LARGE: A leadership perception framework for group recommendation

Group recommendation systems are crucial in applications such as group travel or collective entertainment. Conventional approaches have achieved notable success by matching the aggregated preferences of all group members, under the assumption that the preferences of each member can influence group decisions. Challenging this paradigm, our paper introduces a transformative hypothesis centered on the concept of the leadership group, where the dominant preferences of a leader significantly influence group recommendations, potentially diminishing the contributions of other members. To bridge this innovative hypothesis with practical applications, our research unfolds in two main phases. First, we use real-world datasets to empirically investigate the presence and influence of leadership groups, providing compelling evidence to support our hypothesis. Following that, informed by these insights, we develop advanced methodologies for identifying leadership groups and integrate this concept into a novel recommendation framework. This framework is meticulously designed to capture the essence of group decision- making dynamics, therefore enhancing recommendation accuracy and user satisfaction. Its compatibility with various underlying models ensures broad applicability and potential for widespread adoption. Comprehensive experiments, benchmarked against state-of-the-art group recommendation systems, demonstrate that our innovative framework significantly improves accuracy, underscoring the importance of accounting for leadership dynamics in group recommendations.