{"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/learningword-embeddings-for-low-resource","title":"LearningWord Embeddings for Low-resource Languages by PU Learning","arxiv_id":"1805.03366","date":"2018-05-09","proceeding":null,"authors":["Chao Jiang","Hsiang-Fu Yu","Cho-Jui Hsieh","Kai-Wei Chang"],"abstract":"Word embedding is a key component in many downstream applications in\nprocessing natural languages. Existing approaches often assume the existence of\na large collection of text for learning effective word embedding. However, such\na corpus may not be available for some low-resource languages. In this paper,\nwe study how to effectively learn a word embedding model on a corpus with only\na few million tokens. In such a situation, the co-occurrence matrix is sparse\nas the co-occurrences of many word pairs are unobserved. In contrast to\nexisting approaches often only sample a few unobserved word pairs as negative\nsamples, we argue that the zero entries in the co-occurrence matrix also\nprovide valuable information. We then design a Positive-Unlabeled Learning\n(PU-Learning) approach to factorize the co-occurrence matrix and validate the\nproposed approaches in four different languages.","url_abs":"http://arxiv.org/abs/1805.03366v1","url_pdf":"http://arxiv.org/pdf/1805.03366v1.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":"learningword-embeddings-for-low-resource","repo_url":"https://github.com/uclanlp/PU-Learning-for-Word-Embedding","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1805.03366","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}