THEMATIC ROLE STRUCTURES: BRIDGING FRAMENET AND NATURAL LANGUAGES FOR ENHANCED LINGUISTIC ANALYSIS

Аннотация

FrameNet is a valuable lexical resource that captures the meaning and structure of words in terms of frames and their associated lexical units. However, its applicability to diverse natural languages is often hindered by language-specific variations in thematic role structures. This research proposes a novel approach to link FrameNet with multiple natural languages by establishing universal thematic role structures. By aligning thematic roles across languages, this study aims to enhance linguistic analysis and facilitate cross-lingual information retrieval, machine translation, and sentiment analysis. The proposed method leverages linguistic typology and cross-lingual learning techniques to create a unified framework for integrating FrameNet with various languages, promoting a deeper understanding of lexical semantics and facilitating language technology applications.