{"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/topiceq-a-joint-topic-and-mathematical","title":"TopicEq: A Joint Topic and Mathematical Equation Model for Scientific Texts","arxiv_id":"1902.06034","date":"2019-02-16","proceeding":null,"authors":["Michihiro Yasunaga","John Lafferty"],"abstract":"Scientific documents rely on both mathematics and text to communicate ideas.\nInspired by the topical correspondence between mathematical equations and word\ncontexts observed in scientific texts, we propose a novel topic model that\njointly generates mathematical equations and their surrounding text (TopicEq).\nUsing an extension of the correlated topic model, the context is generated from\na mixture of latent topics, and the equation is generated by an RNN that\ndepends on the latent topic activations. To experiment with this model, we\ncreate a corpus of 400K equation-context pairs extracted from a range of\nscientific articles from arXiv, and fit the model using a variational\nautoencoder approach. Experimental results show that this joint model\nsignificantly outperforms existing topic models and equation models for\nscientific texts. Moreover, we qualitatively show that the model effectively\ncaptures the relationship between topics and mathematics, enabling novel\napplications such as topic-aware equation generation, equation topic inference,\nand topic-aware alignment of mathematical symbols and words.","url_abs":"http://arxiv.org/abs/1902.06034v3","url_pdf":"http://arxiv.org/pdf/1902.06034v3.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":[],"tasks":[{"task_slug":"articles","task_name":"Articles"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"topic-models","task_name":"Topic Models"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/topic-models-on-arxiv","task":"Topic Models","dataset":"Arxiv HEP-TH citation graph","model":"TopicEq","rank_in_archive_order":2,"of":2,"metrics":{"Topic Coherence@50":"0.097"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1902.06034","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}