利用依存句法分析器探討失語症中的動詞論元結構

// Characterizing Verb Argument Structure in Aphasia Using Dependency Parsers

本研究開發並驗證一種自動化語法依存分析法，用於量化腦中風失語症患者敘事語言中的動詞論元結構（VAS）使用｜結果顯示，布洛卡失語症及無語法症患者使用較少動詞論元，且論元數量顯著低於正常對照組｜此方法能有效分析自然語言中VAS，支持臨床語言評估工具開發，強調失語症語法功能受損的重要性。

// This study aimed to develop and validate an automated method using a dependency parser to quantify verb argument structure (VAS) use in narrative speech from people with poststroke aphasia. The new automated VAS measures were validated against manual annotations and applied to a large aphasia sample, revealing that Broca's aphasia and agrammatic participants use verbs with fewer arguments and produce fewer arguments than controls. The findings confirm that dependency parsers can effectively analyze VAS in spontaneous speech with minimal manual effort. This approach supports the development of clinical tools for complex language assessment and highlights VAS impairments in aphasia.

Published 2026年5月18日