Research on intelligent link prediction model of friend influence based on big data and complex network

Abstract

In the brand community, we can expand the brand influence of the product by recommending users with high influence overlapping nodes of two or more product circles to become friends. In this paper, by considering the social network structure of target customers, using the naive bayes link based on local similarity index prediction model algorithm could become friends of a certain high influence overlapping nodes users recommend to another type of high-impact overlapping nodes, lead product circle ordinary users produce purchasing motivation, to produce to product purchasing behavior, for marketers to target customers personalized marketing provides a new theory and new method, using the influence of the friend to realize personalized marketing. Through the recommendation of friends in the social network brand community, the interaction of users in the community can be strengthened, and the vitality of the whole network and brand marketing efficiency can be improved.