Context based multi-source feature aware event recommendation algorithm

In the research, EBSN event recommendation algorithm is usually based on the user's simple relationship or social relationship. However, there are many factors that affect users' choice in EBSN, such as group based social information in EBSN context information and event text information, location information, time information, tag information. In order to enhance the effect of recommendation, fundamentally solve the problem of data sparsity and cold start of recommendation, this paper proposes an event recommendation algorithm based on context multi-source feature perception, establishes the scoring model of multi-source features, applies Random Forest algorithm to realize the weight calculation and extraction of multi-source features, and finally generates the event recommendation list. The experimental results are compared with other three ranking based recommendation algorithms. In the standard of NDCG@10, the MFA-RF algorithm proposed in this paper has obvious advantages, and solves the problems of data sparsity and cold start in the recommendation.