Design And Tool Implementation Of User-Attention Behavioral Value Evaluation Model

With the rapid development of the Internet social industry,most websites owner implement analyzes toward user behavior:commercial websites analyze user's purchase behavior to determine the user's business value,news and music websites analyze user's click behavior and recommend related content for their users.However,the analysis on the website side covers the whole community,thus cannot customize for single user.As the user's preference list increases,the environment is increasingly inflated and user may find it difficult to obtain valuable content from it.Thus,this paper proposes a user-attention behavioral value assessment model for the lack of custom analysis of Internet users' attention environment.Based on Commercial user value model(RFM),we summarize and select the corresponding indicators from the three aspects of value,fr-equency and proximity,then collecting and analyzing the data related to attention user group in order to build the user-attention behavioral value evaluation model.This paper also propose an assess method based on rating mechanism and dynamic score correction system for all user-related features in certain time domain,which reduces the impact of excessive feature counts on the performance of the evaluation model.Besides,this paper proposes a Fuzzy-C-Means(FCM)initial sensitivity optimization algorithm to solve initial sensitivity problem,which is based on particle swarm optimization.In this algorithm,we integrate the Levi flight formula with standard PSO algorithm and eventually applied in FCM algorithm by introducing global random walk mechanism to enhance particle activities and controlling the balance of local walking and global random walking with a switching parameter.This paper conducts clustering tests and validity analysis on several UCI standard datasets to testify its accuracy and fitness performance.Experimental results show that compared with the FCM algorithm,the PSO-LF-FCM algorithm enhances the clustering accuracy and the global search performance in the later iteration of the algorithm,which implies its superior global convergence and optimal solution search ability.Based on research of algorithm and the of user-attention behavioral value evaluation model the user-attention behavioral value evaluation tool is designed and implemented.The tool provides users with three primary modules:user data crawling,user value assessment,and user-attention environment information feedback.These modules provide website users with a number of features from data collecting to evaluate result presentation.The result from tests on the real data from pixiv.net website shows that all modules of the tool can operate in stability.