As important tools for information filtering, recommendation systems have greatly improved the efficiency of users' access to information in daily life by providing personalized suggestions. However, as people's reliance on it grows, recent studies have gradually revealed their potential risks of social unfairness, such as gender discrimination that may result from job recommendations. The unfairness not only harms the interests of specific individuals or groups but also threatens the credibility and long-term sustainability of systems. Therefore, building fairness-aware recommendation systems that proactively identify and mitigate unfairness is crucial for achieving responsible recommendation services. This study focuses on systematically evaluating the effectiveness of current fairness intervention strategies. Specifically, pre-processing methods (including data relabeling and resampling) and post-processing methods (including re-ranking, calibration, and equity of attention) are selected and implemented on the two datasets MovieLens-1M and Lastfm-NL, then comprehensively evaluated in terms of two types of metrics: accuracy and fairness. The experimental results show that different methods are effective in improving different fairness targets, with varying degrees of accuracy loss or gain. This paper further explores the trade-offs between maintaining accuracy and improving fairness on intervention methods, and proposes future improvement directions for fairness-aware recommendation systems in light of the experimental results.

Adolescence is a pivotal period of identity formation, significantly influenced by media consumption, particularly music. While recommender sytems (RS) play a central role in shaping adolescents' music exposure, they are typically designed without considering developmental needs. We do not yet understand how identity development manifests in behavioral data over time. And without this understanding, we cannot design systems that meaningfully support it. This study addresses this critical gap by applying James Marcia's established Identity Status Theory to real-world listening behavior to get insight into how adolescents traverse and transition betweeen distinct developmental phases.
Utilizing the LFM2-b dataset, we operationalize James Marcia's dimensions of exploration and commitment through Shannon Entropy and Jensen-Shannon Divergence of genre distributions, respectively, to identify and characterize distinct behavioral phases (Diffusion, Foreclosure, Moratorium, Achievement).
Our results show that while most users briefly enter all four phases, sustained engagement is much rarer, particularly for Foreclosure and Moratorium, which appear less stable. In contrast, Diffusion and Achievement are both more frequently sustained, with Achievement emerging as the most common final phase (40.78%). A strong directional transition was observed from Moratorium to Achievement (0.67 probability), while reciprocal transitions between Diffusion and Foreclosure highlight re-evaluation among users. These patterns support the dynamic and recursive nature of adolescent identity development. By linking digital behavior to psychological theory, this work offers both developmental insights and practical implications. It lays the foundation for designing recommender systems that are not only personalized but also support healthy identity formation.