A Knowledge-aware Multi-tasks-based Disease Network Construction on Biomedical Literature

• Project/Area Number: 24K15097
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 61030:Intelligent informatics-related
• Research Institution: National Institute of Advanced Industrial Science and Technology
• Principal Investigator: Sohrab Mohammad 国立研究開発法人産業技術総合研究所, 情報・人間工学領域, 主任研究員 (50788247)
• Project Period (FY): 2024-04-01 – 2027-03-31
• Project Status: Granted (Fiscal Year 2024)
• Budget Amount: ¥4,680,000 (Direct Cost: ¥3,600,000、Indirect Cost: ¥1,080,000); Fiscal Year 2026: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000); Fiscal Year 2025: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000); Fiscal Year 2024: ¥1,950,000 (Direct Cost: ¥1,500,000、Indirect Cost: ¥450,000)
• Keywords: NLP / Information Extraction / Langauge Model / Knowledge Integration / Disease Network

Outline of Research at the Start

We will introduce a knowledge-aware neural disease network (KANDN) that will extract diseases, proteins, genes, compounds that interact and regulate each other as events in scientific articles. We will construct a knowledge-aware transformer-based neural pre-trained language model that represents the context-aware input vector by integrating knowledge. In this implementation, several NLP techniques including neural NER, EL, RE, and EE will be addressed as an end-to-end joint learning approach to generate more precise and accurate KANDN from the biomedical domain to facilitate the end users.