An Ontology-Based Architecture for Natural Language Access to Relational Databases

摘要

Natural language (NL) access to databases is a problem that has interested researchers for many years. We demonstrate that an ontology-based approach is technically feasible to handle some of the challenges facing NL query processing for database access. This paper presents the architecture, algorithms and results from the prototype thereof which indicate a domain and language independent architecture with high precision and recall rates. Studies are conducted for each of English and Swahili queries, both for same language and cross-lingual retrieval, from which we demonstrate promising precision and recall rates, language and domain independence, and that for language pairs it is sufficient to incorporate a machine translation system at the gazetteer level.