Harnessing Light for Cold-Start Recommendations: Leveraging Epistemic Uncertainty to Enhance Performance in User-Item Interactions

TL;DR

This paper introduces CREU, a novel recommendation framework that leverages epistemic uncertainty to improve cold-start user-item interaction performance by efficiently utilizing training knowledge.

Contribution

It proposes a new approach to measure and optimize epistemic uncertainty in recommendation models, enhancing cold-start performance beyond existing methods.

Findings

CREU outperforms baseline models in cold-start scenarios.

Efficient measurement of epistemic uncertainty improves recommendation accuracy.

CREU demonstrates robustness across multiple public datasets.

Abstract

Most recent paradigms of generative model-based recommendation still face challenges related to the cold-start problem. Existing models addressing cold item recommendations mainly focus on acquiring more knowledge to enrich embeddings or model inputs. However, many models do not assess the efficiency with which they utilize the available training knowledge, leading to the extraction of significant knowledge that is not fully used, thus limiting improvements in cold-start performance. To address this, we introduce the concept of epistemic uncertainty to indirectly define how efficiently a model uses the training knowledge. Since epistemic uncertainty represents the reducible part of the total uncertainty, we can optimize the recommendation model further based on epistemic uncertainty to improve its performance. To this end, we propose a Cold-Start Recommendation based on Epistemic Uncertainty (CREU) framework. Additionally, CREU is inspired by Pairwise-Distance Estimators (PaiDEs) to efficiently and accurately measure epistemic uncertainty by evaluating the mutual information between model outputs and weights in high-dimensional spaces. The proposed method is evaluated through extensive offline experiments on public datasets, which further demonstrate the advantages and robustness of CREU. The source code is available at https://github.com/EsiksonX/CREU.