Personalized Information Service Applications Based On Click-stream Technology

In the vast world of Internet information, it is increasingly difficultto people to efficiently obtain the required information. The reasons are: onthe one hand, there is the conflict between the decentralized nature ofinformation resources and the limited and dynamics nature of users'information needs. On the other hand, the current search engines and informationservices tools cannot satisfy the information users' need, because of thedefault of their inaccurate, unreal-time. The solution is to seek a betterquality information service. With rapid developments in information technology,according to each user's interest to provide personalized information serviceshave become one of the development trends, which is the best way to solve thecurrent contradiction between demand and need. But through studying the currentpersonalized information services tools, the paper discovered that the currentpersonalized information services tools have the shortcomings that cannotprovide accurate and across-the-aboard information, and cannot track thedynamic information interest changing of users, and need to customize and feedback frequently by users. In this these issues, this paper introducedclickstream technology, presented a new personalized information servicesscheme—the personalized information services based on the clickstreamtechnology. The ideas is making use of clickstream record left behind when userbrowsed information on website page to analyze users' information interest,to get the information interest type assembly ,then recommend information touser according to the information type assembly. The personalized informationservices scheme is characterized by timely access to current users'information interest, recommending accurate and comprehensive information tousers without users' participation.By in-depth studying services scheme and system structure of personalizedinformation services based on clickstream, the paper proposed a generalimplementation process, which is: at first, establish clickstream datawarehouses to collect clickstream data, in this stem focus on the pretreatmentof web log files that is the main source of clickstream data and establishmentof clickstream data warehouses of Theme-oriented analysis;secondly, get theinformation interest characteristic assembly by connecting web usage miningand web content mining;finally, recommend information to users on the basisof the information interest characteristic assembly. In order to enhance theavailability and promotion of the scheme, design the practical applicationscheme in jinyao website which is the portal network of pharmaceuticals industry.In the drug network based on the clicking of the flow of personalized informationservices technology applied research, this paper focuses on the design of thepretreatment of clickstream data and more the interest analysis of users accessto information, and established the clickstream data marts on the basis of theanalysis theme.