Abstract:Implicit recommendation refers to the users' feedback on items derived from their interactions with items, i.e., clicks, and purchases. The methods in the implicit recommendation scenario usually regard all the adopted items as their favorites and indiscriminately assign a uniform confidence weight of their preference toward all the adopted items. In practice, however, a user's preferences toward different items vary a lot. Treating them equally in existing implicit feedback recommender systems may limit the capacity of learning algorithms. To address this problem, we propose a novel Gaussian Personalized Recommendation OPTimization criterion (GPR-OPT), and our aim is to make the unknown preference confidence of users toward their adopted items in implicit feedback recommendation be learnable, so as to improve the accuracy of implicit recommender systems. In particular, we assume the user's interests in items follow Gaussian distributions. By maximizing the posterior probability of items derived from the Gaussian distribution of user features, GPR-OPT is able to self-adaptively learn the confidence of users' preferences from the implicit user–item interactions. We conduct extensive experiments on three real-world datasets, i.e., Movielens 1M, Amazon Book, and Yelp, which show an average of 11.64% improvements over different kinds of collaborative filtering algorithms. GPR-OPT is a generic optimization criterion and can be easily integrated into most existing collaborative filtering recommendation models, leading to a great impact on implicit recommender systems.

Gaussian process for recommender systems

Hybrid Gaussian pLSA model and item based collaborative filtering recommendation

Improved Metric-Based Recommender by Historical Interactions.

Gaussian-Gamma Collaborative Filtering: A Hierarchical Bayesian Model for Recommender Systems

Multi-output Gaussian Processes for Uncertainty-aware Recommender Systems

Grid-based Gaussian Processes Factorization Machine for Recommender Systems

A Novel Framework To Process The Quantity And Quality Of User Behavior Data In Recommender Systems

Towards Context-aware Social Recommendation Via Individual Trust

Towards Non-Linear Social Recommendation Using Gaussian Process

Gaussian process latent variable model factorization for context-aware recommender systems

A Novel Non-Gaussian Embedding Based Model for Recommender Systems

Modeling user and item biases with Gaussian distribution for collaborative filtering

A Group-Specific Recommender System

A Recommendation Algorithm Combining User Grade-Based Collaborative Filtering and Probabilistic Relational Models

GPR-OPT: A Practical Gaussian optimization criterion for implicit recommender systems

Applying Probabilistic Latent Semantic Analysis to Multi-Criteria Recommender System

Modeling Users' Dynamic Preference for Personalized Recommendation.

Incorporating Community Detection and Clustering Techniques into Collaborative Filtering Model

Algorithm of sequential recommendation based on Gaussian distribution modeling

Enhancing Recommender Diversity Using Gaussian Cloud Transformation

Real preference Gaussian mixture model