Research Of Key Technologies For Chinese Intelligent Search Engine

In this thesis, to improve the intelligence of search engine, The new method, concept retrieval, is studied, and new theory, etwork to Network?is set forward. Some systems are designed and developed successfully following this theory. They are Concept Retrieval Based on Knowledge Base, Knowledge Base Management System to support Concept Retrieval Based on Knowledge Base, Concept Base Management System to make concept retrieval based n-gram more intelligent, Knowledge Mining based on Corpus to support the concept retrieval, Intelligent Retrieval System for Database. Chapter 1 shows the significance and objectives of the thesis. The technologies of Nature Language Processing are introduced in this part, and the methods of study in this thesis are summarized. In the end, it shows the contents of every chapter and gives the frame of the thesis. Concept Retrieval is studied in Chapter 2. First, the theory of 揘etwork to Network? is introduced. Then, the work principle and properties of Concept Retrieval are researched. Finally, we give some the implementation systems of the concept retrieval ---Concept Retrieval Based on Knowledge base, Concept Retrieval Based on N-gram, Knowledge Mining Based on Corpus, Intelligent Retrieval System for Database. In Chapter 3, the construct of concept semantic network and the principle of development of Concept Retrieval based on Knowledge Base are studied. The process of development and the key technologies of this system are showed in detail. Finally, the function of this system is analyzed and summarized. The Knowledge Base Management System is studied in Chapter 4, which is an important component of Concept Retrieval Based on Knowledge Base. In this part, the structure and function of knowledge base are studied. Furthermore, The key technologies of the development of this system are showed in depth. We mainly study Knowledge Mining based on Corpus in Chapter 5. In the process of study, we use two methods of Natural Language Processing---rule based and corpus based. The successful development of this system is not only an important support for two kinds of concept retrieval, but also the great help for automatic abstracting. In Chapter 6, we study Concept Base Management System. First, the knowledge of n- gram and Knowledge Mining Based Corpus are reviewed. Then, we introduce the structures of two kinds of Knowledge Bases used by this system. In the end, the deve1opmeflt of this system is introduced steP by steP.We study the Intelligent Retrieval System for Database in ChaPter 7. First, we studythe principle of this system. In the following, we mainly analyze and design kinds ofKnowledge Bases for this system. Finally, I intrOduce the process of the development ofthis system in detail.ChaPter 8 sums uP the Whole thesis and gives the prosPects for the further study inthis field.