The Research On The Influential Factors Of The Social-network Based Referral

In recent years, recommender systems become increasingly popular for eliminating the information overload and providing users with personalized experiences. For the purpose of boosting the users’ acceptance rate of recommendation as well as their satisfaction, the researchers have tried every means. Some of them focus on improving the classic recommendation algorithms while others incorporate users’ friendships in recommender systems due to the development of the social network.Although friends’ referral has numerous advantages, the usage of this method seems less effective than the direct recommendation. The reason is that some referrers are not willing to deliver the recommendation items to their friends which lead to the loss of information transfer.As a matter of fact, there are many factors that may influence users’ recommendation willingness. This paper considers the factors of the referrers themselves, the recommendation contents as well as recommendation methods, and the social relationship strength between the referrers and the target users as a whole. Via incorporating the popular social network, we then construct a hypotheses model to explain users’ willingness to deliver the recommendations in social network environment.According to the in-depth user study and the calculation of the social network data, we analyze the correlation between the influence factors and the users’ recommendation willingness. Based on these results, we firstly modify our original model, and then design a control experiment to evaluate the modified model. Thus, we could locate the appropriate referrers as well as the target users, and choose the suitable contents and methods so as to improve the social recommendation willingness and efficiency.