Abstract
Recently, group recommendations have gained much attention. Nevertheless, most approaches consider only one round of recommendations. However, in a real-life scenario, it is expected that the history of previous recommendations is exploited to tailor the recommendations towards meeting the needs of the group members. Such history should include not only which items the system suggested, but also the reaction of the members to these items. This work introduces the problem of sequential group recommendations, by exploiting the concept of satisfaction and disagreement. Satisfaction describes how well the group received the suggested items. Disagreement describes the satisfaction bias among the group members. We utilize these concepts in three new aggregation methods, SDAA, SIAA and Average+, designed to address the specific challenges introduced by sequential group recommendations. We experimentally show the effectiveness of our methods using big real datasets for both stable and ephemeral groups.
Highlights
Recommendations have become a standard in many services that target a consumer
We introduce the notion of sequential group recommendations where the system considers the entire sequence of interactions for a group in order to produce its recommendations
The input to the algorithm is the group G, the iteration j, and the size k of the group recommendation list Grj, which we report to the group after each iteration is finished. n Line 1 we construct a set, Gl, that contains all the items that appear in the group members individuals recommendation lists
Summary
Recommendations have become a standard in many services that target a consumer. From listening to music to health information, recommender systems are employed to make the user experience better and smoother. One of the most widely used methods to produce a user recommendation is the Collaborative Filtering (CF) method. Each friend has his/her preferences for what kind of movie he/she would like to watch. The recommender system needs to address these preferences adequately and report a set of items that have a degree of relevance to each group member. The first is to create a pseudo user by combining each group member’s data and applying a standard recommendation method. The second and most used approach is to apply a recommendation method to each member individually and aggregate the separate lists into one for the group
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
