Title:Improving Sequential Recommendations via Bidirectional Temporal Data Augmentation with Pre-training

Abstract:Sequential recommendation systems are integral to discerning temporal user preferences. Yet, the task of learning from abbreviated user interaction sequences poses a notable challenge. Data augmentation has been identified as a potent strategy to enhance the informational richness of these sequences. Traditional augmentation techniques, such as item randomization, may disrupt the inherent temporal dynamics. Although recent advancements in reverse chronological pseudo-item generation have shown promise, they can introduce temporal discrepancies when assessed in a natural chronological context. In response, we introduce a sophisticated approach, Bidirectional temporal data Augmentation with pre-training (BARec). Our approach leverages bidirectional temporal augmentation and knowledge-enhanced fine-tuning to synthesize authentic pseudo-prior items that \emph{retain user preferences and capture deeper item semantic correlations}, thus boosting the model's expressive power. Our comprehensive experimental analysis confirms the superiority of BARec across both short and elongated sequence contexts. Moreover, theoretical examination and visual representation of item embeddings offer further insight into the model's logical processes and interpretability. The source code for our study is available at \textcolor{blue}{\href{this https URL}{this https URL}}.