Skip to main content
SpringerLink
Log in

A unified latent variable model for contrastive opinion mining

  • Research Article
  • Published:
Frontiers of Computer Science Aims and scope Submit manuscript

Abstract

There are large and growing textual corpora in which people express contrastive opinions about the same topic. This has led to an increasing number of studies about contrastive opinion mining. However, there are several notable issues with the existing studies. They mostly focus on mining contrastive opinions from multiple data collections, which need to be separated into their respective collections beforehand. In addition, existing models are opaque in terms of the relationship between topics that are extracted and the sentences in the corpus which express the topics; this opacity does not help us understand the opinions expressed in the corpus. Finally, contrastive opinion is mostly analysed qualitatively rather than quantitatively. This paper addresses these matters and proposes a novel unified latent variable model (contraLDA), which: mines contrastive opinions from both single and multiple data collections, extracts the sentences that project the contrastive opinion, and measures the strength of opinion contrastiveness towards the extracted topics. Experimental results show the effectiveness of our model in mining contrasted opinions, which outperformed our baselines in extracting coherent and informative sentiment-bearing topics. We further show the accuracy of our model in classifying topics and sentiments of textual data, and we compared our results to five strong baselines.

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

Access this article

Log in via an institution

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

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

References

  1. Fang Y, Si L, Somasundaram N, Yu Z. Mining contrastive opinions on political texts using cross-perspective topic model. In: Proceedings of the International Conference on Web Search and Data Mining. 2012, 63–72

    Google Scholar 

  2. Trabelsi A, Zaïane O R. A joint topic viewpoint model for contention analysis. In: Proceedings of the International Conference on Applications of Natural Language to Data Bases/Information Systems. 2014, 114–125

    Google Scholar 

  3. Lerman K, McDonald R. Contrastive summarization: an experiment with consumer reviews. In: Proceedings of the HLT Annual Conference of the North American Chapter of the Association for Computational Linguistics. 2009, 113–116

    Google Scholar 

  4. Paul M, Girju R. Cross-cultural analysis of blogs and forums with mixed-collection topic models. In: Proceedings of the ACL Conference on Empirical Methods in Natural Language Processing. 2009, 1408–1417

    Google Scholar 

  5. Elahi M F, Monachesi P. An examination of cross-cultural similarities and differences from social media data with respect to language use. In: Proceedings of the 8th International Conference on Language Resources and Evaluation. 2012, 4080–4086

    Google Scholar 

  6. Ibeke E, Lin C, Wyner A, Barawi M H. Extracting and understanding contrastive opinion through topic relevant sentences. In: Proceedings of the 8th International Joint Conference on Natural Language Processing. 2017, 395–400

    Google Scholar 

  7. Barawi M H, Lin C, Siddharthan A. Automatically labelling sentiment-bearing topics with descriptive sentence labels. In: Proceedings of the 22nd International Conference on Natural Language and Information Systems. 2017, 299–312

    Google Scholar 

  8. Zhai C, Velivelli A, Yu B. A cross-collection mixture model for comparative text mining. In: Proceedings of the International Conference on Knowledge Discovery and Data Mining. 2004, 743–748

    Google Scholar 

  9. Ahmed A, Xing E P. Staying informed: supervised and semi-supervised multi-view topical analysis of ideological perspective. In: Proceedings of the Conference on Empirical Methods in Natural Language Processing. 2010, 1140–1150

    Google Scholar 

  10. Mukherjee A, Liu B. Mining contentions from discussions and debates. In: Proceedings of the International Conference on Knowledge Discovery and Data Mining. 2012, 841–849

    Google Scholar 

  11. Thonet T, Cabanac G, Boughanem M, Pinel-Sauvagnat K. VODUM: a topic model unifying viewpoint, topic and opinion discovery. In: Proceedings of the European Conference on Information Retrieval. 2016, 533–545

    Google Scholar 

  12. Paul M J, Zhai C, Girju R. Summarizing contrastive viewpoints in opinionated text. In: Proceedings of the Conference on Empirical Methods in Natural Language Processing. 2010, 66–76

    Google Scholar 

  13. Paul M J, Girju R. A two-dimensional topic-aspect model for discovering multi-faceted topics. Urbana, 2010, 51(61801): 36

    Google Scholar 

  14. Guo J, Lu Y, Mori T, Blake C. Expert-guided contrastive opinion summarization for controversial issues. In: Proceedings of the 24th International Conference on World Wide Web Companion. 2015, 1105–1110

    Chapter  Google Scholar 

  15. Ren Z, de Rijke M. Summarizing contrastive themes via hierarchical non-parametric processes. In: Proceedings of the 38th SIGIR International Conference on Research and Development in Information Retrieval. 2015, 93–102

    Google Scholar 

  16. He L, Li W, Zhuge H. Exploring differential topic models for comparative summarization of scientific papers. In: Proceedings of COLING International Conference on Computational Linguistics. 2016, 1028–1038

    Google Scholar 

  17. Nakasaki H, Kawaba M, Utsuro T, Fukuhara T. Mining cross-lingual/cross-cultural differences in concerns and opinions in blogs. In: Proceedings of the 22nd International Conferee on Computer Processing of Oriental Languages. Language Technology for the Knowledge-based Economy. 2009, 213–224

    Chapter  Google Scholar 

  18. Guo H, Zhu H, Guo Z, Zhang X, Su Z. Opinionlt: a text mining system for cross-lingual opinion analysis. In: Proceedings of the 19th ACM International Conference on Information and Knowledge Management. 2010, 1199–1208

    Google Scholar 

  19. Gutiérrez E D, Shutova E, Lichtenstein P, de Melo G, Gilardi L. Detecting cross-cultural differences using a multil, 2016, 4: 47–60

    Google Scholar 

  20. Lin C, He Y. Joint sentiment/topic model for sentiment analysis. In: Proceedings of the 18th ACM conference on Information and Knowledge Management. 2009, 375–384

    Google Scholar 

  21. Lin C, He Y, Everson R, Ruger S. Weakly supervised joint sentiment-topic detection from text. Journal of IEEE Transactions on Knowledge and Data Engineering, 2012, 24(6): 1134–1145

    Article  Google Scholar 

  22. Wallach H, Mimno D, McCallum A. Rethinking LDA: why priors matter. Advances in Neural Information Processing Systems, 2009, 22: 1973–1981

    Google Scholar 

  23. Minka T. Estimating a dirichlet distribution. Technical Report, 2003

    Google Scholar 

  24. Ibeke E, Lin C, Coe C, Wyner A, Liu D, Barawi M H, Yusof N F A. A curated corpus for sentiment-topic analysis. In: Proceedings of the LREC 2016 Workshop on Emotion and Sentiment Analysis. 2016, 32–39

    Google Scholar 

  25. Chang J, Gerrish S, Wang C, Boyd-Graber J L, Jordan L, Blei D M. Reading tea leaves: how humans interpret topic models. In: Proceedings of the 22nd Conference on Neural Information Processing Systems. 2009, 288–296

    Google Scholar 

  26. Mimno D, Wallach H M, Talley E, Leenders M, McCallum A. Optimizing semantic coherence in topic models. In: Proceedings of the Conference on Empirical Methods in Natural Language Processing. 2011, 262–272

    Google Scholar 

  27. Bouma G. Normalized (pointwise) mutual information in collocation extraction. In: Proceedings of the German Society for Computational Linguistics and Language Technology. 2009, 31–40

    Google Scholar 

  28. Newman D, Lau J H, Grieser K, Baldwin T. Automatic evaluation of topic coherence. In: Proceedings of the HLT Annual Conference of the North American Chapter of the Association for Computational Linguistics. 2010, 100–108

    Google Scholar 

  29. Aletras N, Stevenson M. Evaluating topic coherence using distributional semantics. In: Proceedings of the 10th International Conference on Computational Semantics. 2013, 13–22

    Google Scholar 

  30. Steyvers M, Griffiths T. Probabilistic topic models. Handbook of Latent Semantic Analysis, 2007, 427(7): 424–440

    Google Scholar 

  31. Cano A B, He Y, Xu R. Automatic labelling of topic models learned from Twitter by summarisation. In: Proceedings of the 52nd Annual Meeting of Association for Computational Linguists. 2014, 618–624

    Google Scholar 

  32. Ramage D, Hall D, Nallapati R, Manning C D. Labeled LDA: a supervised topic model for credit attribution in multi-labeled corpora. In: Proceedings of the Conference on Empirical Methods in Natural Language Processing. 2009, 248–256

    Google Scholar 

  33. Mcauliffe J D, Blei D M. Supervised topic models. In: Proceedings of the 21st Annual Conference on in Neural Information Processing Systems. 2008, 121–128

    Google Scholar 

  34. Kim Y. Convolutional neural networks for sentence classification. In: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing. 2014, 1746–1751

    Google Scholar 

  35. Ganu G, Marian A, Elhadad N. URSA-user review structure analysis: understanding online reviewing trends. DCS Technical Report: Cite-seer, 2010

    Google Scholar 

Download references

Acknowledgments

This work was supported by the award made by the UK Engineering and Physical Sciences Research Council (EP/P005810/1).

Author information

Authors and Affiliations

  1. Department of Computing Science, University of Aberdeen, Aberdeen, AB24 3FX, UK

    Ebuka Ibeke, Chenghua Lin, Adam Wyner & Mohamad Hardyman Barawi

Authors
  1. Ebuka Ibeke
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Chenghua Lin
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Adam Wyner
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Mohamad Hardyman Barawi
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Chenghua Lin.

Additional information

Ebuka Ibeke is currently a PhD student at the University of Aberdeen and a Part-time Lecturer at the School of Computing Science and Digital Media, Robert Gordon University, UK. He received his BSc degree (2.1 division) in Computer Science from Nnamdi Azikiwe University, Nigeria in 2005 and an MSc degree (1st class Honours) in Information Engineering from Robert Gordon University, UK in 2011. His research interests include sentiment analysis, topic modelling, argumentation mining, and natural language generation/processing.

Chenghua Lin received the BEng degree in electrical engineering and automation from Beihang University, China in 2006, the MEng degree (First class Honours) in electronic engineering from the University of Reading (2007), and the PhD degree in computer science from the University of Exeter (2011). Currently, he is a SICSA Senior Lecturer (Associate Professor) in Computing Science at the University of Aberdeen, UK. His current research interests include integration of machine learning and natural language processing for sentiment analysis, intention mining, text summarisation, and natural language generation.

Adam Wyner is a Lecturer in Computing Science at the University of Aberdeen, UK. He has a PhD in Linguistics (Cornell University, USA, 1994) on the formal syntax and semantics of adverbial modification as well as a second PhD in Computer Science (King's College London, UK, 2008) on the representation and automation of legal concepts for e-contracting. He has worked as a research associate on two EU projects to formalise the law and argumentation to automatically process and to support policymaking. He is currently a co-investigator in a project to make historical legal texts machine readable. He has numerous publications on legal informatics, text analysis, language processing, and argumentation.

Mohamad Hardyman Barawi is currently a PhD student at the University of Aberdeen, UK and a Lecturer at the University Malaysia Sarawak, Malaysia. Prior to the academic post, he was operation support speciahst for Hewlett Packard Malaysia. He received his BSc degree (Hons) in Computer Science from University Putra Malaysia in 2003 and an MLIS degree in Library and Information Science from University of Malaya, Malaysia in 2006. His research interests include topic modelling, sentiment analysis, and text summarisation.

Electronic supplementary material

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ibeke, E., Lin, C., Wyner, A. et al. A unified latent variable model for contrastive opinion mining. Front. Comput. Sci. 14, 404–416 (2020). https://doi.org/10.1007/s11704-018-7073-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11704-018-7073-5

Keywords

Search

Navigation