Incremental Chinese Lexicon Extraction with Minimal Resources on a Domain-Specific Corpus

This article presents an original lexical unit extraction system for Chinese. The method is based on an incremental process driven by an association score featuring a minimal resources statistically aided linguistic approach. We also introduce a linguistics-based lexical unit definition and use it to describe an evaluation protocol dedicated to the task. The experimental results on a domain specific corpus show that the method performs better than other approaches. The extraction results, evaluated on a random sample of the working corpus, show a recall of 68.4 % and precision of 37.1 %.

• Type: Journal
• Year: 2010
• Where: COLING
• Authors: Gaël Patin
• Keywords: COLING 2010, Computational Linguistics, Lexical Unit, Lexical Unit Extraction, Linguistics-based Lexical Unit