{"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/remedying-bilstm-cnn-deficiency-in-modeling","title":"Why Attention? Analyze BiLSTM Deficiency and Its Remedies in the Case of NER","arxiv_id":"1908.11046","date":"2019-08-29","proceeding":null,"authors":["Peng-Hsuan Li","Tsu-Jui Fu","Wei-Yun Ma"],"abstract":"BiLSTM has been prevalently used as a core module for NER in a sequence-labeling setup. State-of-the-art approaches use BiLSTM with additional resources such as gazetteers, language-modeling, or multi-task supervision to further improve NER. This paper instead takes a step back and focuses on analyzing problems of BiLSTM itself and how exactly self-attention can bring improvements. We formally show the limitation of (CRF-)BiLSTM in modeling cross-context patterns for each word -- the XOR limitation. Then, we show that two types of simple cross-structures -- self-attention and Cross-BiLSTM -- can effectively remedy the problem. We test the practical impacts of the deficiency on real-world NER datasets, OntoNotes 5.0 and WNUT 2017, with clear and consistent improvements over the baseline, up to 8.7% on some of the multi-token entity mentions. We give in-depth analyses of the improvements across several aspects of NER, especially the identification of multi-token mentions. This study should lay a sound foundation for future improvements on sequence-labeling NER. (Source codes: https://github.com/jacobvsdanniel/cross-ner)","url_abs":"https://arxiv.org/abs/1908.11046v3","url_pdf":"https://arxiv.org/pdf/1908.11046v3.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":"remedying-bilstm-cnn-deficiency-in-modeling","repo_url":"https://github.com/jacobvsdanniel/cross-ner","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"remedying-bilstm-cnn-deficiency-in-modeling","repo_url":"https://github.com/jacobvsdanniel/cross_ner","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"remedying-bilstm-cnn-deficiency-in-modeling","repo_url":"https://github.com/ckiplab/ckiptagger","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"cg","task_name":"NER"},{"task_slug":"named-entity-recognition-ner","task_name":"Named Entity Recognition (NER)"}],"methods":[{"method_slug":"bilstm","method_name":"BiLSTM"},{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/named-entity-recognition-ner-on-ontonotes-v5","task":"Named Entity Recognition (NER)","dataset":"Ontonotes v5 (English)","model":"Att-BiLSTM-CNN","rank_in_archive_order":20,"of":28,"metrics":{"F1":"88.4","Precision":"88.71","Recall":"88.11"},"uses_additional_data":false},{"leaderboard":"/sota/named-entity-recognition-on-wnut-2017","task":"Named Entity Recognition (NER)","dataset":"WNUT 2017","model":"Cross-BiLSTM-CNN","rank_in_archive_order":20,"of":23,"metrics":{"F1":"42.85","Precision":"58.28","Recall":"33.92"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}