Proposal for A Framework for the High-Precision Identification ofLinguistic Relationships

Abstract

Current research in Information Retrieval and Information Extraction demands high-precision syntactic and semantic information from natural language text. We propose a plan for developing a framework to identify, with high-precision, the linguistic relationships between pairs of words in natural language text. Related research is reviewed and preliminary results are given. In our plan we outline the linguistic identification task: a new way to evaluate NLP systems.