Abstract:There exist complex interactions among a large number of latent factors behind the decision making processes of different individuals, which drive the various user behavior patterns in recommender systems. These factors hidden in those diverse behaviors demonstrate highly entangled patterns, covering from high-level user intentions to low-level individual preferences. Uncovering the disentanglement of these latent factors can benefit in enhanced robustness, interpretability, and controllability during representation learning for recommendation. However, the large degree of entanglement within latent factors poses great challenges for learning representations that disentangle them, and remains largely unexplored in literature. In this paper, we present the SEMantic MACRo-mIcro Disentangled Variational Auto-Encoder (SEM-MacridVAE) model for learning disentangled representations from user behaviors, taking item semantic information into account. Our SEM-MacridVAE model achieves macro disentanglement by inferring the high-level concepts associated with user intentions (e.g., to buy a pair of shoes or a laptop) through a prototype routing mechanism, as well as capturing the individual preferences with respect to different concepts separately. The micro disentanglement is guaranteed through a micro-disentanglement regularizer stemming from an information-theoretic interpretation of VAEs, which forces each dimension of the representations to independently reflect an isolated low-level factor (e.g., the size or the color of a shirt). The semantic information including visual and categorical signals extracted from candidate items is utilized to further boost the recommendation performance of the proposed SEM-MacridVAE model. Empirical experiments demonstrate that our proposed approach is able to achieve significant improvement over the state-of-the-art baselines. We also show that the learned representations are interpretable and controllable, capable of potentially leading to a new paradigm for recommendation where users have fine-grained control over some target aspects of the recommendation candidates.

Semi-Disentangled Representation Learning in Recommendation System

Attribute-driven Disentangled Representation Learning for Multimodal Recommendation

Disentangled Representation Learning for Recommendation

Disentangled Item Representation for Recommender Systems

Disentangled causal representation learning for debiasing recommendation with uniform data

Knowledge-Guided Disentangled Representation Learning for Recommender Systems

Multimodal Disentangled Representation for Recommendation

Causal Disentanglement for Semantic-Aware Intent Learning in Recommendation

Disentangled Multi-interest Representation Learning for Sequential Recommendation

Causal Disentanglement for Semantics-Aware Intent Learning in Recommendation

DualVAE: Dual Disentangled Variational AutoEncoder for Recommendation

Learning Disentangled Representations for Recommendation

When Search Meets Recommendation: Learning Disentangled Search Representation for Recommendation

Disentangled Hypergraph Collaborative Filtering for Social Recommendation

Preference or Intent? Double Disentangled Collaborative Filtering

Representation Learning Via Dual-Autoencoder for Recommendation

Disentangled Representation for Diversified Recommendations

Denoising Item Graph with Disentangled Learning for Recommendation

Curriculum Co-disentangled Representation Learning Across Multiple Environments for Social Recommendation.

Discrete Social Recommendation.

On Causally Disentangled State Representation Learning for Reinforcement Learning based Recommender Systems