Abstract:
Accurately predicting a user’s rating to a service is a challenging task in the presence of malicious users that manipulate the ratings to services. Many existing service rating systems lack the ability that counter the manipulation of rating systems. This paper proposed an artificial neural network (ANN) based service rating scheme that counters the manipulation of service ratings. The scheme takes into account of both similarity-based rating and the ratings given by representative users when predicting a user’s rating to a service. Some experiments were carried out to compare the prediction accuracy of the proposed scheme with a well-known existing scheme WSRec [26]. The results show that the proposed scheme provides more accurate rating predictions in the presence of a large amount of malicious users.