A Decision Making Method Based on TOPSIS and Considering the Social Relationship

With the rapid development of social networks, the decision-making of users is deeply influenced by online ratings, and suggestions from friends and relatives, and they have become important references for consumers to make decisions. However, the multi-sources information makes it difficult for users to make efficient decisions. In order to overcome this difficulty, our research lays a basis on helping consumers to make decisions. We consider the impacts from both friends in social networks such as Facebook and online buyers of online shopping websites such as Amazon. Furthermore, for generating the impact weights of friends, we utilize the Hawks process to simulate the interaction process with three typical interactive behaviors. For making a higher accuracy of product ratings, we filter online buyers and adopt the ratings from the buyers who had similar interests with a target individual. At last according to friends' suggestions, online ratings, and the weights generated formerly, we provide the option preference by the TOPSIS (Technique for order preference by similarity to an ideal solution) method. The training results analysis of the parameters involved in our research is provided as well as the decision making results evaluation. Compared with the state-of -art methods, our method can provide more accurate results. The architecture and algorithms we provided also can be applied to other kinds of social network based decision-making processes.