This paper proposes FairGENRec, a fair sequential recommendation system based on the Diffusion Model (DM). To solve the problem that existing recommendation systems learn correlations with users' sensitive features (such as gender and age) and cause unfairness, we utilize the uncertainty modeling and diversity representation capabilities of DM. FairGENRec injects random noise into the original distribution through a sensitive feature recognition model and reconstructs items with a sequential denoising model. At the same time, it models the fairness of recommendations by injecting various interest information that removes the bias of sensitive user features into the generated results. In the inference phase, noise is added using users' past interactions and the target item representation is reconstructed through back-iteration. The experimental results on three datasets show that FairGENRec is effective in improving both accuracy and fairness, and the degree of fairness improvement is visualized through case analysis.
