Information Personalized Service Based On The Network

With the computer technology and network technology, Internet on the amount of information in order to exponentially grow and expand rapidly, but their information organization is heterogeneous, pluralistic and distributed. On the one hand people are useful to the urgency of the growing information needs; the other hand, people face due to constantly updated information and increase the formation of the "information explosion", how to derive the desired knowledge of their own, but could do nothing. For ordinary users, Internet on the "information Trek" and "information overload" has become an increasingly serious problem.This information personalized service from start, the set of personalized service status information, content based on personalized service for information related to theory, using the key technology and personalized service evaluation studies, and two The new algorithm into the personalized service, "cold start" problem analysis and solution of. Current Web-based personalized information services used by a variety of recommended technologies, collaborative filtering-based recommendation technology is the most widely used and most successful. The technologies such as content-based filtering for recommendation technology, based on knowledge of the recommended techniques, etc. in technology to achieve the degree of difficulty, the algorithm's time complexity, space complexity, etc all have great advantages, but it also has its own shortcomings, such as: accuracy problem, sparsity problem, "cold start" problem, scalability issues have been plaguing the further development of collaborative filtering technology. The "cold start" problem is in the extreme case of sparse data, a new project the first time, it made no assessment of the user, or a new user does not any evaluation of a project done, then the user-based collaborative filtering algorithms simply can not. "Sparse problems" and "cold start problem" directly affect the personalized information service system, the normal operation of recommended paper on the "cold start problem" in-depth research, innovation is for the "new user" and "new projects" The "cold start problem" proposed a new solution algorithm, we design a personalized service evaluation method, using two data sets on the authority of the new algorithm was tested, and conclusions are given according to test results.