Dual Preference Distribution Learning for Item Recommendation

TL;DR

DUPLE introduces a dual Gaussian-principled framework to capture both a user’s general item preferences and specific attribute preferences. By learning a general distribution over item embeddings and a transformed specific distribution over item attributes, DUPLE jointly models inter-preference relationships via covariance and enables explainable recommendations through a summarized preferred attribute profile. The approach combines a general-specific transformation, a BPR-inspired attribute supervision, and a probabilistic ranking objective, and it demonstrates superior performance over multiple baselines on six public datasets while providing interpretable explanations. This dual-preference, probabilistic perspective offers improved recommendation quality and a transparent rationale grounded in user attribute preferences, with potential for more robust and explainable recommender systems in practice.

Abstract

Recommender systems can automatically recommend users with items that they probably like. The goal of them is to model the user-item interaction by effectively representing the users and items. Existing methods have primarily learned the user's preferences and item's features with vectorized embeddings, and modeled the user's general preferences to items by the interaction of them. In fact, users have their specific preferences to item attributes and different preferences are usually related. Therefore, exploring the fine-grained preferences as well as modeling the relationships among user's different preferences could improve the recommendation performance. Toward this end, we propose a dual preference distribution learning framework (DUPLE), which aims to jointly learn a general preference distribution and a specific preference distribution for a given user, where the former corresponds to the user's general preference to items and the latter refers to the user's specific preference to item attributes. Notably, the mean vector of each Gaussian distribution can capture the user's preferences, and the covariance matrix can learn their relationship. Moreover, we can summarize a preferred attribute profile for each user, depicting his/her preferred item attributes. We then can provide the explanation for each recommended item by checking the overlap between its attributes and the user's preferred attribute profile. Extensive quantitative and qualitative experiments on six public datasets demonstrate the effectiveness and explainability of the DUPLE method.

Figures (7)

• Figure 1: Illustration of the proposed dual preference distribution learning framework (DUPLE) for the explainable item recommendation, which jointly learns the user's preferences to general item (blue flow) and specific attribute (green flow) for the better recommendation.
• Figure 2: Illustration of the explanation production of the proposed DUPLE. We first derive the user's preferred attribute profile with the mean vector of the learned user's specific preference distribution. We then use the overlapped attributes between the learned attribute profile and the item's attributes as the explanation.
• Figure 3: The training curves of our DUPLE model on the six datasets. The training loss is shown in grey line corresponding to the right longitudinal axis. MRR and HR@10 on the validation set are shown in orange and blue lines, respectively, corresponding to the left longitudinal axis.
• Figure 4: Visualization of the relationships among the user's preferences of a random user in Women's Clothing (a) and MovieLens-1M (b) datasets, respectively. The darker color indicates the higher relation, where the two highest and lowest preference pairs are surrounded by yellow and red boxes, respectively.
• Figure 5: Examples of the explainable recommendation of the proposed DUPLE method. Each example displays a user's historical preferred items, the preferred attribute profile summarized by DUPLE (the frontier position of the attribute means a bigger preference degree), and a recommended item with its exxplanations.