Search papers, labs, and topics across Lattice.
This paper introduces DUMoE, a novel framework for drift-aware multimodal user representation learning that effectively captures the evolving preferences of users on social media. By integrating a temporal dynamics-aware backbone with a sparse mixture-of-experts approach, DUMoE disentangles multiple latent interests and adapts to user behavior over varying time scales. Experimental results demonstrate that DUMoE significantly outperforms existing methods in predicting user interests and interactions, highlighting its robustness in dynamic environments.
User preferences can shift dramatically over time, but DUMoE's innovative approach captures these changes with remarkable accuracy, outperforming traditional models.
Understanding user preferences from noisy and temporally evolving social media behaviors is fundamentally challenging due to interest drift, where user preferences shift across time and exhibit both multi-scale temporal patterns and diverse co-existing interests. To address this, we propose DUMoE, a unified framework for drift-aware multimodal user representation learning. Our model consists of (i) a temporal dynamics-aware backbone that captures and integrates static profiles, short-term behavioral signals, and long-term dependencies into a coherent representation, and (ii) a sparse mixture-of-experts (MoE) interest adapter that disentangles multiple latent interests via expert specialization and adaptive routing. Each expert models a distinct interest subspace, while a gating network dynamically selects and aggregates a sparse subset of relevant experts for each user. To enable stable and effective optimization, we further introduce a three-stage training strategy that decouples backbone learning, expert specialization, and gating optimization. Extensive experiments on real-world social media datasets show that DUMoE consistently outperforms state-of-the-art methods on both user interest prediction and interaction prediction tasks.
