Users interest discovery can be belonged to category of classification knowledge discovery, which can divide user information need into two groups as interested group and uninterested group by classification forecast analysis of user information need which reflects historical visiting information behavior of user. After analysis the use of a several methods of user classification prediction, this paper presents a modified model combining clustering algorithm was presented for improving the forecasting accuracy of SVM. The efficiency of the proposed method was tested by the user access informations log data. The results have shown that the higher accuracy is expressed in this proposed model, and it is applicable to practice.

Under the innovation system integrated with production, teaching and researching, the evolution mechanism of knowledge theme becomes central point in academic fields. However, previous researches mostly focused on analyzing the tendency of academic achievements, didn’t bring the potential interaction between the production and researching into consideration. In this paper we will use the co-integration analysis and Granger causality test methods, which are prevalent in economics research, to study the evolution of specific knowledge theme – “Cloud computing”. Our main purpose is to reveal the evolution mechanism as well as general principle of popularity of one theme.

No abstract