This paper presents the results of a feasibility study for a bootstrapping natural language database query interface which uses natural language generation (NLG) technology to address the interpretation problem faced by existing NLIDB systems. In particular we assess the feasibility of automatically acquiring the requisite semantic and linguistic resources for the NLG component using the database metadata and data content, a domain-specific ontology and a corpus of associated text documents, such as end-user manuals, for example.
展开▼
