|
摘要
|
The rapid development of the Internet has transformed the way people access information. Users can now easily search for keywords to obtain diverse data, and recommender systems help manage data overload and find interesting items. Traditional recommender systems focus on individual user preferences, which reflect underlying interests, but in group settings, social factors are crucial for decision-making. Previous research rarely considered these factors. In human psychology, curiosity in social contexts motivates seeking interestingness, emphasizing the unknown and unexpected impact on perception and interest. This study proposes a group recommender system based on social curiosity, inspired by Berlyne’s curiosity theory. The system incorporates social network information and quantifies curiosity-stimulating factors (surprise, uncertainty, and conflict) for modeling. Evaluations using real-world datasets investigate whether curiosity enhances group user satisfaction. Results show that incorporating social curiosity significantly improves the precision, coverage, and diversity of group recommendations compared to collaborative filtering. The findings indicate that curiosity-driven recommendations align better with group interests, offering more diverse options and higher satisfaction among members. |
GoogleAnalytics