CoBaR: Confidence-Based Recommender

TL;DR

This paper introduces CoBaR, a confidence-based hierarchical clustering method that dynamically determines optimal user group sizes, improving recommendation accuracy over traditional fixed-size neighborhood algorithms.

Contribution

The paper presents a novel confidence interval and hierarchical clustering approach for adaptive user grouping in neighborhood-based collaborative filtering.

Findings

Outperformed traditional algorithms on four datasets

Adaptive user grouping improves recommendation quality

Hierarchical clustering effectively determines optimal group sizes

Abstract

Neighborhood-based collaborative filtering algorithms usually adopt a fixed neighborhood size for every user or item, although groups of users or items may have different lengths depending on users' preferences. In this paper, we propose an extension to a non-personalized recommender based on confidence intervals and hierarchical clustering to generate groups of users with optimal sizes. The evaluation shows that the proposed technique outperformed the traditional recommender algorithms in four publicly available datasets.