A novel user behavior analysis and prediction algorithm based on mobile social environment

For service behavior prediction, if merely depending on behavior history of a target user, the quantity and category of samples are quite limited; if utilizing correlated users' samples by means of result fusion, the predicted results are very likely to be interfered by noise samples. Therefore, based on the mobile social environment (MSE) of a target user in this paper, the correlated user with the closest long-term habits and that with the greatest short-term influences for the target user are respectively obtained by using optimization theory. Their behavior samples are integrated into the sample database of target user to construct a sample enriched mechanism with minimized noise for improving the accuracy of user behavior prediction remarkably. First, according to the characteristics of MSE, two optimization models based on similarity degree and interaction degree respectively are formulated to select the corresponding optimal correlated users for analyzing two main factors of the target user's behaviors; furthermore, an adaptive update strategy based on fuzzy theory is proposed to describe the importance of two factors in real time and quantitative manners. Second, an improved Apriori theory is introduced to predict user next service behaviors accurately; particularly, a new update mechanism of Apriori sample database is constructed to effectively integrate the samples of optimal correlated users. Finally, extensive simulation results show that the proposed algorithm outperforms several related algorithms in terms of prediction accuracy and operation efficiency.