Research And Implementation On Natural Chinese Language Generic Interface To Database

Natural Language Interface to Database (NLIDB) is a branch of artificialintelligence,whichisto makecomputer'understand'naturallanguage,suchas Chinese. And thus,users can access information stored in database byarticulating questions directlyin their ownlanguage. With development of 40years,some of NLIDB demonstrated impressive characteristics in certainapplication areas. However,to be addressed sentences with implicitknowledge, few of performance of the systems were adopted by industry.AftersomecomprehensivestudyofDescriptionLogic,asemanticmodelwasrealized to bridge natural Chinese language and Database. The emphasis wasput on implicit knowledge understanding and the domain of the system wasextended. In order to transplant the system, a domain knowledge extractorbased on inference engine of Description Logic was also designed. The mainwork of this thesis was on knowledge representation and the content of theresearch was from the sentences with lexical analysis to SQL,including themodule of semantic model with Description Logic, the module of implicitknowledge understanding and the module of Domain knowledge Extracting.The system is validated by the data of personnel database in our university,and the result indicates sentences with implicit knowledge can be produced well. Finally by analyzing an example of a domain demonstrated how thesystem semi-auto extractsnewknowledge.