A matching recommendation algorithm for celebrity endorsement on social network

In this paper, a new matching recommendation algorithm is proposed to help an enterprise to find one or several proper celebrities as their product endorsement. The fans group of a celebrity, his impaction value in the social network and the matching degree between the celebrity and the product are used to determine the most suitable celebrity in social network. The attribute similarities between the target customers and the fans of the celebrity are calculated via the Pearson similarity formula. Then, considering the impaction value of the celebrity and the matching degree of the celebrity and the product which can be accessed on the website or usually available from the enterprise, an evaluation index is proposed. We use some data from Sina Micro-blog to show the effectiveness of our proposed matching recommendation algorithm. Moreover, the analysis shows that the products may be different which are endorsed by the same celebrity in and off the social network.