{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/re-ranking-words-to-improve-interpretability","title":"Re-Ranking Words to Improve Interpretability of Automatically Generated Topics","arxiv_id":"1903.12542","date":"2019-03-29","proceeding":"WS 2019 5","authors":["Areej Alokaili","Nikolaos Aletras","Mark Stevenson"],"abstract":"Topics models, such as LDA, are widely used in Natural Language Processing.\nMaking their output interpretable is an important area of research with\napplications to areas such as the enhancement of exploratory search interfaces\nand the development of interpretable machine learning models. Conventionally,\ntopics are represented by their n most probable words, however, these\nrepresentations are often difficult for humans to interpret. This paper\nexplores the re-ranking of topic words to generate more interpretable topic\nrepresentations. A range of approaches are compared and evaluated in two\nexperiments. The first uses crowdworkers to associate topics represented by\ndifferent word rankings with related documents. The second experiment is an\nautomatic approach based on a document retrieval task applied on multiple\ndomains. Results in both experiments demonstrate that re-ranking words improves\ntopic interpretability and that the most effective re-ranking schemes were\nthose which combine information about the importance of words both within\ntopics and their relative frequency in the entire corpus. In addition, close\ncorrelation between the results of the two evaluation approaches suggests that\nthe automatic method proposed here could be used to evaluate re-ranking methods\nwithout the need for human judgements.","url_abs":"http://arxiv.org/abs/1903.12542v1","url_pdf":"http://arxiv.org/pdf/1903.12542v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"re-ranking-words-to-improve-interpretability","repo_url":"https://github.com/areejokaili/topic_reranking","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"interpretable-machine-learning","task_name":"Interpretable Machine Learning"},{"task_slug":"re-ranking","task_name":"Re-Ranking"},{"task_slug":"retrieval","task_name":"Retrieval"}],"methods":[{"method_slug":"interpretability","method_name":"Interpretability"},{"method_slug":"lda","method_name":"LDA"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1903.12542","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}