Research On Technologies Of Personalized Information Service

Currently, personalized information service is a very important research issue in the field of Web information technology. Aiming to the user's different demands, it adopt different mode and provide different service to assist user in gaining information rapidly duly and effectively. Aiming to the characteristic and demands of Web information service, the dissertation proposes the system frame of personalized information service. The Web text resource modeling, organizing, managing and user demand modeling are studied comprehensively. These technologies are employed to realize the web information aspects such as search & navigation of resources, personalized information customization, knowledge active recommendation and so on. The main work of this dissertation is dissertated in the followings:In the 1^st chapter, the research background and practical significance of personalized information service are introduced firstly. Aiming to the research content of Web information service, the current advances of the interrelated fields, such as search engine, semantic Web, personalized information service and so on, are also reviewed. The service characteristic and function of three types of personalized information service system are compared specially. And then the research content and architecture of the dissertation are presented.In the 2^nd chapter, the general situation of personalized information service is addressed. The definition, logic process, characteristic and service mode is researched.Then, the system characteristic and function component of the personalized information service system researched by the dissertation is brought forward.In the 3^rd chapter, the formalized description (feature modeling) method is discussed chiefly. A feature modeling method based on vector space model and HowNet is presented. The material realization technologies, such as weighting method of the characteristic element, eliminating different semantic meanings, semantic similarity calculating, are researched. And to achieve the maximizing avail of Web resource, the classification of Web text resource is studied. At last, through experiments, the dissertation carries on verification of the above-mentioned theories and methods.In the 4^th chapter, the dissertation discusses the user-oriented user modeling technology. Three sub-missions of user modeling, covering data gathering, model describing and updating, are researched. The key technologies are studied, including using Centroid-based Classification Method to build user model, optimizing user model based on HowNet and employing Rocchio feedback algorithm to update user model. Then the information filtering technology is studied. Finally the user modeling technology given by the dissertation is validating through experiments.In the 5^th chapter, the dissertation takes Web resource discovery and sharing as the application background, and introduces the examples about integration application of personalized information service in manufacturing interrelated field, which includes search & navigation of resources, personalized information customization and knowledge active recommendation. The theories and methods studied by the dissertation are validated basically.The last chapter summarizes the achievements and innovations of the dissertation, and prospects the future work.