Improvement of quality of interviews by AI-based summary/analysis for occupational mental health

2021 Fiscal Year Final Research Report

• Project/Area Number: 19K22740
• Research Category: Grant-in-Aid for Challenging Research (Exploratory)
• Allocation Type: Multi-year Fund
• Review Section: Medium-sized Section 58:Society medicine, nursing, and related fields
• Research Institution: The University of Tokyo
• Principal Investigator: Kawakami Norito 東京大学, 大学院医学系研究科(医学部), 教授 (90177650)
• Co-Investigator(Kenkyū-buntansha): 西 大輔 東京大学, 大学院医学系研究科(医学部), 准教授 (40450605) ; 渡辺 和広 北里大学, 医学部, 講師 (60822682) ; 今村 幸太郎 東京大学, 大学院医学系研究科(医学部), 特任講師 (80722793)
• Project Period (FY): 2019-06-28 – 2022-03-31
• Keywords: 人工知能 / 産業保健 / 面接 / 音声

Outline of Final Research Achievements

This study aimed to automatically create a POMR record, assess his/her psychopathology, course and prognosis, based on a voice record of one-on-one interview between a worker (case) and occupational health professionals, with applying artificial intelligence (AI) technologies. It also intended to develop a training system for evaluating the quality of interview of occupational health professionals and making a feedback, to promote he quality of occupational mental health service. However, there was a technical problem in the process of making sentences for recorded interviews, which cannot be solved within a study period, despite of various efforts to overcome it to make the system applicable in the practice.

Free Research Field

精神保健学

Academic Significance and Societal Importance of the Research Achievements

録音データからの文章生成技術の開発というステップで技術的困難に遭遇し、さまざまな工夫によりシステムのプロトタイプを作成し試行を重ねたが、必ずしも実用に耐える水準での精度が研究期間内に達成できなかった。産業保健面談における実用に耐える音声記録の文書生成のためには、文章生成アプリケーションの基礎技術の一層の改善が求められることが明らかとなった。