{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,2]],"date-time":"2026-02-02T20:13:33Z","timestamp":1770063213814,"version":"3.49.0"},"reference-count":15,"publisher":"China Science Publishing & Media Ltd.","issue":"4","license":[{"start":{"date-parts":[[2021,5,24]],"date-time":"2021-05-24T00:00:00Z","timestamp":1621814400000},"content-version":"vor","delay-in-days":143,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2021,10,25]]},"abstract":"<jats:p>Currently, as a basic task of military document information extraction, Named Entity Recognition (NER) for military documents has received great attention. In 2020, China Conference on Knowledge Graph and Semantic Computing (CCKS) and System Engineering Research Institute of Academy of Military Sciences (AMS) issued the NER task for test evaluation, which requires the recognition of four types of entities including Test Elements (TE), Performance Indicators (PI), System Components (SC) and Task Scenarios (TS). Due to the particularity and confidentiality of the military field, only 400 items of annotated data are provided by the organizer. In this paper, the task is regarded as a few-shot learning problem for NER, and a method based on BERT and two-level model fusion is proposed. Firstly, the proposed method is based on several basic models fine tuned by BERT on the training data. Then, a two-level fusion strategy applied to the prediction results of multiple basic models is proposed to alleviate the over-fitting problem. Finally, the labeling errors are eliminated by post-processing. This method achieves F1 score of 0.7203 on the test set of the evaluation task.<\/jats:p>","DOI":"10.1162\/dint_a_00102","type":"journal-article","created":{"date-parts":[[2021,5,24]],"date-time":"2021-05-24T20:34:17Z","timestamp":1621888457000},"page":"568-577","update-policy":"https:\/\/doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":22,"title":["Few-shot Learning for Named Entity Recognition Based on BERT and\n                    Two-level Model Fusion"],"prefix":"10.3724","volume":"3","author":[{"given":"Yuan","family":"Gong","sequence":"first","affiliation":[{"name":"AI Lab, KingSoft Corp, Beijing 100086, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Lu","family":"Mao","sequence":"additional","affiliation":[{"name":"AI Lab, KingSoft Corp, Beijing 100086, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Changliang","family":"Li","sequence":"additional","affiliation":[{"name":"AI Lab, KingSoft Corp, Beijing 100086, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"2026","published-online":{"date-parts":[[2021,10,25]]},"reference":[{"key":"2021102517000361800_ref1","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2020.2981314","volume-title":"A survey on deep learning for named entity recognition","author":"Li","year":"2020"},{"key":"2021102517000361800_ref2","volume-title":"BERT: Pre-training of deep bidirectional transformers for language\n                        understanding","author":"Devlin","year":"2018"},{"key":"2021102517000361800_ref3","first-page":"28","article-title":"Transfer learning for scientific data chain extraction in\n                        small chemical corpus with joint BERT-CRF model","volume-title":"BIRNDL SIGIR","author":"Pang","year":"2019"},{"key":"2021102517000361800_ref4","doi-asserted-by":"crossref","DOI":"10.1088\/1742-6596\/1267\/1\/012017","volume-title":"New\n                        research on transfer learning model of named entity recognition","author":"Guan","year":"2019"},{"issue":"1","key":"2021102517000361800_ref5","first-page":"68","article-title":"Big data analysis by infinite deep neural\n                        networks","volume":"53","author":"Lei","year":"2016","journal-title":"Journal of Computer Research and\n                        Development"},{"key":"2021102517000361800_ref6","volume-title":"A maximum entropy approach to named entity recognition","author":"Borthwick","year":"1999"},{"key":"2021102517000361800_ref7","first-page":"473","article-title":"Named entity recognition using an HMM-based chunk\n                        tagger","volume-title":"Proceedings of the 40th Annual Meeting\n                        of the Association for Computational Linguistics","author":"Zhou","year":"2002"},{"key":"2021102517000361800_ref8","first-page":"1","article-title":"Named entity recognition from biomedical text using\n                        SVM","volume-title":"The 5th International Conference on\n                        Bioinformatics and Biomedical Engineering","author":"Ju","year":"2011"},{"key":"2021102517000361800_ref9","doi-asserted-by":"crossref","first-page":"229","DOI":"10.1109\/NLPKE.2007.4368038","article-title":"Automatic recognition of Chinese organization name based on\n                        conditional random fields","volume-title":"2007 International\n                        Conference on Natural Language Processing and Knowledge\n                    Engineering","author":"Zhang","year":"2007"},{"key":"2021102517000361800_ref10","volume-title":"A survey on recent advances in named entity recognition from deep\n                        learning models","author":"Yadav","year":"2019"},{"key":"2021102517000361800_ref11","volume-title":"Bidirectional LSTM-CRF models for sequence tagging","author":"Huang","year":"2015"},{"issue":"76","key":"2021102517000361800_ref12","first-page":"2493","article-title":"Natural language processing (almost) from\n                        scratch","volume":"12","author":"Collobert","year":"2011","journal-title":"Journal of Machine Learning\n                        Research"},{"key":"2021102517000361800_ref13","doi-asserted-by":"crossref","first-page":"1532","DOI":"10.3115\/v1\/D14-1162","article-title":"Glove: Global vectors for word\n                    representation","volume-title":"Proceedings of the 2014 Conference\n                        on Empirical Methods in Natural Language Processing (EMNLP)","author":"Pennington","year":"2014"},{"key":"2021102517000361800_ref14","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/ICSESS47205.2019.9040736","article-title":"Research on Chinese naming recognition model based on BERT\n                        embedding","volume-title":"2019 IEEE 10th International\n                        Conference on Software Engineering and Service Science (ICSESS)","author":"Cai","year":"2019"},{"key":"2021102517000361800_ref15","doi-asserted-by":"crossref","DOI":"10.1088\/1742-6596\/1550\/3\/032149","volume-title":"Named entity recognition method of Brazilian legal text based on\n                        pre-training model","author":"Wang","year":"2020"}],"container-title":["Data Intelligence"],"original-title":[],"language":"en","link":[{"URL":"http:\/\/direct.mit.edu\/dint\/article-pdf\/3\/4\/568\/1968571\/dint_a_00102.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"http:\/\/direct.mit.edu\/dint\/article-pdf\/3\/4\/568\/1968571\/dint_a_00102.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,3,14]],"date-time":"2025-03-14T07:42:31Z","timestamp":1741938151000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.sciengine.com\/doi\/10.1162\/dint_a_00102"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021]]},"references-count":15,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2021,10,25]]}},"URL":"https:\/\/doi.org\/10.1162\/dint_a_00102","relation":{},"ISSN":["2641-435X"],"issn-type":[{"value":"2641-435X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2021]]},"published":{"date-parts":[[2021]]}}}