Research And Application Of Full-Text Retrieval Based On Lucene.Net

The rapid development of IT and the popularity of the Internet increased digital resources such as digital documents and information database apace, showing the characteristic of abundance. How to search information users interested in quickly from the resources has been a hotspot of information research. Full-text retrieval can search the huge amounts of data efficiently. Words can be scanned, indexed, pointed out its position and frequency by an indexing program. A retrieval program searched the index file and returned results to users when they searching the data. This mechanism improved searching speed and efficiency greatly.With the deepening of information system applications, data stored and managed are more and more. Traditional approach based on database fields has failed to meet the needs of the query. Users need to search resources of information systems quickly through full-text retrieval mechanism. So how to combine full-text retrieval with information systems has been an important problem of exploitation.Lucene.Net is a high-powered open-source Full-text retrieval toolkit. With a very obvious object-oriented feature, it provides a simple but powerful interface. Developers can insert it to application programs conveniently to construct a new Full-text retrieval system. This paper researched Lucene.Net deeply and developed different full-text retrieval systems according to different application requirements. The main researches include the following three points:1,Analyzed each of the modules of Lucnen.Net and extended its function. Improved the Chinese segmentation mechanism of the Analyzer and optimized the efficiency of indexing and the order of searching results.2,Nowadays safety of information systems are emphasized more and more. Important data should be encrypted and stored in database. But traditional systems can not provide full-text retrieval for encrypted data. According to the shortcoming of current solutions, this paper built a full-text retrieval engine above the data access layer based on Lucene.Net to indexing data when encrypted and stored in database, and to achieving association between index and encrypted data. Full-text retrieval was implemented effectively while the data security was achieved, and users'need was fulfilled.3,In practice of information system two kinds of data were installed on different computer servers: stored based on file system or database. Each data source was highly autonomic. A full-text retrieval system with unified mode for querying data from varies of data source was needed. A distributed full-text retrieval system was implemented in this paper based on Lucene.Net and SOA architecture. Unified format index can be created that users can query shared information effectively stored on different servers. Data querying requests from users can be distributed so that local data of each data source can be retrieved. The querying results were gathered together and transferred to users. It can meet the new demand of full-text retrieval.