Automatic Analysis and Generation of Japanese Complex Predicates based on the Complete List of Predicate Components

2021 Fiscal Year Final Research Report

• Project/Area Number: 20K21809
• Research Category: Grant-in-Aid for Challenging Research (Exploratory)
• Allocation Type: Multi-year Fund
• Review Section: Medium-sized Section 61:Human informatics and related fields
• Research Institution: Nagoya University
• Principal Investigator: Sato Satoshi 名古屋大学, 工学研究科, 教授 (30205918)
• Project Period (FY): 2020-07-30 – 2022-03-31
• Keywords: 自然言語処理 / 機能語 / 文末述語生成 / 文末述語解析

Outline of Final Research Achievements

Japanese language has very complex predicates such as "Kaka-re-tei-ta-nichigai-arimasen" that are composed by "kaku (write)", "reru (passive), "iru (aspect)", "ta (past)", "nichigainai (must）", and "masu (polite)". This study has compiled the complete list of these components, which are 137 particles, one copula, 133 auxiliary verbs, 74 predicate suffixes, and 301 compound auxiliary verbs. By defining the grammatical properties of conjugation type and connection type for each component, we have realized a system that automatically generates such complicated predicates.

Free Research Field

自然言語処理

Academic Significance and Societal Importance of the Research Achievements

本研究は、機械的な生成の観点から日本語の複雑な述語表現に対する見取り図を与えたという意味において、学術的意義がある。特に、接続型の導入により、活用形を陽に指定しなくとも文法的に正しい述語表現を生成できるようになった。本研究で実現した生成システムを利用することにより、たとえば、書き言葉から話し言葉への自動変換、キャラクタ設定に応じた発話文の自動生成などが可能となる。