Research And Realization On The Personalized Search Engine In The Field Of Health Consultation

To a certain extent, the traditional search engines have solved the problems of"information overload" and "resource labyrinth". However, the united search results cannot meet the growing personalized demand of users. Therefore, the personalized searchengine has become the research focus in the modern information retrieval. Aimed to thecharacteristics of the users’ strong demand in personalization and the complexinformation in health consultation, this thesis constructs a personalized search enginebased on the user interest model.Firstly, this thesis analyses the system architecture of personalized search engine,and expatiates on the core technologies of constructing the user interest model and thealgorithm of personal ranking. Then, after studying the characteristics of the user’sbackground and the browsing behavior in the health consultation, this thesis collects theinterest information of user through combining the methods of explicit and implicitinformation gathering. After analyzing and dealing with the different information,Vector Space Model is used to build user interest model. And the factor of weight decayis introduced to update the user interest model. In order to realize the personal rankingof the search results, the user interest model is applied to the search engine’s rankingalgorithm. Finally, this thesis tests the system’s performance by calculating the user’sdegree of satisfaction. Experimental results show that the system improves the user’sdegree of satisfaction with the search results effectively. And to a certain extent, it helpsthe user to lock the required information resources quickly and accurately. And thissystem has some value of application in the field of health consultation.