Research on analyzing Mongolian legal documents using deep learning
| Project/Area Number |
21K12600
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 90020:Library and information science, humanistic and social informatics-related
|
|---|
| Research Institution | Ritsumeikan University |
|---|
Principal Investigator |
美龍 硬幸 (バトジャルガル ビルゲサイハン) 立命館大学, 総合科学技術研究機構, プロジェクト研究員 (30725396)
|
|---|
| Project Period (FY) |
2021-04-01 – 2025-03-31
|
| Project Status |
Granted (Fiscal Year 2023)
|
| Budget Amount *help |
¥4,160,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥960,000)
Fiscal Year 2024: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2023: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2022: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2021: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
|
|---|
| Keywords | deep learning / legal documents / text classification / Mongolian / deep learming / machine learning / text mining |
|---|
| Outline of Research at the Start |
Recently, numerous Mongolian legal documents have been made digitally public. Nevertheless, analysis of these documents has not been done due to the lack of Mongolian Natural Language Processing (NLP) tools. A reliable computerized analysis is necessary, which requires developing an innovative method for analyzing these documents. Manual reading and analyzing documents are not effective, on a massive scale. Demands are growing for precise analysis of legal documents, hence an artificial intelligence (AI) system that aids expert decisions that clarify conflicts and disputes is essential.
|
| Outline of Annual Research Achievements |
In the FY2023, the following tasks have been performed: 1) collecting and preparing datasets, 2) proposing deep learning models for Mongolian legal documents and, 3) enhancing the web-based system.
Regarding the datasets, a natural language interface (NLI) dataset was prepared manually by selecting 829 questions related to Mongolian civil law from the Mongolian bar exam questions. We also proposed deep-learning-based models for recognizing textual inference in the Mongolian language and experiments were conducted.
Moreover, in the FY2023, based on the research results obtained in the past, we enhanced our web-based system to integrate for recognizing textual inference in the Mongolian legal documents.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
Our research has been conducted according to the research plan. As planned, the achievements in the FY2023 allowed advancing my research towards to the research goal in developing a deep-learning-based method to analyze Mongolian legal documents. Surveys were conducted in Mongolia, as well as feedback and data were obtained through business trips. Research results and achievements have been published in two international journal articles.
|
| Strategy for Future Research Activity |
In the FY2024, business trips for 1) conducting evaluations among overseas users, and 2) obtaining analyses and feedbacks will also be conducted.
We are also planning to perform evaluations by several experts at the National University of Mongolia and Ritsumeikan University in Japan. Feedback from the researchers will be received in a timely manner. Users’ evaluation will also be conducted by experts and users. Further improvements of the method will be carried based on the evaluation results and user feedback.
Research assistance of experts and students are necessary on a part-time basis to 1) evaluate the proposed system 2) conduct experiments, and 3) analyze the experimental results.
Feedback from the researchers will be received in a timely manner. Further improvements of the system will be carried out based on the evaluation results and user feedbacks.
Methods and contents that have no copyright issues will be available freely to the public on the Internet during the FY2023-FY2024.
|
Report
(3 results)
Research Products
(13 results)
