Fumihito Nishino
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Fumihito Nishino has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
Hao Yu
7 shared publications- Ha-Thanh Nguyen
National Institute of Informatics
4 shared publications - Ken Satoh
Center for Juris-Informatics, ROIS-DS, Tokyo, Japan
4 shared publications - May Myo Zin
Center for Juris-Informatics, ROIS-DS, Tokyo, Japan
3 shared publications Yao Meng
Fujitsu R&D Center CO., LTD., Chaoyang District, 100025, Beijing China
3 shared publicationsGaolin Fang
2 shared publications
Latest Publications
- Open Access
JBE-QA: Japanese Bar Exam QA Dataset for Assessing Legal Domain Knowledge
We introduce JBE-QA, a Japanese Bar Exam Question-Answering dataset to evaluate large language models' legal knowledge. Derived from the multiple-choice (tanto-shiki) section of the Japanese bar exam (2015-2024), JBE-QA provides the first comprehensive benchmark for Japanese legal-domain evaluation of LLMs. It covers the Civil Code, the Penal Code, and the Constitution, extending beyond the Civil Code focus of prior Japanese resources. Each question is decomposed into independent true/false judgments with structured contextual fields. The dataset contains 3,464 items with balanced labels. We evaluate 26 LLMs, including proprietary, open-weight, Japanese-specialised, and reasoning models. Our results show that proprietary models with reasoning enabled perform best, and the Constitution questions are generally easier than the Civil Code or the Penal Code questions.
Open AccessJBE-QA: Japanese Bar Exam QA Dataset for Assessing Legal Domain Knowledge
We introduce JBE-QA, a Japanese Bar Exam Question-Answering dataset to evaluate large language models' legal knowledge. Derived from the multiple-choice (tanto-shiki) section of the Japanese bar exam (2015-2024), JBE-QA provides the first comprehensive benchmark for Japanese legal-domain evaluation of LLMs. It covers the Civil Code, the Penal Code, and the Constitution, extending beyond the Civil Code focus of prior Japanese resources. Each question is decomposed into independent true/false judgments with structured contextual fields. The dataset contains 3,464 items with balanced labels. We evaluate 26 LLMs, including proprietary, open-weight, Japanese-specialised, and reasoning models. Our results show that proprietary models with reasoning enabled perform best, and the Constitution questions are generally easier than the Civil Code or the Penal Code questions.
- 28 May 2024
- Published by Springer Nature
- p. 77-92
Addressing Annotated Data Scarcity in Legal Information Extraction
- 28 May 2024
- Published by Springer Nature
- p. 77-92
Addressing Annotated Data Scarcity in Legal Information Extraction
- Open Access
- 7 December 2023
- Published by IOS Press
- p. 177-186
Information Extraction from Lengthy Legal Contracts: Leveraging Query-Based Summarization and GPT-3.5
In the legal domain, extracting information from contracts poses significant challenges, primarily due to the scarcity of annotated data. In such situations, leveraging large language models (LLMs), such as the Generative Pretrained Transformer (GPT) models, offers a promising solution. However, the inherent token limitations of these models can be a bottleneck for processing lengthy legal contracts. This paper presents an unsupervised two-step approach to address these challenges. First, we propose a query-based summarization model that extracts sentences pertinent to predefined queries, concisely representing lengthy contracts. This summarization ensures that the core information remains intact while simultaneously addressing the token limitation issue. Subsequently, the generated summary is fed to GPT-3.5 for precise information extraction. Our approach effectively overcomes the challenges of token limitations and zero resources, enabling efficient and scalable information extraction from legal contracts. We compare our results with those obtained from supervised models that have been finetuned on domain-specific annotated data. Experimental results demonstrate the remarkable effectiveness of our approach, as it achieves state-of-the-art performance without the need for domain-specific training data.
- 7 December 2023
- Published by IOS Press
- p. 177-186
Open AccessInformation Extraction from Lengthy Legal Contracts: Leveraging Query-Based Summarization and GPT-3.5
In the legal domain, extracting information from contracts poses significant challenges, primarily due to the scarcity of annotated data. In such situations, leveraging large language models (LLMs), such as the Generative Pretrained Transformer (GPT) models, offers a promising solution. However, the inherent token limitations of these models can be a bottleneck for processing lengthy legal contracts. This paper presents an unsupervised two-step approach to address these challenges. First, we propose a query-based summarization model that extracts sentences pertinent to predefined queries, concisely representing lengthy contracts. This summarization ensures that the core information remains intact while simultaneously addressing the token limitation issue. Subsequently, the generated summary is fed to GPT-3.5 for precise information extraction. Our approach effectively overcomes the challenges of token limitations and zero resources, enabling efficient and scalable information extraction from legal contracts. We compare our results with those obtained from supervised models that have been finetuned on domain-specific annotated data. Experimental results demonstrate the remarkable effectiveness of our approach, as it achieves state-of-the-art performance without the need for domain-specific training data.
- Open Access
- 19 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 462-466
Improving Translation of Case Descriptions into Logical Fact Formulas using LegalCaseNER
- 19 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 462-466
Open AccessImproving Translation of Case Descriptions into Logical Fact Formulas using LegalCaseNER
- 1 February 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 268-271
Link scientific publications using linked data
Scientific publication management services are changing drastically. On the one hand, researchers demand intelligent search services to discover scientific publications. On the other hand, publishers need to incorporate semantic information to better organize their digital assets and make publications more discoverable. For this purpose, we investigate how to manage scientific publications using Linked Data and introduce FELinker, an entity linking component that links scientific publications with DBPedia. In particular, this paper introduces advantages of linking scientific publications with Linked Data, discusses major challenges, and outlines the proposed method. Experiment shows the proposed method could get promising performance in scientific publication linkage.
- 1 February 2015
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 268-271
Link scientific publications using linked data
Scientific publication management services are changing drastically. On the one hand, researchers demand intelligent search services to discover scientific publications. On the other hand, publishers need to incorporate semantic information to better organize their digital assets and make publications more discoverable. For this purpose, we investigate how to manage scientific publications using Linked Data and introduce FELinker, an entity linking component that links scientific publications with DBPedia. In particular, this paper introduces advantages of linking scientific publications with Linked Data, discusses major challenges, and outlines the proposed method. Experiment shows the proposed method could get promising performance in scientific publication linkage.
- 1 January 2009
- Published by Springer Nature
- p. 689-696
Web Orchestration: Customization and Sharing Tool for Web Information
- 1 January 2009
- Published by Springer Nature
- p. 689-696
Web Orchestration: Customization and Sharing Tool for Web Information
- 1 January 2006
- Published by Springer Nature in Lecture Notes in Computer Science
- p. 1118-1125
A Web-Based Method for Building Company Name Knowledge Base
- 1 January 2006
- Published by Springer Nature in Lecture Notes in Computer Science
- p. 1118-1125
A Web-Based Method for Building Company Name Knowledge Base
- 1 January 2005
- Published by Springer Nature in Lecture Notes in Computer Science
- p. 1004-1016
Web-Based Terminology Translation Mining
- 1 January 2005
- Published by Springer Nature in Lecture Notes in Computer Science
- p. 1004-1016
Web-Based Terminology Translation Mining
- 1 January 2005
- Published by Springer Nature
- p. 542-552
A Lexicon-Constrained Character Model for Chinese Morphological Analysis
- 1 January 2005
- Published by Springer Nature
- p. 542-552
A Lexicon-Constrained Character Model for Chinese Morphological Analysis