Disentangled Item Representation for Recommender Systems

Abstract

Item representations in recommendation systems are expected to reveal the properties of items. Collaborative recommender methods usually represent an item as one single latent vector. Nowadays the e-commercial platforms provide various kinds of attribute information for items (e.g., category, price and style of clothing). Utilizing these attribute information for better item representations is popular in recent years. Some studies use the given attribute information as side information, which is concatenated with the item latent vector to augment representations. However, the mixed item representations fail to fully exploit the rich attribute information or provide explanation in recommender systems. To this end, we propose a fine-grained Disentangled Item Representation (DIR) for recommender systems in this paper, where the items are represented as several separated attribute vectors instead of a single latent vector. In this way, the items are represented at the attribute level, which can provide fine-grained information of items in recommendation. We introduce a learning strategy, LearnDIR, which can allocate the corresponding attribute vectors to items. We show how DIR can be applied to two typical models, Matrix Factorization (MF) and Recurrent Neural Network (RNN). Experimental results on two real-world datasets show that the models developed under the framework of DIR are effective and efficient. Even using fewer parameters, the proposed model can outperform the state-of-the-art methods, especially in the cold-start situation. In addition, we make visualizations to show that our proposition can provide explanation for users in real-world applications.CCS Concepts: • Applied computing → Online shopping.

Controllable Recommenders using Deep Generative Models and Disentanglement

“…The use of deep generative recommenders allow for disentangled representations, regaining some interpretability [9,24,34]. Most such models use the VAE framework, where the objective is to reconstruct the input with high probability, with an additional loss term constraining the latent space.…”

“…There have been several works that use disentanglement in recommendation: Ma et al [25] assume that disentanglement is generated by user behaviours on 'macro' and 'micro' levels, and show that their models outperform non-disentangled baselines. Ma et al [26] apply disentanglement to the sequential recommendation task, while Wang et al [34] disentangle diverse user-intents using graph based collaborative filtering; Cui et al [9] propose DGCF, which have 'implicit' (unknown) and 'explicit' (known) signals that are disentangled using an RNN based model with a two-step method since some computations are non-differentiable. Wang et al [32] propose using weakly-supervised disentanglement objective on pairs of items.…”