A Continuous Rating Model for News Recommendation

With the explosive growth of online news, adaptive news recommendation models based on social connections have developed as a result of the growth of online news websites. Several models have relayed better outcomes for users. Most of these models establish users' social connections according to their binary news rating scores. In various real systems, mainly based on a non-binary pattern, the rating score is often multigrade. In this paper, we introduce a new continuous ratings model for news recommendations. This model uses exponential distance to measure the similarities of users to increase its robustness to noise. Simulation results show that the proposed model has significantly improved the performance of personalized news recommendation.