Evaluating Large-Scale Construction Grammars on the Tasks of Semantic Frame Extraction and Semantic Role Labeling

Abstract

This study investigates the application of computational construction grammars to semantic frame extraction (SFE) and semantic role labeling (SRL) in natural language processing (NLP). SRL and SFE discern relationships within sentence constituents to extract meaning. The study evaluates various computational construction grammars within the [project and tool name anonymized], a tool which employs Fluid Construction Grammar (FCG) to detect frame-semantic patterns in text corpora. The analysis assesses the performance of these grammars, highlighting both their strengths and weaknesses, to guide future advancements. Results reveal significant differences in grammar performance based on configurations and heuristics, and the advantage of clustering PropBank role sets into semantically similar groups using VerbAtlas mappings for a comprehensive understanding of grammar performance across semantic categories. The findings contribute to the further development and operationalization of large-scale, computational construction grammars, providing insights into the strengths and areas for improvement of current grammars, and suggesting directions for future research.