{"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/latent-tree-models-for-hierarchical-topic","title":"Latent Tree Models for Hierarchical Topic Detection","arxiv_id":"1605.06650","date":"2016-05-21","proceeding":null,"authors":["Peixian Chen","Nevin L. Zhang","Tengfei Liu","Leonard K. M. Poon","Zhourong Chen","Farhan Khawar"],"abstract":"We present a novel method for hierarchical topic detection where topics are\nobtained by clustering documents in multiple ways. Specifically, we model\ndocument collections using a class of graphical models called hierarchical\nlatent tree models (HLTMs). The variables at the bottom level of an HLTM are\nobserved binary variables that represent the presence/absence of words in a\ndocument. The variables at other levels are binary latent variables, with those\nat the lowest latent level representing word co-occurrence patterns and those\nat higher levels representing co-occurrence of patterns at the level below.\nEach latent variable gives a soft partition of the documents, and document\nclusters in the partitions are interpreted as topics. Latent variables at high\nlevels of the hierarchy capture long-range word co-occurrence patterns and\nhence give thematically more general topics, while those at low levels of the\nhierarchy capture short-range word co-occurrence patterns and give thematically\nmore specific topics. Unlike LDA-based topic models, HLTMs do not refer to a\ndocument generation process and use word variables instead of token variables.\nThey use a tree structure to model the relationships between topics and words,\nwhich is conducive to the discovery of meaningful topics and topic hierarchies.","url_abs":"http://arxiv.org/abs/1605.06650v2","url_pdf":"http://arxiv.org/pdf/1605.06650v2.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":"latent-tree-models-for-hierarchical-topic","repo_url":"https://github.com/kmpoon/hlta","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"GPL-3.0"}}],"tasks":[{"task_slug":"clustering","task_name":"Clustering"},{"task_slug":"topic-models","task_name":"Topic Models"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1605.06650","atlas_url":"https://app.syntology.ai/?focus=1605.06650","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}