Abstract:Recommendation techniques greatly promote the development of online service in the interconnection environment. Personalized recommendation has attracted researchers’ special attention because it is more targeted to individual tasks with the characteristics of diversification and novelty. However, the data sets that personalized recommendation process usually possess the characteristics of data sparseness and information loss, which is more likely to have problems such as cognitive deviation and interest drift. To solve these issues, in recent years people gradually notice the important role in which trust factor plays in promoting the development of personalized recommendation. Given the difference between online social trust and traditional offline social trust in facilitating personalized recommendation, this paper proposes a novel technique of online social trust reinforced personal recommendation to improve the recommendation performance. Compared with traditional offline social trust-based personal recommendation, our work synthesizes both factors of online social trust and offline social trust to identify private and public trusted user communities. The trusted degree or the accredited degree can be deduced by Bayesian network probabilistic inferences. In this way, the performance of personalized recommendation can be improved by avoiding excessive interest deviation. Moreover, we also get time sequence into personal recommendation model to effectively track how user’s interest changes over time. Accordingly, the recommendation accuracy can be improved by eliminating the unfavorable effect of interest drift caused by temporal variation. Empirical experiments on typical Yelp testing data set illustrate the effectiveness of the proposed recommendation technique.

Preference of Online Users and Personalized Recommendations

User Preference Quantity Versus Recommendation Performance: A Preliminary Study

User Interest Dynamics on Personalized Recommendation

Heterogeneity Involved Network-based Algorithm Leads to Accurate and Personalized Recommendations

Hybrid Recommendation Model Based on Social Media and Social Networks

Personalized recommendation system based on social relationships and historical behaviors

A Personalized Hybrid Recommendation Strategy Based on User Behaviors and Its Application

A Personalized Recommendation Algorithm Via Biased Random Walk

Research on Personalized Hybrid Recommendation System

Alleviating Bias Leads to Accurate and Personalized Recommendation

A Joint Optimization Approach for Personalized Recommendation Diversification.

Empirical Research on the Impact of Personalized Recommendation Diversity.

Promotional effect on cold start problem and diversity in a data characteristic based recommendation method

Can Dissimilar Users Contribute to Accuracy and Diversity of Personalized Recommendation?

A Time-aware Hybrid Algorithm for Online Recommendation Services

Online Social Trust Reinforced Personalized Recommendation

Personalized Recommendation by Exploring Social Users' Behaviors

Combining user interest and feedback in an online book personalized recommendation system

Personalized Recommendation Algorithm Based on Preference Features

A joint optimization approach for personalized recommendation diversification

Personalized Recommender Systems Based on Social Relationships and Historical Behaviors