ABSTRACT
Traditional topic modeling approaches generally rely on document-term co-occurrence statistics to find latent topics in a collection of documents. However, relying only on such statistics can yield incoherent or hard to interpret results for the end-users in many applications where the interest lies in interpreting the resulting topics (e.g. labeling documents, comparing corpora, guiding content exploration, etc.). In this work, we propose to leverage external common sense knowledge, i.e. information from the real world beyond word co-occurrence, to find topics that are more coherent and more easily interpretable by humans. We introduce the Common Sense Topic Model (CSTM), a novel and efficient approach that augments clustering with knowledge extracted from the ConceptNet knowledge graph. We evaluate this approach on several datasets alongside commonly used models using both automatic and human evaluation, and we show how it shows superior affinity to human judgement. The code for the experiments as well as the training data and human evaluation are available at https://github.com/D2KLab/CSTM.
- Mehdi Allahyari, Seyedamin Pouriyeh, Krys Kochut, and Hamid Reza Arabnia.Google Scholar
- . A Knowledge-based Topic Modeling Approach for Automatic Topic Labeling. IJACSA 2017 ([n. d.]).Google Scholar
- Dr. Hiteshwar Kumar Azad and A. Deepak. 2019. Query Expansion Techniques for Information Retrieval: a Survey. Inf. Process. Manag. 56 (2019), 1698--1735.Google ScholarDigital Library
- DavidMBlei, AndrewY Ng, and Michael I Jordan. 2003. Latent dirichlet allocation. the Journal of machine Learning research 3 (2003), 993--1022.Google Scholar
- Jonathan Chang, Jordan Boyd-Graber, Sean Gerrish, Chong Wang, and David M. Blei. 2009. Reading Tea Leaves: How Humans Interpret Topic Models (NIPS'09). Red Hook, NY, USA, 288--296.Google ScholarDigital Library
- Zhiyuan Chen, Arjun Mukherjee, Bing Liu, Meichun Hsu, Malu Castellanos, and Riddhiman Ghosh. 2013. Discovering Coherent Topics Using General Knowledge. In CIKM '13 (San Francisco, California, USA). New York, NY, USA, 209--218.Google Scholar
- Caitlin Doogan and Wray Buntine. 2021. Topic Model or Topic Twaddle? Reevaluating Semantic Interpretability Measures. In NAACL '21.Google Scholar
- Anjie Fang, Craig Macdonald, Iadh Ounis, and Philip Habel. 2016. Using Word Embedding to Evaluate the Coherence of Topics from Twitter Data. In SIGIR '16 (Pisa, Italy) (SIGIR '16). New York, NY, USA.Google ScholarDigital Library
- Adriana Ferrugento, Ana Alves, Hugo Gonçalo Oliveira, and Filipe Rodrigues. 1957. A synopsis of linguistic theory 1930--1955.Google Scholar
- Adriana Ferrugento, Ana Alves, Hugo Gonçalo Oliveira, and Filipe Rodrigues. 2015. Towards the Improvement of a Topic Model with Semantic Knowledge, Vol. 9273. Portuguese Conference on Artificial Intelligence, 759--770.Google Scholar
- Derek Greene and Pádraig Cunningham. 2006. Practical Solutions to the Problem of Diagonal Dominance in Kernel Document Clustering. In ICML 2006.Google ScholarDigital Library
- Antonio Gulli. 2005. AG's corpus of news articles.Google Scholar
- Ismail Harrando and Raphaël Troncy. 2021. Explainable Zero-Shot Topic Extraction Using a Common-Sense Knowledge Graph. In LDK 2021. Dagstuhl, Germany.Google Scholar
- Alexander Miserlis Hoyle, Pranav Goel, Denis Peskov, Andrew Hian-Cheong, Jordan L. Boyd-Graber, and P. Resnik. 2021. Is Automated Topic Model Evaluation Broken?: The Incoherence of Coherence. ArXiv abs/2107.02173 (2021).Google Scholar
- Ming-Hung Hsu, Ming-Feng Tsai, and Hsin-Hsi Chen. 2006. Query Expansion with ConceptNet andWordNet: An Intrinsic Comparison. In Information Retrieval Technology. Berlin, Heidelberg.Google Scholar
- Filip Ilievski, Pedro Szekely, and Bin Zhang. 2021. CSKG: The CommonSense Knowledge Graph. Extended Semantic Web Conference (ESWC) (2021).Google Scholar
- Ken Lang. 1995. Newsweeder: Learning to filter netnews. In 12??? International Conference on Machine Learning (ICML). 331--339.Google ScholarCross Ref
- Natalia Loukachevitch, Michael Nokel, and Kirill Ivanov. 2018. Combining Thesaurus Knowledge and Probabilistic Topic Models. In Analysis of Images, Social Networks and Texts.Google Scholar
- David Mimno, Hanna M.Wallach, Edmund Talley, Miriam Leenders, and Andrew McCallum. 2011. Optimizing Semantic Coherence in Topic Models. In EMNLP '11 (Edinburgh, United Kingdom). Association for Computational Linguistics, USA.Google Scholar
- Janna Omeliyanenko, Albin Zehe, Lena Hettinger, and Andreas Hotho. [n. d.]. LM4KG: Improving Common Sense Knowledge Graphs with Language Models. In ISWC 2020. Cham.Google ScholarDigital Library
- Andrew Rosenberg and Julia Hirschberg. 2007. V-Measure: A Conditional Entropy-Based External Cluster Evaluation Measure. In EMNLP-CoNLL '07. Association for Computational Linguistics, Prague, Czech Republic, 410--420.Google Scholar
- Charlotte Rudnik, Thibault Ehrhart, Olivier Ferret, Denis Teyssou, Raphaël Troncy, and Xavier Tannier. 2019. Searching News Articles Using an Event Knowledge Graph Leveraged by Wikidata. In 5??? Wiki Workshop. 1232--1239.Google ScholarDigital Library
- Maarten Sap, Ronan Le Bras, Emily Allaway, Chandra Bhagavatula, Nicholas Lourie, Hannah Rashkin, Brendan Roof, Noah A. Smith, and Yejin Choi. [n. d.]. ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning. In AAAI 2019.Google Scholar
- Suzanna Sia, Ayush Dalmia, and Sabrina J. Mielke. 2020. Tired of Topic Models? Clusters of Pretrained Word Embeddings Make for Fast and Good Topics too!. In EMNLP '20. Association for Computational Linguistics, Online, 1728--1736.Google Scholar
- Dandan Song, Jingwen Gao, Jinhui Pang, Lejian Liao, and Lifei Qin. 2020. Knowledge Base Enhanced Topic Modeling. In ICKG 2020. 380--387.Google Scholar
- Robyn Speer, Joshua Chin, and Catherine Havasi. 2017. ConceptNet 5.5: An Open Multilingual Graph of General Knowledge. 4444--4451.Google Scholar
- Ilaria Tiddi, Mathieu d'Aquin, and Enrico Motta. 2015. Using Linked Data Traversal to Label Academic Communities. In WWW 2015 (Florence, Italy) (WWW '15 Companion). New York, NY, USA.Google ScholarDigital Library
- Hanna M. Wallach, Iain Murray, Ruslan Salakhutdinov, and David Mimno. 2009. Evaluation Methods for Topic Models. In ICML '09 (Montreal, Quebec, Canada) (ICML '09). New York, NY, USA, 1105--1112.Google Scholar
- Wei Xu, Xin Liu, and Yihong Gong. 2003. Document Clustering Based on Non-Negative Matrix Factorization (SIGIR '03). Association for Computing Machinery, New York, NY, USA, 267--273.Google Scholar
Index Terms
- Discovering Interpretable Topics by Leveraging Common Sense Knowledge
Recommendations
Jointly Discovering Fine-grained and Coarse-grained Sentiments via Topic Modeling
MM '14: Proceedings of the 22nd ACM international conference on MultimediaThe ever-increasing user-generated contents in social media and other web services make it highly desirable to discover opinions of users on all kinds of topics. Motivated by the assumption that individual word and paragraph in documents will deliver ...
Text, Topics, and Turkers: A Consensus Measure for Statistical Topics
HT '15: Proceedings of the 26th ACM Conference on Hypertext & Social MediaTopic modeling is an important tool in social media analysis, allowing researchers to quickly understand large text corpora by investigating the topics underlying them. One of the fundamental problems of topic models lies in how to assess the quality of ...
Group topic model: organizing topics into groups
AbstractLatent Dirichlet allocation defines hidden topics to capture latent semantics in text documents. However, it assumes that all the documents are represented by the same topics, resulting in the “forced topic” problem. To solve this problem, we ...
Comments