Adaptive information filtering for dynamic recommender systems

TL;DR

This paper introduces two incremental diffusion algorithms for dynamic recommender systems that provide real-time personalized suggestions without cumulative errors, demonstrated on movie and social bookmarking datasets.

Contribution

The paper presents novel incremental diffusion-based algorithms that ensure fast, accurate, and non-cumulative error recommendations in dynamic environments.

Findings

Algorithms achieve fast response times.

Errors do not accumulate over time.

Effective on multiple real-world datasets.

Abstract

The dynamic environment in the real world calls for the adaptive techniques for information filtering, namely to provide real-time responses to the changes of system data. Where many incremental algorithms are designed for this purpose, they are usually challenged by the worse and worse performance resulted from the cumulative errors over time. In this Letter, we propose two incremental diffusion-based algorithms for the personalized recommendations, which integrate some pieces of local and fast updatings to achieve the approximate results. In addition to the fast responses, the errors of the proposed algorithms do not cumulate over time, that is to say, the global recomputing is unnecessary. This remarkable advantage is demonstrated by several metrics on algorithmic accuracy for two movie recommender systems and a social bookmarking system.