Performance of Recommender Systems: Based on Content Navigator and Collaborative Filtering

TL;DR

This paper evaluates the performance of recommender systems, focusing on collaborative filtering and a new clustering method called Contents Navigator, aiming to improve recommendation accuracy and computational efficiency in big data environments.

Contribution

It introduces a novel clustering approach, Contents Navigator, inspired by complex networks, to enhance collaborative filtering for more stable and efficient recommendations.

Findings

Improved recommendation stability with Contents Navigator

Reduced computational complexity in large-scale systems

Enhanced accuracy over traditional collaborative filtering methods

Abstract

In the world of big data, many people find it difficult to access the information they need quickly and accurately. In order to overcome this, research on the system that recommends information accurately to users is continuously conducted. Collaborative Filtering is one of the famous algorithms among the most used in the industry. However, collaborative filtering is difficult to use in online systems because user recommendation is highly volatile in recommendation quality and requires computation using large matrices. To overcome this problem, this paper proposes a method similar to database queries and a clustering method (Contents Navigator) originating from a complex network.