Research On Chinese Personalized Retrieval System Based On User Model

With the rapid development of the information society, people access to the society of a high-speed transmission of information from the information occlusion. This effectively transmitted information comes from network, which makes the world smaller by obtaining instantaneous information. Now a highly effective and convenient digital library is to replace the library from which people access mainly to information in the past. However, the development of digital library mainly concentrates on collecting and arranging digital information. The users take a lot of time when they retrieve information from the network. Therefore, the digital library doesn't improve much from users'perspective.In fact, at a time each user is only interested in a certain aspect of information, which is difficult to retrieve. So, the retrieval model is to improve and the users'different retrieval needs is to meet. Thus, the digital library can get better development.Personalized digital resources retrieval system which is shortened for PDRRS hereafter is developed, based on the users'demand purpose and features by analyzing users'interests and hobbies. In this context, this thesis adopts users'modeling technology, web data mining technology and collaborative filtering technology on the basis of the existing personalized information retrieval system and establishes a digital library with an individualized Chinese text information retrieval system according to the characteristics and complexity of Chinese text information resources in the digital library. Therefore, the users can get the most information with the minimum operation.This paper is divided into five parts.The first part mainly introduces the present situation and existing problems of the research on the digital library with personalized information retrieval system and puts forward the corresponding solutions to the problems. Meanwhile, a series of background knowledge is studied and analyzed, which is needed when the personalized digital resources retrieval system is developed.The second part analyzes the characteristics of the PDRRS and puts forward the goals the retrieval system can accomplish around these characteristics.The third and the fourth part is the key in this paper.The third part builds a systematic core working model and explains the working principle of each corresponding function module according to the analysis of the former part.In the expression of the document features, the PDRRS divides the document into phrases or entries by using Chinese text information processing technology. That is, each document is expressed as characteristic vector with entries and as semantic conversion with a mathematical calculation model.The PDRRS builds the users modeling by completing the information acquisition of users'needs from three aspects. First, the users provide the retrieval information on their own. Second, the PDRRS analyzes the users'needs and abstracts their retrieval needs out of the key words they provide. Finally, the PDRRS analyzes and obtains the users'retrieval habits and their changing process by tracking the users'querying retrieval process.The fourth part is about the realization of the PDRRS. The PDRRS expands the users'queries through users'modeling, provides the relevant information according to users'habits and shields the irrelevant information so as to improve the efficiency of the query. In this way, different users with different interests and hobbies can get different retrieval results, though they enter the same retrieval words.The fifth part works out a summary of the current work and proposes a further suggestion for improvement.