{"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/from-pos-tagging-to-dependency-parsing-for","title":"From POS tagging to dependency parsing for biomedical event extraction","arxiv_id":"1808.03731","date":"2018-08-11","proceeding":null,"authors":["Dat Quoc Nguyen","Karin Verspoor"],"abstract":"Background: Given the importance of relation or event extraction from\nbiomedical research publications to support knowledge capture and synthesis,\nand the strong dependency of approaches to this information extraction task on\nsyntactic information, it is valuable to understand which approaches to\nsyntactic processing of biomedical text have the highest performance. Results:\nWe perform an empirical study comparing state-of-the-art traditional\nfeature-based and neural network-based models for two core natural language\nprocessing tasks of part-of-speech (POS) tagging and dependency parsing on two\nbenchmark biomedical corpora, GENIA and CRAFT. To the best of our knowledge,\nthere is no recent work making such comparisons in the biomedical context;\nspecifically no detailed analysis of neural models on this data is available.\nExperimental results show that in general, the neural models outperform the\nfeature-based models on two benchmark biomedical corpora GENIA and CRAFT. We\nalso perform a task-oriented evaluation to investigate the influences of these\nmodels in a downstream application on biomedical event extraction, and show\nthat better intrinsic parsing performance does not always imply better\nextrinsic event extraction performance. Conclusion: We have presented a\ndetailed empirical study comparing traditional feature-based and neural\nnetwork-based models for POS tagging and dependency parsing in the biomedical\ncontext, and also investigated the influence of parser selection for a\nbiomedical event extraction downstream task. Availability of data and material:\nWe make the retrained models available at\nhttps://github.com/datquocnguyen/BioPosDep","url_abs":"http://arxiv.org/abs/1808.03731v2","url_pdf":"http://arxiv.org/pdf/1808.03731v2.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":"from-pos-tagging-to-dependency-parsing-for","repo_url":"https://github.com/datquocnguyen/BioNLP","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"from-pos-tagging-to-dependency-parsing-for","repo_url":"https://github.com/datquocnguyen/BioPosDep","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"dependency-parsing","task_name":"Dependency Parsing"},{"task_slug":"event-extraction","task_name":"Event Extraction"},{"task_slug":"pos","task_name":"POS"},{"task_slug":"pos-tagging","task_name":"POS Tagging"},{"task_slug":"part-of-speech-tagging","task_name":"Part-Of-Speech Tagging"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/dependency-parsing-on-genia-las","task":"Dependency Parsing","dataset":"GENIA - LAS","model":"BiLSTM-CRF","rank_in_archive_order":1,"of":3,"metrics":{"F1":"91.92"},"uses_additional_data":false},{"leaderboard":"/sota/dependency-parsing-on-genia-uas","task":"Dependency Parsing","dataset":"GENIA - UAS","model":"BiLSTM-CRF","rank_in_archive_order":1,"of":3,"metrics":{"F1":"92.84"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}