An Optimized trust model integrated with linear features for cyber-enabled recommendation services

The growth of cyberspace brings more information to service recommendation. The scores of item are used in most of the recommendation algorithms, but the attributes of users and items are rarely involved in trust recommendation in cyberspace. Both the rating features and attribute information are important for trust recommendation results. In this paper, we combine the heterogeneous information in cyberspace and propose a novel trust recommendation model based on the latent factor model and trusty neighborhood fitting model. We utilize the feature based Latent Factor Model and study the linear features integrated model To solve the failure problem of the latent factor model in the integrated model under the cold-start situations, we propose two optimized methods, which contain the filling method based on feature similarity and the filling method based on feature regression through mapping attributes to features. Experimental results show that the improved method outperforms traditional collaborative recommendation in terms of recommendation accuracy. Meanwhile, our proposed method has been verified to free from the impact of the cold-start problem.