With Natural Language Understanding And Information Mining Capabilities, Search Engine Development

The background of this thesis is to develop an enterprise search engine which can either be used for enterprise portal to provide online self-service, information searching or can be used as an internal knowledge management system for a large organization.The author participated this project to realize an intelligence search engine with nature language process and data mining capability.This search engine is based on traditional search engine architecture, and has lots of enhancement and customization. The information retrieval has been improved from keyword level to concept level. It uses ontology to analysis syntactic, semantic and pragmatic information of massive data.This paper tries to handle different kinds of documentation with various formats, using semantic engine to parse content and build index, with question-answer based query method, it can provides personalized searching service.In general, this paper hasn't made any breakout in nature language process and data mining area. It is just an implementation with different technologies. It still needs a lot of improvement on speed, efficiency and query accuracy.