. Health care can be a very different environment based on many aspects. Being a public National Health Service [ 19 ], the boundary between “patient,” “consumer,” and “customer” is difficult to define in health care Therefore, currently available sentiment analysis methods may not be accurate.

The aim of this study was to review the methods used to measure sentiment for Twitter-based health care studies. The first objective was to review what methods of sentiment analysis have been used and in which health care setting. The second objective was to explore to what extent the methods were trained and validated for the study data, and if any justification for their methodology use was offered.

Identification and Screening

In May 2015, a computerized search of the literature was conducted, following Preferred Reporting Items for Systemic Reviews guidelines [ 22 ]. MEDLINE (OvidSP) and EMBASE (OvidSP) were searched using the terms. References were checked from papers and reviews, and citations were checked from included studies. The titles and abstracts were screened from the retrieved search to identify relevant studies. A supplementary hand search was carried out in September 2016 in key journals. Studies had to include one of the following search terms in the title, abstract, or keywords: “Twitter” or associated terms “tweet” or “microblog” and “Sentiment” or associated search terms “opinion” or “emoti” or “happi” or “Senti.” There were 3 inclusion criteria for the study. First, the study must have Twitter as its primary focus. The aim of this review was to explore research into the methods of sentiment analysis on Twitter messages only. Second, the papers must be relating to a health care subject. This included all aspects of health and health care delivery, health care research, policy, and organizational and professional use. Finally, papers that used a quantitative method to analyze both positive and negative sentiments of the messages, for example, “−1,” were included.

Eligibility and Inclusion

The studies were restricted to those published in English. A total of 69 full-text articles were assessed for eligibility. Of these, 15% (10/69) were rejected because they looked at social media in general (not Twitter specifically), for example, the use of social media by surgical colleagues [ 23 ]. Moreover, 36% (25/69) were rejected because the study did not pertain to health care, for example, public perceptions of nonmedical use of opioids [ 24 ]. Furthermore, 32% (22/69) papers were excluded because the sentiment analysis was either not measured, not quantitative or did not discuss positive and negative sentiments specifically, for example, characterizing sleep issues using Twitter [ 25 ]. The criteria used to compare the methods in each study looked at the method of tool production, in which setting it was used, and the method of testing the tool. For assessment, a comparison of the number of annotators used to manually annotate tweets, if any, and the level of agreement between them was used. Furthermore, the proportion of tweets used to train an algorithm compared with the final sample analyzed was also assessed.

Overall Results

In total, 12 papers were found that satisfied all 3 inclusion criteria (see Table 1 for overview). These were published between 2011 and 2016 with data collected from Twitter between 2006 and 2016. Moreover, 2 papers examined global data, 9 in the United States, and 1 in the United Kingdom. Comments from 2 papers suggest that on an average 46% (92/2) of health care tweets contain some form of sentiment, that is, not neutral [ 12 , 26 ]. Many studies conducted analysis on public health–related subjects (n=7). In addition, 3 papers examined the sentiment toward an aspect of disease: the disease itself (n=1), symptoms (n=1), or treatment (n=1). Finally, 2 papers studied an emergency medical situation and a medical conference.

Tools used for sentiment analysis.

A total of 5 of the 12 studies conducted a manual sentiment analysis of a sample of their data using annotators to train their tool. One study used 13.58% (1000/7362) of their final data sample to train their developed method [ 34 ]. Three studies used an average of 0.7% of their total dataset to train their tool (1.46%, 250/17,098; 0.55%, 2216/404,065; and 0.1%, 250/198,499). One paper compared the accuracy of their chosen methods with a manually annotated corpus of their data [ 30 ]. Moreover, 2 papers from the group commented on justification of the sentiment analysis tools used.

There were 3 categories of sentiment analysis methods found (see Table 2 ), a tool specifically produced and trained for that study data, open source tools, and commercially available software. This distinction was made based on the required level of expertise in computer programming needed to implement that method and if predefined lexicons were used. Tools produced specifically for the study required the most amount of programming knowledge as these sometimes required the use of machine learning techniques to train a tool or rule-based methods. Alternatively, using commercially available software required the least knowledge as these are designed to be quick and easy to use. Half of the studies conducted quantitative sentiment analysis using an automated method developed by the study group themselves using algorithms or machine learning techniques. Moreover, 3 studies used commercially available sentiment analysis products. The remaining 3 papers used open source, freely available sentiment analysis software, which required little programming experience. In addition, 1 study from the open source and 1 from commercial method studies used a method of manual training to tailor the tool for their specific study data [ 33 ].

Sentiment tools based on type of tool: KNN: k-nearest-neighbors; N/A: not applicable; NB: Naïve Bayes; SVM; support vector machines.

A total of 5 studies commented on the number of annotators used for the manual classification of sentiment to train their final tool (average=3 annotators, range 2-6). A single study used a method of outsourcing the task of manual classification to multiple anonymous annotators via Amazon Mechanical Turk [ 38 ].

Self-Produced Sentiment Analysis Tools

Of the 12 studies reviewed, 6 produced sentiment analysis tools within their own department, specifically designed for their study using already defined algorithms. Liu describes the different types of algorithms that can be used, and they produce different kinds of summaries [ 39 , 40 ]. Moreover, 2 different types of algorithms were found to be used, a standard supervised machine learning algorithm and a classification method (such as AFINN named after the author, Finn Arup Neilsen). These methods produce their own classifier trained to detect polarity using their original data. These may be different from the open source tools, which use already pretrained classifiers in premade software systems designed more toward an end user.

A total of 3 papers used a similar method of sentiment via categorization, all examining opinions toward smoking. Sofean et al produced an automated sentiment tool based on identifying 250 positive and 250 negative tweets from a smaller sample to train their tool [ 36 ]. There was no further detail into the annotation and analysis process. A limitation to their tool was that it screened out emoticons (symbols used to express emotion) before producing a tool. This is a method often used by users to convey emotion [ 39 ]. Myslin et al analyzed the sentiment toward emerging tobacco products on 7362 tweets, where Cole-Lewis et al looked specifically at sentiment toward electronic cigarettes on 17,098 tweets [ 29 , 34 ]. Neither of the studies commented on why a self-produced solution was used. Tweets were broadly categorized into “positive,” “neutral,” or “negative” by the annotators. The intensity of the sentiment was not recorded. To find the relationship between the sentiment and subject, 3 machine learning algorithms were used, Naïve Bayes, K-Nearest-Neighbor, and Support Vector Machine [ 41 ]. An automated sentiment analysis tool was produced based on the manual analysis of sentiment of a sample of tweets during the pilot phase of each study. This represented 13.58% (1000/7362) for Myslin. The study by Cole-Lewis used only 1.46% (250/17,098) of their total sample to train their algorithms. This represents a very small percentage of their sample and may result in their method being less accurate than intended. However, no comment is made by the study group to why only this number was used.

Desai et al used the AFINN (named after the author, Finn Arup Neilsen), to measure the sentiment of Twitter activity during Kidney Week 2011 from 993 tweets [ 31 ]. AFINN is a rule-based approach combined with statistical modeling to create a hybrid approach to sentiment classification [ 7 ]. This is based on comparing a sample of data with a list of weights of positive or negative keywords using the affective norms for English words dataset [ 42 ]. The AFINN consists of a list of manually labeled English words that have been given an integer value between −5 (highly negative) to +5 (highly positive). A value is assigned for each word in a tweet using the lexicon. The values are averaged to calculate the sentiment score for the whole message. This method has been validated for use in microblogs such as Twitter [ 43 ]. Tighe et al used this method to assess the sentiment of tweets pertaining to pain, suggesting a rule-based classifier has greater methodological advantage due to its deterministic results compared with human annotators which can have poor interannotator agreement with sentiment [ 37 ]. In addition, they supplemented AFINN with the use of emoticon terminology to enhance the accuracy of the rule-based classifier [ 39 , 44 ]. One study sought to compare different supervised machine learning (SML) techniques with each other, and to a rule-based open source lexicon for drug-related tweets [ 30 ]. They found that by using manually annotated tweets specifically from that subject to train SML techniques was more accurate than a preprepared lexicon due to the variation in language used. They also compare types of SML techniques to show that they all performed to a similar level.

Open Source Sentiment Software

Open source software is a computer software that has its source code made available to the public to modify [ 45 ]. The developers or copyright holders of the software give the rights to study and distribute the software for any purpose for free. Moreover, 4 papers used open source software for their sentiment analysis. None of these tools were initially produced using health care messages. Ramagopalan et al investigated the opinions of specific multiple sclerosis treatments using 60,037 tweets [ 26 ]. They used an open source sentiment analysis tool called package twitteR R [ 46 ] in combination with Jeffrey Breen’s sentiment analysis code [ 47 ]. This software was developed for the analysis of consumer sentiment toward a product and compares the frequency of positive or negative words against a predefined list. The overall sentiment score of each message is calculated by subtracting the number of negative words from the number of positive words. A sentiment score of >0 suggests that the message has an overall positive opinion. Of their dataset, 52% of messages contained a non-neutral sentiment. This study showed that there was a statistically significant difference in sentiment toward different types of multiple sclerosis medications. There was no comment on analysis of the tool itself or justification of its use.

Bhattacharya et al used SentiStrength [ 20 , 48 ], a popular open source software to analyze the sentiment of 164,104 tweets from 25 Federal Health Agencies in the United States and their 130 accounts. SentiStrength has been designed to measure the sentiment of short informal messages and has been widely used for Twitter analysis [ 49 ]. It was used in this case because it outperforms other lexical classifiers [ 42 ]. No manual sentiment analysis was conducted.

SentiStrength was developed in 2009 to extract sentiment strength from informal English text, giving a rating between −5 and +5. The algorithm was developed on an initial set of 2600 MySpace comments used for pilot testing. A set of 3 same gender (female) coders were used for initial testing and this was optimized by machine learning into its final version. It can detect positive emotion with 60.6% accuracy and negative emotion with 72.8% accuracy. SentiStrength outperforms a wide range of other machine learning approaches. SentiStrength has not yet been validated specifically for health care–based messages.

Hawkins et al measured patient-perceived quality of care in US hospitals using Twitter [ 33 ]. Over 404,000 tweets were analyzed for their sentiment and compared with established quality measures over a 1-year period. Natural language processing was used to measure the sentiment of the patient experience tweets. This was based on a Python library TextBlob [ 50 ]. TextBlob is trained from human annotated words commonly found in product reviews based on the Pattern Library [ 51 ]. The sentiment score can range from −1 to +1, with a score of 0 suggesting a tweet that is neutral. This was the first study that adopted Amazon Mechanical Turk [ 38 ] to use multiple outsourced anonymous curators to train their tool. They found a weak association between the positive sentiment toward a hospital and the readmission rate.

Commercial Software

There are numerous commercial software packages available to analyze the sentiment of tweets. These range in price depending on the number of tweets or duration of use. In this study, 2 papers were found using commercial software. Neither tool was developed with health care messages as its foundation, and no justification for their use is offered for either.

The largest number of messages analyzed by Nwosu measured the sentiment of over 683,000 tweets based around palliative medicine and end of life care [ 35 ]. Discussion about end of life can be difficult and sometimes missed [ 52 ]. TopsyPro was used to measure the sentiment of tweets [ 53 ]. This software was created in 2015 as an Web based tool for Twitter analytics and sentiment analysis and is based on an annual subscription costing US $12,000 per year per named user (for the “Pro” version which enables more detailed analysis). There is no information currently available on the methods used by Topsy Labs, Inc. on how the sentiment analysis is conducted.

Radian6 [ 54 ] is another piece of “listening” social media software to collect and analyze data. It has been previously used to collect data during a medical conference, with analysis focused on the major Twitter influencers [ 55 ]. The software does not require the user to have any programming knowledge and is deigned to be easy to use. Black et al used this software to analyze tweets based around public health emergency response during the Japanese earthquake and tsunami in March 2011. There was no comment on why this software was used. Radian6 can “listen” automatically to large-scale Twitter conversation based on specific keywords.

A study conducted by Greaves et al was found looking at hospital quality in the United Kingdom, and it measured the sentiment of over 198,000 tweets directed toward NHS hospitals in 2012 [ 32 ]. The commercially available software used was developed by TheySay Ltd (Oxford, UK). TheySay is based on compositional sentiment parsing, described by work from Moilanen and Pulman, using 5 automated ways of natural language processing [ 56 ]. For academic purposes, the software costs roughly £350 for a similar volume of data to the mentioned study to be analyzed.

Principal Findings

On average, 46% (92/2) of health-based tweets contain some form of positive or negative sentiment [ 12 , 26 ]. A relationship between sentiment on Twitter and hospital statistics has already been proven [ 33 ]. It is important to conduct sentiment analysis for health care tweets that is accurate and consistent. This study has found that there is a large disparity in the types of methods used, from basic categorizations to seemingly sophisticated and expensive commercially available software. Between the same subject matter such as hospital quality, different sentiment analysis methods have been used which makes it difficult to compare the results between the two [ 32 , 33 ]. Chew et al conducted a content analysis of tweets during the 2009 H1N1 outbreak and chose to use only a qualitative method for sentiment analysis of tweets, categorizing tweets based on emotive words, for example, “Humour” or “Concern” [ 10 ]. On the basis of complexity of implementation, 3 broad categories of methods have emerged: (1) self-produced methods using algorithms, (2) open source methods, and (3) commercially available software. Only 1 method in this study was produced with health care language as its foundation using a corpus of manually annotated health care setting–specific tweets for training [ 30 ]. Many methods were based on tools trained on product reviews and nonspecific social media messages that may not be appropriate for use in the health care setting [ 20 , 57 ]. The language used to convey sentiment in medicine is likely to be different than that toward a product as the boundary between “patient,” “consumer,” and “customer” is difficult to define and terms can have varying usage and meanings [ 11 , 18 , 19 ]. Health-related tweets represent a unique type of content, and their communication on Twitter carries special characteristics as found in pain-related tweets [ 37 ].

Most studies did not justify the reason for their selected method. Furthermore, there was no evidence of analysis of accuracy of the method before being used for the larger respective data. Researchers tend to assume a method selected will be accurate. Most self-produced methods train their tool using a very small percentage of their final dataset, in one case less than 2% [ 29 ]. A formal process for checking the accuracy occurred in one of the author’s study that compared types of supervised machine learning techniques. Software products and open source tools being currently used tend to be designed originally to identify opinions about products in the commercial setting rather than behaviors. This questions their accuracy when used in a medical setting.

Recommendations

This research shows that different approaches are used for the sentiment analysis of tweets in the health care setting. The evidence suggests that there is a need for the production and analysis of accuracy of a sentiment analysis tool trained using setting-specific health care tweets. Twitter is used globally, and health care can vary greatly depending on the setting. On the basis of this study, such a tool would ideally be trained using a health care subject-specific corpus of labeled tweets to train supervised machine learning classifiers [ 30 ]. Semantic Evaluation Exercises (SemEval 2016) held in San Diego is an event where programmers are tasked with producing a sentiment analysis tool on a range of Twitter subjects such as a political candidate or product, using a pre-annotated corpus. This collaborative approach could be used to produce a more advanced and accurate tool for the health care setting using subject-specific lexicons and complementary health care–based features [ 11 , 18 , 58 ]. Furthermore, it could measure the intensity of sentiment using an aggregation of methods (eg, emoticons, natural language processing, and supervised machine learning), and it could check for accuracy against a slightly larger manually annotated dataset before being used on much larger sample sizes. This could allow future research in health care–based tweets to accurately and consistently measure the sentiment of setting specific health care–based messages.

Abbreviations

Conflicts of Interest: None declared.

IEEE Account

• Change Username/Password
• Update Address

Purchase Details

• Payment Options
• Order History
• View Purchased Documents

Profile Information

• Communications Preferences
• Profession and Education
• Technical Interests
• US & Canada: +1 800 678 4333
• Worldwide: +1 732 981 0060
• Contact & Support
• About IEEE Xplore
• Accessibility
• Terms of Use
• Nondiscrimination Policy
• Privacy & Opting Out of Cookies

A not-for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology for the benefit of humanity. © Copyright 2024 IEEE - All rights reserved. Use of this web site signifies your agreement to the terms and conditions.

• Data Center
• Applications
• Open Source

Datamation content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More .

Sentiment analysis, also known as opinion mining, is the process of using computational techniques to extract subjective information from textual data like emails, customer reviews, or social media feeds. Sentiment analysis tools determine the emotional tone or overall sentiment expressed toward a topic, product, service, brand, or individual. Businesses and organizations can use these tools to monitor online conversations, analyze customer feedback, understand public opinion, manage brand reputation, and predict trends.

While sentiment analysis initially focused on classifying text as positive, negative, or neutral, the field has become more sophisticated, using such methods as natural language processing (NLP), machine learning (ML), and deep learning to identify emotions and emotional undertones behind text.

Table of Contents

Featured Partners: Data Analysis Software

Understanding How Sentiment Analysis Works

Sentiment analysis involves a step-by-step process that transforms raw text—from social media content, blog posts, reviews, customer support tickets, case studies, webchats, or community forums, for example—into a quantifiable sentiment score. The process includes preprocessing the data, engineering features to help the model identify the sentiments it communicates, integrating a lexicon to quantify those sentiments, using machine learning to process the data, and refining the results.

Step 1: Preprocessing Data

Data preparation is a foundational step to ensure the quality of the sentiment analysis by cleaning and preparing text before feeding it to a machine learning model. Common data preprocessing techniques include lowercasing, which converts text to lowercase for consistency; tokenization, which breaks down text into words or phrases (tokens); stop word removal, which eliminates unnecessary words; and normalization, which corrects spelling errors and handles slang or abbreviations.

Step 2: Engineering Features

The next step is to establish features to help the model identify sentiments. This process involves the creation, transformation, extraction, and selection of the features or variables most suitable for creating an accurate machine learning algorithm. Part-of-speech (POS) tagging is used to identify the grammatical functions of each word, as some parts of speech carry more sentiment weight, and n-grams, which analyze sequences of words that hold stronger sentiment value than individual words.

Step 3: Integrating a Sentiment Lexicon

Lexicon-based sentiment analysis is a popular technique for extracting the emotional polarity of text. It relies on lexicons, or predefined dictionaries of words and phrases that have an associated sentiment value. These lexicons are integrated into the analysis to allow models to assign sentiment scores based on the presence of sentiment-laden words within the text.  Examples of sentiment lexicons that can be integrated include NRCLexicon and SentiWordNet .

Step 4: Performing Machine Learning Analysis

In this step, machine learning algorithms are used for the actual analysis. These can include supervised or unsupervised learning methods. Supervised learning trains the model on labeled data where the text is paired with its corresponding sentiment, while unsupervised learning allows the model to identify sentiment clusters within the data without pre-labeled examples, which enables exploring emerging sentiment trends.

Step 5: Evaluating and Refining

The final step involves evaluating the model’s performance on unseen data by setting metrics to help assess how well the model identifies the sentiment. Users can refine the model through other methods, such as parameter tuning or exploring a different algorithm based on these evaluations.

5 Types of Sentiment Analysis

Sentiment analysis goes beyond classifying text as positive or negative. It can be categorized in different ways based on the level of granularity and the methods used. Popular methods include polarity based, intent based, aspect-based, fine-grained, and emotion detection.

Polarity-Based Sentiment Analysis

Polarity-based sentiment analysis determines the overall sentiment behind a text and classifies it as positive, negative, or neutral. Polarity can be expressed with a numerical rating, known as a sentiment score, between -100 and 100, with 0 representing neutral sentiment. This method can be applied for a quick assessment of overall brand sentiment across large datasets, such as social media analysis across multiple platforms.

Intent-Based Analysis

Intent-based analysis can identify the intended action behind a text—for instance, whether a customer wants to seek information, purchase a product, or file a complaint. This type of sentiment analysis can be applied to developing chatbots for efficient conversation routing or helping marketers identify the right B2B campaign for their target audience.

Fine-Grained Analysis

Fine-grained analysis delves deeper than classifying text as positive, negative, or neutral, breaking down sentiment indicators into more precise categories. Fine-grained analysis provides a more nuanced understanding of opinions, as it identifies why customers or respondents feel the way they do.

Emotion Detection Analysis

Emotion detection analysis defines and evaluates specific emotions within a text, such as anger, joy, sadness, or fear. This type of sentiment analysis is ideal for businesses or brands that aim to deliver empathic customer service, as it can help them understand the emotional triggers in advertising or marketing campaigns.

Aspect-Based Analysis

Aspect-based analysis identifies the sentiment toward a specific aspect of a product, service, or topic. This technique categorizes data by aspect and determines the sentiment attributed to each. It is usually applied for analyzing customer feedback, targeting product improvement, and identifying the strengths and weaknesses of a product or service.

Sentiment Analysis Methodologies

Sentiment analysis uses computational techniques to determine the emotions and attitudes within textual data. Natural language processing (NLP) and machine learning (ML) are two of the major approaches that are used.

Natural Language Processing (NLP)

NLP is a branch of artificial intelligence (AI) that combines computational linguistics with statistical and machine learning models, enabling computers to understand human language. In sentiment analysis, NLP techniques play a role in such methods as tokenization, POS tagging, lemmatization or stemming, and sentiment dictionaries.

Machine Learning (ML)

ML is a branch of AI and computer science that uses algorithms that learn from massive amounts of data to identify patterns and make predictions. It enables AI to imitate how humans learn and has revolutionized the field of sentiment analysis in many ways. With ML, algorithms can be trained on labeled data (supervised learning) or it can identify patterns in unlabeled data (unsupervised learning). It also allows advanced neural networks to extract complex data from text through deep learning.

Data Collection and Preparation for Sentiment Analysis

The reliability of results depends on the quality and relevance of the data being analyzed—as such, careful consideration must be given to choosing the sources and strategies of data collection. It’s also important to address challenges in the data collection process accordingly and follow the best practices in processing data for sentiment analysis.

Gathering Data for Sentiment Analysis

Data for sentiment analysis can be mined from a variety of sources for both online and offline platforms. Choosing the right data source for sentiment analysis depends on the specific goals, needs, and research questions of a business or organization. Here are the most common sources of data:

• Social Media: Social media platforms provide a constant source of public opinions on topics, brands, and current events that can be collected through manual searching, social listening tools, and API data retrieval.
• Review Sites: Textual data can be gathered from review sites like Yelp, Amazon, and Google Reviews through web scraping and APIs to provide insights into how customers feel about a company, product, or service.
• Surveys and Focus Groups: These examples of structured data gathering are aimed at collecting feedback on specific topics.
• Articles and Publications: Online news, websites, blogs, and industry publications are powerful tools for understanding public sentiment and social issues. They can be searched using news aggregators or web scraping methods.
• Voice of Customer (VoC) Data: This non-traditional data source includes chat histories, customer support transcripts, customer emails, and more, and can be incorporated using CRM integrations or transcript analysis tools.
• Public Information: Public information available on government or company websites, including press releases and financial reports, can be a rich source of useful data.
• Employee Feedback: Companies can collect employee feedback through internal surveys, performance reviews, and communication channels using human resources (HR) systems and feedback platforms.
• Electronic Medical Records: Patient notes and records within the healthcare system can be accessed through HIPAA-compliant (Health Insurance Portability and Accountability Act of 1996) platforms or CRMs. Medical records require secure access, so it’s important to collaborate with healthcare providers and adhere to privacy regulations.
• Podcasts: Podcasts are a great source of information, such as commentaries and discussions found within podcast transcripts.
• Gaming Platforms: Data from in-game chats, forum posts, and reviews is also a rich source of information for sentiment analysis.

How to Gather Data for Sentiment Analysis

Before collecting data, define your goals for what you want to learn through sentiment analysis. If you’re conducting a study, determine your research questions—be as specific as possible—and identify opinions or emotions you’re interested in, such as customer satisfaction, brand perception, or attitude towards a social issue.

Select the type of data suitable for your project or research and determine your data collection strategy. When gathering data online, make sure to comply with the websites’ terms and conditions, and if you’re interacting directly with respondents or customers, ensure that their privacy is protected and secure informed consent forms—especially if you’re using surveys and focus groups.

Best Practices for Processing Data for Sentiment Analysis

Processing raw data before conducting sentiment analysis ensures that the data is clean and ready for algorithms to interpret. While there are several methodical measures that you can take in processing data for sentiment analysis, it still depends on your goals and the characteristics of the dataset you have.

Data Cleaning

This process involves removing redundant, incorrect, and irrelevant data not meant for analysis, including HTML tags, hyperlinks, special characters, and other text that shouldn’t be in your dataset. Methods of data cleaning and preprocessing include the following:

• Normalization: Normalizes the data for analysis by reducing noise and complexity and transforming it into a consistent format using processes like stemming (reducing words to their root form) or lemmatization (grouping words to form a common base).
• Tokenization: Breaks down textual data into tokens to make understanding the individual components of the text easier.
• POS Tagging: Used for more nuanced sentiment analysis, especially for detecting sarcasm or negation—NLP techniques can identify the grammatical function of each word and other categories such as tense, number, and more.
• Negation Handling: Techniques can automatically detect the extent of negation in a text, such as rule-based methods or trained models.
• Stop Word Removal: This optional method can remove stop words such as “the,” “a,” or “an” to improve efficiency in analyzing the text.

Select Your Model

Choose a sentiment analysis model that’s aligned with your objectives, size, and quality of training data, your desired level of accuracy, and the resources available to you. The most common models include the rule-based model and a machine learning model.

A rule-based model involves data labeling, which can be done manually or by using a data annotation tool. A machine learning model can be built by training a vast amount of data to analyze text to give more accurate and automated results.

Analyze and Evaluate

In processing data for sentiment analysis, keep in mind that both rule-based and machine learning models can be improved over time. It’s important to assess the results of the analysis and compare data using both models to calibrate them.

3 Top Sentiment Analysis Tools and Technologies

Sentiment analysis tools use AI and deep learning techniques to decode the overall sentiment of a text from various data sources. The best tools can use various statistical and knowledge techniques to analyze sentiments behind the text with accuracy and granularity. Three of the top sentiment analysis solutions on the market include IBM Watson, Azure AI Language, and Talkwalker.

IBM Watson Natural Language Understanding

IBM Watson Natural Language Understanding (NLU) is an AI-powered solution for advanced text analytics. This platform uses deep learning to extract meaning and insights from unstructured data, supporting up to 12 languages. Users can extract metadata from texts, train models using the IBM Watson Knowledge Studio, and generate reports and recommendations in real-time.

IBM Watson NLU stands out as a sentiment analysis tool for its flexibility and customization, especially for users who are working with a massive amount of unstructured data. It’s priced based on the NLU item, equivalent to one text unit or up to 10,000 characters. The standard tier starts at $0.0002 per NLU item per month.

Azure AI Language

Microsoft’s Azure AI Language, formerly known as Azure Cognitive Service for Language, is a cloud-based text analytics platform with robust NLP features. This platform offers a wide range of functions, such as a built-in sentiment analysis tool, key phrase extraction, topic moderation, and more.

What sets Azure AI Language apart from other tools on the market is its capacity to support multilingual text, supporting more than 100 languages and dialects. It also offers pre-built models that are designed for multilingual tasks, so users can implement them right away and access accurate results. Azure AI Language offers free 5,000 text records per month and costs $25 per 1,000 succeeding text records.

Talkwalker is a sentiment analysis tool designed for social media monitoring. As a leading social listening platform, it offers robust tools for analyzing brand sentiment, predicting trends, and interacting with target audiences online.

Users can leverage AI-powered sentiment analysis tools to detect negative comments or sarcasm on social media posts, forums, and images to provide companies and organizations with an in-depth understanding of their online brand perception. Talkwalker offers four pricing tiers, and potential customers can contact sales to request quotes.

Learn more about our picks in our review of the best sentiment analysis tools for 2024 .

5 Practical Application Examples of Sentiment Analysis

Sentiment analysis has become a valuable tool for organizations in a wide range of industries. Companies can use it for social media monitoring, customer service management, and analysis of customer data to improve operations and drive growth.

Customer Experience (CX)

Sentiment analysis is a valuable tool for improving customer satisfaction through brand monitoring, product evaluation, and customer support enhancement. Sentiment analysis tools can help businesses and organizations monitor what customers say about their brands, products, or services, and sentiment analysis data can help brands resolve issues, retain customers, and identify the best communication platforms to engage with leads and customers.

Marketing and Sales

Sentiment analysis tools enable sales teams and marketers to identify a problem or opportunity and adapt strategies to meet the needs of their customer base. They can help companies follow conversations about their business and competitors on social media platforms through social listening tools. Organizations can use these tools to understand audience sentiment toward a specific topic or product and tailor marketing campaigns based on this data.

Social Issues and Politics

Sentiment analysis tools are valuable in understanding today’s social and political landscape. For instance, users can understand public opinion by tracking sentiments on social issues, political candidates, or policies and initiatives. It can also help in identifying crises in public relations and provide insights that are crucial for the decision-making process of policymakers.

Sentiment analysis is essential for financial markets, as it helps professionals analyze news articles, social media channels, and financial reports to predict stock trends. Sentiment analysis tools can help calibrate investor sentiments toward companies that could affect stock prices.

Healthcare practitioners can leverage patient sentiment data to understand their needs and support them, which is a helpful tool in advancing mental health research. Sentiment analysis also enables service providers to analyze patient feedback to improve their satisfaction and overall experience.

Advanced Topics in Sentiment Analysis

Sentiment analysis is a complex field and has played a pivotal role in the realm of data analytics . Ongoing advancements in sentiment analysis are designed for understanding and interpreting nuanced languages that are usually found in multiple languages, sarcasm, ironies, and modern communication found in multimedia data.

Multilingual Sentiment Analysis

Analyzing sentiments across multiple languages and dialects increases the complexity of data analysis . Different languages and dialects have unique vocabularies, cultural contexts, and grammatical structures that could affect how a sentiment is expressed. To understand the sentiments behind multiple languages, you can make use of AI-driven solutions or platforms that include language-specific resources and sentiment-aware models.

Sarcasm, Irony, and Other Complexities

Despite the advancements in text analytics, algorithms still struggle to detect sarcasm and irony. Rule-based models, machine learning, and deep learning techniques can incorporate strategies for detecting sentiment inconsistencies and using real-world context for a more accurate interpretation.

Sentiment Analysis and Multimedia Data

Multimodal sentiment analysis extracts information from multiple media sources, including images, videos, and audio. Analyzing multimodal data requires advanced techniques such as facial expression recognition, emotional tone detection, and understanding the impact between modalities.

Challenges and Limitations of Sentiment Analysis

Sentiment analysis is a powerful tool for businesses that want to understand their customer base, enhance sales marketing efforts, optimize social media strategies, and improve overall performance. However, sentiment analysis also has challenges and limitations.

It requires accuracy and reliability, but even the most advanced algorithms can still misinterpret sentiments. Accuracy in understanding sentiments is influenced by several factors, including subjective language, informal writing, cultural references, and industry-specific jargon. Continuous evaluation and fine-tuning of models are necessary to achieve reliable results.

Sentiment analysis should also adhere to ethical considerations, as the process involves personal opinions and private data. In conducting sentiment analysis, prioritize the respondents’ privacy and observe responsible data collection processes. Identify and address potential biases in datasets by using diverse and representative data that covers different demographics, cultures, and viewpoints, or by employing re-sampling and specialized algorithms.

Bottom Line: The Continuing Relevance of Sentiment Analysis

Sentiment analysis can help organizations understand the emotions, attitudes, and opinions behind an ever-increasing amount of textual data. While certain challenges and limitations exist in this field, sentiment analysis is widely used for enhancing customer experience, understanding public opinion, predicting stock trends, and improving patient care.

In the future, sentiment analysis systems might employ more advanced techniques for recognizing nuanced languages and capturing sentiments more accurately. Ultimately, sentiment analysis will remain an essential tool for businesses and researchers alike to better understand their audience and stay on top of the latest trends.

Learn more about other things you can discover through different types of analysis in our articles on key benefits of big data analytics and statistical analysis .

Subscribe to Data Insider

Learn the latest news and best practices about data science, big data analytics, artificial intelligence, data security, and more.

Similar articles

11 top data collection trends emerging in 2024, 6 top data classification trends, 7 data management trends: the future of data management, get the free newsletter.

Subscribe to Data Insider for top news, trends & analysis

Latest Articles

Exploring multi-tenant architecture: a..., 8 best data analytics..., common data visualization examples:..., what is data management....

Online sentiment analysis monitoring is an essential strategy for brands aiming to understand their audience’s perceptions towards their brand. By analyzing online conversations, brands gain valuable insights and identify trends. This helps them make data-driven decisions to improve marketing, customer service, and product development. This article will present the top 10 online sentiment monitoring platforms for brands, highlighting their key features, benefits, and applications.

Table of Contents

• Talk Walker
• Hootsuite Insights
• Linkfluence

1. TalkWalker

• Comprehensive social listening and analytics platform: TalkWalker is a powerful tool for social listening and analytics, providing comprehensive data from across social media, news, blogs, and forums. It monitors over 150 million sources, ensuring brands stay in the loop with relevant conversations. This extensive data collection enables businesses to keep their finger on the pulse of public opinion and trends, aiding in proactive decision-making.
• Advanced sentiment analysis capabilities: TalkWalker leverages artificial intelligence (AI) to perform nuanced sentiment analysis. This means it captures not just positive or negative sentiments but also complex emotions like anger, joy, or sadness. For example, a brand can quickly identify if a viral tweet about their product is causing joy or outrage. Therefore, this allows them to respond appropriately and swiftly.
• Integration with popular marketing tools and platforms: The platform integrates seamlessly with a variety of marketing tools such as Hootsuite, HubSpot, and Marketo. This integration facilitates a smooth flow of data and enhances analytics, enabling marketers to take timely, informed actions. For example, a sports apparel brand might use TalkWalker to monitor social media discussions about their new product launch. If TalkWalker detects a surge in negative sentiment due to sizing issues, the brand can quickly address the problem, preventing a potential PR crisis.

• Real-time social media monitoring and analytics: Brand24 provides real-time monitoring of social media and online conversations. This helps brands stay ahead of discussions and track online sentiment effectively. By tracking mentions, keywords, and hashtags, businesses can identify key trends, sentiments, and influencers. This ensures they remain informed about their industry’s pulse. For example, a sports apparel brand can monitor the buzz around their latest release and quickly adjust their strategy based on immediate feedback.
• In-depth sentiment analysis and reporting: The platform includes powerful sentiment analysis and detailed reporting tools. These tools help brands understand public perception. They analyze whether conversations about a brand are positive, negative, or neutral. This provides insights into the effectiveness of marketing campaigns. For example, a major airline uses Brand24 to identify areas needing improvement in customer service policies. This enhances overall customer satisfaction.
• Customizable alerts and notifications: Users can set up personalized alerts for specific keywords, mentions, or sentiments, keeping them constantly informed about ongoing conversations. This feature is especially useful for catching emerging trends or managing crises promptly. Brand24 ensures that vital information never slips through the cracks, supporting a proactive approach in brand management.

3. Hootsuite Insights

• Social media management and analytics platform: Hootsuite Insights, powered by Brandwatch, merges social media management with comprehensive analytics for online sentiment.
• Advanced sentiment analysis and topic discovery: It offers advanced sentiment analysis and helps users discover trending topics relevant to their brand. This tool allows businesses to gauge customer opinions and emotions related to their products or services. It makes tailoring marketing strategies easier.
• Integration with Hootsuite’s social media scheduling and engagement tools: This integration ensures a cohesive workflow for managing social media scheduling, engagement, and analysis. Users can seamlessly schedule posts, interact with their audience, and monitor performance metrics from a single platform. This enhances efficiency and productivity.

4. Synthesio

• Enterprise-grade social listening and analytics platform: Synthesio is designed for large organizations needing thorough social media monitoring, online sentiment data, and comprehensive analysis. It excels in accessing vast data sources, making it ideal for enterprises. This tool helps businesses stay updated on social conversations.
• Advanced sentiment analysis and natural language processing (NLP) capabilities: Synthesio employs NLP techniques to provide detailed sentiment analysis across a multitude of languages. These advanced capabilities enable organizations to decipher customer opinions, preferences, and concerns from various social media platforms. For example, a company used Synthesio to scrutinize 1.5 million social media mentions. This helped them gain deep insights into consumer behaviors and needs in real-time.
• Customizable dashboards and reporting: Users can design custom dashboards and generate reports to track specific metrics and key performance indicators (KPIs) aligned with their brand goals. This flexibility offers immense value by allowing brands to focus on the data most relevant to their objectives.

5. Brandwatch

According to their website, sentiment accuracy generally falls within the range of 60-75% for supported languages; however, this can fluctuate based on the data source used.

• Powerful social listening and analytics platform: Brandwatch provides deep insights through its extensive data access across various online sources, offering a unique perspective on consumer behavior and trends.
• Advanced sentiment analysis and image recognition: Its sentiment analysis includes image recognition, helping brands understand the impact of visual content on consumer sentiment. By analyzing images shared on social media, brands can gauge how their visual content influences customer opinions. This gives them an edge in marketing strategies.
• Integration with popular marketing and data visualization tools: Seamless integration with tools like Hootsuite and Tableau allows for enhanced data visualization and broader analysis capabilities. This integration makes it easier for brands to visualize data patterns and refine their strategies based on comprehensive analytics.
• Review management platform : Birdeye offers a comprehensive suite of tools for review management, customer feedback, and overall customer experience. Birdeye uses non-gated review management.
• Multi-channel review management : Birdeye aggregates reviews from over 200 sites, allowing businesses to monitor and respond to customer feedback across various platforms efficiently. This wide-ranging coverage helps companies maintain a consistent brand image and address customer concerns promptly. 
• AI-powered insights : The platform utilizes artificial intelligence to analyze customer feedback, providing businesses with actionable insights to improve their products, services, and overall customer experience. This feature helps companies identify trends and make data-driven decisions to enhance their reputation. 
• Automated review generation : Birdeye offers tools to automatically request reviews from customers after interactions, helping businesses build a robust online presence with authentic customer feedback. This proactive approach can significantly boost a company’s online visibility and credibility.
• Integrated customer surveys : The platform includes customizable survey tools that allow businesses to gather detailed feedback from customers. This feature enables companies to gain deeper insights into customer satisfaction and identify areas for improvement.

7. Linkfluence

• Social media intelligence and analytics platform: Linkfluence offers a comprehensive approach to social media intelligence and analytics, capturing real-time data from a wide array of sources. This helps businesses stay connected to the latest trends, customer sentiments, and emerging conversations relevant to their brand.
• Advanced sentiment analysis and NLP capabilities: Linkfluence excels at delivering precise sentiment analysis with cutting-edge natural language processing (NLP) capabilities. This accuracy holds even in complex and multifaceted conversations. For instance, a major sports apparel brand leveraged Linkfluence’s NLP technology to uncover nuanced opinions about their products, leading them to refine their marketing strategies effectively.
• Customizable dashboards and reporting: The platform’s customizable dashboards and reporting features allow users to craft personalized views and generate insightful reports. This customization facilitates targeted analysis and helps businesses make informed decisions. By tailoring these dashboards, companies can focus on the specific metrics and insights most relevant to their objectives.

8. Digimind

• Social listening and analytics platform: Digimind stands out by offering comprehensive tools to businesses. These tools help brands grasp and enhance their online presence. The platform’s strength lies in its ability to monitor vast amounts of data from social media, blogs, forums, and news sites. This makes it indispensable for brand reputation management.
• Advanced sentiment analysis and topic discovery: At the core of Digimind’s capabilities is its sentiment analysis, which pinpoints the emotions in customer conversations in real time. This feature allows brands to gauge customer satisfaction by identifying positive, negative, and neutral sentiments. Additionally, the platform excels at discovering trending topics and uncovering relevant discussions. This keeps brands informed about what’s happening in their industry, what competitors are doing, and what their customers care about.
• Integration with popular marketing and data visualization tools: Digimind’s seamless integration with tools like Hootsuite, Marketo, and Tableau extends its functionality, providing deeper insights. This means brands can merge social listening insights with their existing marketing and analytics workflows effortlessly.
• Visual social listening and analytics platform: YouScan stands out by analyzing visual content in addition to traditional text-based sentiment analysis. This approach allows brands to gain a more comprehensive understanding of consumer sentiments expressed through images and videos.
• Advanced image recognition and sentiment analysis: With its advanced image recognition capabilities, YouScan goes beyond text to identify sentiments embedded in visual content. This enables businesses to decipher customer emotions expressed in photos, enhancing their understanding of market trends and public perception.
• Customizable dashboards and reporting: The platform provides customizable dashboards and reporting tools, allowing users to monitor specific metrics and gain actionable insights tailored to their needs. This feature ensures that companies can track relevant data and make informed decisions in real-time.

10. Mention

• Real-time social media monitoring and analytics: Mention provides real-time monitoring and analysis of social media mentions and conversations. This empowers businesses to stay aware of their online reputation and presence by tracking mentions of their brand, competitors, and industry-specific words.
• Advanced sentiment analysis and reporting: The platform boasts advanced sentiment analysis using natural language processing to gauge the emotional tone behind social media posts. This allows businesses to understand public sentiment towards their brand, products, or services more comprehensively. In addition, Mention offers detailed reporting tools that provide valuable insights and visual data, aiding in smarter decision-making processes.
• Customizable alerts and notifications: Users can set up tailored alerts for specific mentions, keywords, or sentiment changes. This feature ensures they stay informed in real-time, enabling prompt responses to customer complaints, identifying potential sales leads, and keeping tabs on industry trends. This proactive approach can significantly enhance a business’s responsiveness and strategic planning.

What is sentiment analysis?

Sentiment analysis is a technique used to determine the emotional tone behind online text. By leveraging natural language processing (NLP), machine learning, and text analysis, these tools interpret whether the expressed sentiment is positive, negative, or neutral.

How does sentiment analysis work?

Sentiment analysis operates by examining text data from sources like social media, reviews, and comments. NLP algorithms dissect sentences to identify the sentiment behind the words, determining the overall emotion. This involves parsing the text, extracting meaning, and classifying it into sentiment categories.

What is the definition of sentiment analysis?

Sentiment analysis is the computational process of identifying and categorizing opinions in text. It helps in understanding the author’s attitude toward a specific topic, product, or service by analyzing their written expressions.

How can I perform sentiment analysis on my text data?

You can conduct sentiment analysis using various online platforms and tools that specialize in this method. These tools utilize NLP and machine learning to analyze your text data, offering insights into public perception and sentiment trends. Popular platforms include SEMrush, Brandwatch, and Alchemer, which provide detailed sentiment insights driven by robust analytical techniques.

Tags: Reputation Protection .

Ready to Take the Next Step?

Get in touch with our team and we’ll take the first steps toward making you look better online.

Mapreduce framework based sentiment analysis of twitter data using hierarchical attention network with chronological leader algorithm

• Original Article
• Published: 27 August 2024
• Volume 14 , article number  172 , ( 2024 )

Cite this article

• Jayashree Jagdale 1 ,
• R. Sreemathy 2 ,
• Balaso Jagdale 3 &
• Kranti Ghag 4  

16 Accesses

Explore all metrics

Sentiment analysis (SA) or opinion mining is a general dialogue preparation chore that intends to discover sentiments behind the opinions in texts on changeable subjects. Recently, researchers in an area of SA have been considered for assessing opinions on diverse themes like commercial products, everyday social problems and so on. Twitter is a region, wherein tweets express opinions, and acquire an overall knowledge of unstructured data. This process is more time-consuming and the accuracy needs to be improved. Here, the Chronological Leader Algorithm Hierarchical Attention Network (CLA_HAN) is presented for SA of Twitter data. Firstly, the input Twitter data concerned is subjected to a data partitioning phase. The data partitioning of input Tweets are conducted by Deep Embedded Clustering (DEC). Thereafter, partitioned data is subjected to MapReduce framework, which comprises of mapper and reducer phase. In the mapper phase, Bidirectional Encoder Representations from Transformers (BERT) tokenization and feature extraction are accomplished. In the reducer phase, feature fusion is carried out by Deep Neural Network (DNN) whereas SA of Twitter data is executed utilizing a Hierarchical Attention Network (HAN). Moreover, HAN is tuned by CLA which is the integration of chronological concept with the Mutated Leader Algorithm (MLA). Furthermore, CLA_HAN acquired maximal values of f-measure, precision and recall about 90.6%, 90.7% and 90.3%.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Subscribe and save.

• Get 10 units per month
• Download Article/Chapter or eBook
• 1 Unit = 1 Article or 1 Chapter
• Cancel anytime

Price excludes VAT (USA) Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

DAC-BiNet: Twitter crime detection using deep attention convolutional bi-directional aquila optimal network

An Attention Arousal Space for Mapping Twitter Data

Sentiment analysis using semantic similarity and hadoop mapreduce, explore related subjects.

• Artificial Intelligence

Data availability

The data underlying this article are available in Twitter US Airline sentiment dataset, at https://www.kaggle.com/crowdflower/twitter-airline-sentiment .

Al-mashhadani MI, Hussein KM, Khudir ET (2022) Sentiment analysis using optimized feature sets in different facebook/twitter dataset domains using big data. Iraqi J Comput Sci Math 3(1):64–70

Article   Google Scholar  

Andrews SJ, Lamb ME (2017) The structural linguistic complexity of lawyers’ questions and children’s responses in Scottish criminal courts. Child Abuse Negl 65:182–193

Bag S, Kumar SK, Tiwari MK (2019) An efficient recommendation generation using relevant Jaccard similarity. Inf Sci 483:53–64

Barnaghi P, Ghaffari P, Breslin JG (2016) Opinion mining and sentiment polarity on twitter and correlation between events and sentiment. In: Proceedings of 2016 IEEE second international conference on big data computing service and applications, BigDataService. IEEE, pp 52–57

Basiri ME, Nemati S, Abdar M, Cambria E, Acharya UR (2021) ABCDM: an attention-based bidirectional CNN-RNN deep model for sentiment analysis. Futur Gener Comput Syst 115:279–294

Bukhari SAC, Bashir AK, Malik KM (2018) Semantic web in the age of big data: a perspective: Will We Drown in a Data Tsunami or Enter a Knowledge Utopia?. IEEE Microwave and Wireless Technology Letters 3(3):1–6

Choudhury J, Pandey C, Saxena A (2019) Sentimental analysis of Twitter data on Hadoop. In: Computing, communication and signal processing: proceedings of ICCASP 2018. Springer Singapore, pp 457–469

Christian H, Agus MP, Suhartono D (2016) “Single document automatic text summarization using term frequency-inverse document frequency (TF-IDF). ComTech Comput Math Eng Appl 7(4):285–294

Google Scholar  

Dehghani M, Trojovsky P (2022) Hybrid leader based optimization: a new stochastic optimization algorithm for solving optimization applications. Sci Rep 12(1):5549

Demirbaga U (2021) HTwitt: a hadoop-based platform for analysis and visualization of streaming Twitter data. Neural Comput Appl 35:1–16

Devlin J, Chang MW, Lee K, Toutanova K, (2018) Bert: pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805

Guo X, Zhu E, Liu X, Yin J (2018) Deep embedded clustering with data augmentation. In: Proceedings of Asian conference on machine learning, PMLR, pp 550–565

Hussain A, Cambria E (2018) Semi-supervised learning for big social data analysis. Neurocomputing 275:1662–1673

Kamps J, Marx M, Mokken RJ, De Rijke M (2004) Using WordNet to measure semantic orientations of adjectives. In Lrec 4:1115–1118

Malik M, Naaz S, Ansari IR (2019) Sentiment analysis of Twitter data using big data tools and Hadoop ecosystem. In: Proceedings of the international conference on ISMAC in computational vision and bio-engineering 2018 (ISMAC-CVB). Springer International Publishing, pp 857–863

Meher S (2021) Fractional order fish migration optimization algorithm for spam detection in twitter data stream. J Netw Commun Syst 4(4):33–42

Mohammad SM, Kiritchenko S, Zhu X (2013) NRC-Canada: building the state-of-the-art in sentiment analysis of tweets. arXiv preprint arXiv:1308.6242

Poli R, Kennedy J, Blackwell T (2007) Particle swarm optimization: an overview. Swarm Intell 1:33–57

Qian Y, Fan Y, Hu W, Soong FK (2014) On the training aspects of deep neural network (DNN) for parametric TTS synthesis. In: Proceedings of 2014 IEEE international conference on acoustics, speech and signal processing (ICASSP). IEEE, pp 3829–3833

Ramsingh J, Bhuvaneswari V (2021) An efficient map reduce-based hybrid NBC-TFIDF algorithm to mine the public sentiment on diabetes mellitus–a big data approach. J King Saud Univ Comput Inf Sci 33(8):1018–1029

Rao R (2016) Jaya: a simple and new optimization algorithm for solving constrained and unconstrained optimization problems. Int J Ind Eng Comput 7(1):19–34

Rodrigues AP, Chiplunkar NN (2019) A new big data approach for topic classification and sentiment analysis of Twitter data. Evol Intell 15:1–11

Saric F, Glavas G, Karan M, Snajder J, Basic BD (2012) Takelab: systems for measuring semantic text similarity. In: SEM 2012: the first joint conference on lexical and computational semantics–volume 1: proceedings of the main conference and the shared task, and volume 2: proceedings of the sixth international workshop on semantic evaluation, SemEval 2012, pp 441–448

Sharma S, Kumar S (2019) Sentiment analysis on twitter posts using Hadoop. Int. Res. J. Eng. Technol. (IRJET) 06(04):4201–4204

Shetty SD (2021) Sentiment analysis, tweet analysis and visualization on big data using Apache Spark and Hadoop. In: IOP conference series: materials science and engineering. IOP Publishing, vol 1099, no 1, pp 012002

Tamrakar L, Shrivastava D, Ghosh DS (2021) Student sentiment analysis using classification with feature extraction techniques. arXiv preprint arXiv:2102.05439

Turney PD (2002) Thumbs up or thumbs down? Semantic orientation applied to unsupervised classification of reviews. arXiv preprint cs/0212032

Twitter US Airline sentiment dataset is taken from. https://www.kaggle.com/crowdflower/twitter-airline-sentiment . Accessed on July 2023

Yang Z, Yang D, Dyer C, He X, Smola A, Hovy E (2016) Hierarchical attention networks for document classification. In: Proceedings of the 2016 conference of the North American chapter of the association for computational linguistics: human language technologies, pp 1480–1489

Zeidabadi FA, Doumari SA, Dehghani M, Montazeri Z, Trojovsky P, Dhiman G (2022) MLA: a new mutated leader algorithm for solving optimization problems. Comput Mater Contin 70(3):5631–5649

Download references

Acknowledgements

I would like to express my very great appreciation to the co-authors of this manuscript for their valuable and constructive suggestions during the planning and development of this research work.

This research did not receive any specific funding.

Author information

Authors and affiliations.

Department of Information Technology, Pune Institute of Computer Technology, Pune, India

Jayashree Jagdale

Department of Electronics and Telecommunication Engineering, Pune Institute of Computer Technology, Pune, Maharashtra, India

R. Sreemathy

Department of Computer Engineering, Dr. Vishwanth Karad MIT World Peace University, Pune, India

Balaso Jagdale

Department of Computer Engineering, SVKM’s DJ Sanghvi College of Engineering, Mumbai, India

Kranti Ghag

You can also search for this author in PubMed   Google Scholar

Contributions

All authors have made substantial contributions to conception and design, revising the manuscript, and the final approval of the version to be published. Also, all authors agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Corresponding author

Correspondence to Jayashree Jagdale .

Ethics declarations

Conflict of interest.

The authors declare no conflict of interest

Ethical approval

Not Applicable.

Informed consent

Additional information, publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Jagdale, J., Sreemathy, R., Jagdale, B. et al. Mapreduce framework based sentiment analysis of twitter data using hierarchical attention network with chronological leader algorithm. Soc. Netw. Anal. Min. 14 , 172 (2024). https://doi.org/10.1007/s13278-024-01293-y

Download citation

Received : 18 September 2023

Revised : 11 March 2024

Accepted : 25 June 2024

Published : 27 August 2024

DOI : https://doi.org/10.1007/s13278-024-01293-y

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Deep neural network (DNN)
• Mutated leader algorithm (MLA)
• Deep embedded clustering (DEC)
• MapReduce framework
• Hierarchical attention network (HAN)
• Find a journal
• Publish with us
• Track your research