Towards Better Representation of Context Into Recommender Systems
Jinfeng Zhong, Elsa Negre · International Journal of Knowledge-Based Organizations · 2022
Context-aware recommender systems (CARSs) are attracting more and more attention from both the academic community and from industry. Users' contextual situations (e.g., location, time, companion, etc.) which can influence their ratings on items, are taken into consideration. Therefore, more accurate and personalized recommendations can be generated. The integration of contextual information in recommender systems to better model users' preferences under different contextual situations is a key research topic. In this paper, the authors propose a new method for representing contextual situations in recommender systems based on the influence of contextual conditions on ratings using Pearson Correlation Coefficient. The authors show the effectiveness of the proposed method compared to state-of-art methods by experiments on three different datasets widely used in CARSs research community.