Judicial nested named entity recognition method with MRC framework
Judicial named entity recognition (JNER) is a basic task of judicial intelligence and judicial service informatization. At present, the research of JNER has attracted extensive attention. However, the existing JNER methods usually can only assign a single label to a token in the input sequence, whic...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
KeAi Communications Co., Ltd.
2023-06-01
|
| Series: | International Journal of Cognitive Computing in Engineering |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2666307423000128 |
| _version_ | 1797377856587694080 |
|---|---|
| author | Hu Zhang Jiayu Guo Yujie Wang Zhen Zhang Hansen Zhao |
| author_facet | Hu Zhang Jiayu Guo Yujie Wang Zhen Zhang Hansen Zhao |
| author_sort | Hu Zhang |
| collection | DOAJ |
| description | Judicial named entity recognition (JNER) is a basic task of judicial intelligence and judicial service informatization. At present, the research of JNER has attracted extensive attention. However, the existing JNER methods usually can only assign a single label to a token in the input sequence, which is not applicable to nested entities where a token may be assigned two or more different labels at the same time. Therefore, this paper introduces the machine reading comprehension (MRC) framework into JNER, and proposes a judicial nested NER method based on the MRC. Firstly, we design the question template according to the characteristics of judicial nested named entities, and construct the legal text named entity dataset in MRC format. Next, we introduce the span extraction MRC model based on the pre-trained to encode the question and text, and learn the context knowledge of the entity in the question. Finally, we extract the starting and end positions of the matching span respectively through two classifiers, to get the corresponding entities. The experimental results on the information extraction dataset in “CAIL2021” show, compared with the existing baseline models, the proposed method effectively improves the recognition effect of nested entities commonly existing in the judicial field. |
| first_indexed | 2024-03-08T19:58:24Z |
| format | Article |
| id | doaj.art-a2a46172c023441ba404239b57d819c6 |
| institution | Directory Open Access Journal |
| issn | 2666-3074 |
| language | English |
| last_indexed | 2024-03-08T19:58:24Z |
| publishDate | 2023-06-01 |
| publisher | KeAi Communications Co., Ltd. |
| record_format | Article |
| series | International Journal of Cognitive Computing in Engineering |
| spelling | doaj.art-a2a46172c023441ba404239b57d819c62023-12-24T04:46:38ZengKeAi Communications Co., Ltd.International Journal of Cognitive Computing in Engineering2666-30742023-06-014118126Judicial nested named entity recognition method with MRC frameworkHu Zhang0Jiayu Guo1Yujie Wang2Zhen Zhang3Hansen Zhao4School of Computer and Information Technology, Shanxi University, Taiyuan 030006, Shanxi, China; Corresponding author.School of Computer and Information Technology, Shanxi University, Taiyuan 030006, Shanxi, ChinaSchool of Computer and Information Technology, Shanxi University, Taiyuan 030006, Shanxi, ChinaSchool of Computer and Information Technology, Shanxi University, Taiyuan 030006, Shanxi, ChinaDepartment of Computation Science and Engineering, George Institute of Technology, Atlanta, USAJudicial named entity recognition (JNER) is a basic task of judicial intelligence and judicial service informatization. At present, the research of JNER has attracted extensive attention. However, the existing JNER methods usually can only assign a single label to a token in the input sequence, which is not applicable to nested entities where a token may be assigned two or more different labels at the same time. Therefore, this paper introduces the machine reading comprehension (MRC) framework into JNER, and proposes a judicial nested NER method based on the MRC. Firstly, we design the question template according to the characteristics of judicial nested named entities, and construct the legal text named entity dataset in MRC format. Next, we introduce the span extraction MRC model based on the pre-trained to encode the question and text, and learn the context knowledge of the entity in the question. Finally, we extract the starting and end positions of the matching span respectively through two classifiers, to get the corresponding entities. The experimental results on the information extraction dataset in “CAIL2021” show, compared with the existing baseline models, the proposed method effectively improves the recognition effect of nested entities commonly existing in the judicial field.http://www.sciencedirect.com/science/article/pii/S2666307423000128Named entity recognitionMachine reading comprehensionNested named entitiesWisdom justice |
| spellingShingle | Hu Zhang Jiayu Guo Yujie Wang Zhen Zhang Hansen Zhao Judicial nested named entity recognition method with MRC framework International Journal of Cognitive Computing in Engineering Named entity recognition Machine reading comprehension Nested named entities Wisdom justice |
| title | Judicial nested named entity recognition method with MRC framework |
| title_full | Judicial nested named entity recognition method with MRC framework |
| title_fullStr | Judicial nested named entity recognition method with MRC framework |
| title_full_unstemmed | Judicial nested named entity recognition method with MRC framework |
| title_short | Judicial nested named entity recognition method with MRC framework |
| title_sort | judicial nested named entity recognition method with mrc framework |
| topic | Named entity recognition Machine reading comprehension Nested named entities Wisdom justice |
| url | http://www.sciencedirect.com/science/article/pii/S2666307423000128 |
| work_keys_str_mv | AT huzhang judicialnestednamedentityrecognitionmethodwithmrcframework AT jiayuguo judicialnestednamedentityrecognitionmethodwithmrcframework AT yujiewang judicialnestednamedentityrecognitionmethodwithmrcframework AT zhenzhang judicialnestednamedentityrecognitionmethodwithmrcframework AT hansenzhao judicialnestednamedentityrecognitionmethodwithmrcframework |