Tao Li

Shiyu Shen, Bin Pan, Tianyang Shi et al.

Bayesian Neural Networks (BNNs) have become one of the promising approaches for uncertainty estimation due to the solid theorical foundations. However, the performance of BNNs is affected by the ability of catching uncertainty. Instead of only seeking the distribution of neural network weights by in-distribution (ID) data, in this paper, we propose a new Bayesian Neural Network with an Attached structure (ABNN) to catch more uncertainty from out-of-distribution (OOD) data. We first construct a mathematical description for the uncertainty of OOD data according to the prior distribution, and then develop an attached Bayesian structure to integrate the uncertainty of OOD data into the backbone network. ABNN is composed of an expectation module and several distribution modules. The expectation module is a backbone deep network which focuses on the original task, and the distribution modules are mini Bayesian structures which serve as attachments of the backbone. In particular, the distribution modules aim at extracting the uncertainty from both ID and OOD data. We further provide theoretical analysis for the convergence of ABNN, and experimentally validate its superiority by comparing with some state-of-the-art uncertainty estimation methods Code will be made available.

Ramesh Baral, Tao Li

The Location-Based Social Networks (LBSN) (e.g., Facebook) have many factors (for instance, ratings, check-in time, etc.) that play a crucial role for the Point-of-Interest (POI) recommendations. Unlike ratings, the reviews can help users to elaborate their opinion and share the extent of consumption experience in terms of the relevant factors of interest (aspects). Though some of the existing recommendation systems have been using the user reviews, most of them are less transparent and non-interpretable. These reasons have induced considerable attention towards explainable and interpretable recommendation. To the best of our knowledge, this is the first paper to exploit the user reviews to incorporate the sentiment and opinions on different aspects for the personalized and explainable POI recommendation. In this paper, we propose a model termed as PERS (Personalized Explainable POI Recommender System) which models the review-aspect category correlation by exploiting deep neural network, formulates the user-aspect category bipartite relation as a bipartite graph, and models the explainable recommendation using bipartite core-based and ranking-based methods. The major contributions of this paper are: (i) it models users and locations based on the aspects posted by user via reviews, (ii) it exploits a deep neural network to model the review-aspect category correlation, (iii) it provisions the incorporation of multiple contexts (e.g., categorical, spatial, etc.) in the POI recommendation model, (iv) it formulates the preference of users' on aspect category as a bipartite relation, represents it as a location-aspect category bipartite graph, and models the explainable recommendation with the notion of ordered dense subgraph extraction using bipartite core-based and ranking-based approaches, and (v) it evaluates the generated recommendation with three real-world datasets.