Novel and Diverse Recommendations by Leveraging Linear Models with User and Item Embeddings

Nowadays, item recommendation is an increasing concern for many companies. Users tend to be more reactive than proactive for solving information needs. Recommendation accuracy became the most studied aspect of the quality of the suggestions. However, novel and diverse suggestions also contribute to user satisfaction. Unfortunately, it is common to harm those two aspects when optimizing recommendation accuracy. In this paper, we present EER, a linear model for the top-N recommendation task, which takes advantage of user and item embeddings for improving novelty and diversity without harming accuracy.