{"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/label-noise-reduction-in-entity-typing-by","title":"Label Noise Reduction in Entity Typing by Heterogeneous Partial-Label Embedding","arxiv_id":"1602.05307","date":"2016-02-17","proceeding":null,"authors":["Xiang Ren","Wenqi He","Meng Qu","Clare R. Voss","Heng Ji","Jiawei Han"],"abstract":"Current systems of fine-grained entity typing use distant supervision in\nconjunction with existing knowledge bases to assign categories (type labels) to\nentity mentions. However, the type labels so obtained from knowledge bases are\noften noisy (i.e., incorrect for the entity mention's local context). We define\na new task, Label Noise Reduction in Entity Typing (LNR), to be the automatic\nidentification of correct type labels (type-paths) for training examples, given\nthe set of candidate type labels obtained by distant supervision with a given\ntype hierarchy. The unknown type labels for individual entity mentions and the\nsemantic similarity between entity types pose unique challenges for solving the\nLNR task. We propose a general framework, called PLE, to jointly embed entity\nmentions, text features and entity types into the same low-dimensional space\nwhere, in that space, objects whose types are semantically close have similar\nrepresentations. Then we estimate the type-path for each training example in a\ntop-down manner using the learned embeddings. We formulate a global objective\nfor learning the embeddings from text corpora and knowledge bases, which adopts\na novel margin-based loss that is robust to noisy labels and faithfully models\ntype correlation derived from knowledge bases. Our experiments on three public\ntyping datasets demonstrate the effectiveness and robustness of PLE, with an\naverage of 25% improvement in accuracy compared to next best method.","url_abs":"http://arxiv.org/abs/1602.05307v1","url_pdf":"http://arxiv.org/pdf/1602.05307v1.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":"label-noise-reduction-in-entity-typing-by","repo_url":"https://github.com/shanzhenren/PLE","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"GPL-3.0"}},{"paper_slug":"label-noise-reduction-in-entity-typing-by","repo_url":"https://github.com/INK-USC/AFET","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"GPL-3.0"}},{"paper_slug":"label-noise-reduction-in-entity-typing-by","repo_url":"https://github.com/shanzhenren/AFET","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"GPL-3.0"}}],"tasks":[{"task_slug":"entity-typing","task_name":"Entity Typing"},{"task_slug":"semantic-similarity","task_name":"Semantic Similarity"},{"task_slug":"semantic-textual-similarity","task_name":"Semantic Textual Similarity"},{"task_slug":"type","task_name":"Vocal Bursts Type Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1602.05307","atlas_url":"https://app.syntology.ai/?focus=1602.05307","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}