Incremental Collaborative Filtering for Highly-Scalable Recommendation Algorithms

Abstract

Abstract. Most recommendation systems employ variations of Collaborative Filtering (CF) for formulating suggestions of items relevant to users' interests. However, CF requires expensive computations that grow polynomially with the number of users and items in the database. Methods proposed for handling this scalability problem and speeding up recommendation formulation are based on approximation mechanisms and, even when performance improves, they most of the time result in accuracy degradation. We propose a method for addressing the scalability problem based on incremental updates of user-to-user similarities. Our Incremental Collaborative Filtering (ICF) algorithm (i) is not based on any approximation method and gives the potential for high-quality recommendation formulation (ii) provides recommendations orders of magnitude faster than classic CF and thus, is suitable for online application.

Utilizing a personalization-enabled access node in support of converged cross-domain scoring and advertising

“…However, in the adjacent field of collaborative filtering [20], in order to compute the similarities between users, a variety of similarity measures have been proposed, such as Pearson correlation, cosine vector similarity, Spearman correlation, entropy-based uncertainty measure, and mean square difference. Some studies, including those by Breese [4] and Herlocker [16], suggest that Pearson correlation performs better.…”

Mining user navigation patterns for personalizing topic directories

“…Collaborative filtering systems [14,15,16,22] are used for generating recommendations and have been broadly used in e-commerce. Such systems are based on the assumption that users with common interests and behaviour present similar searching/browsing behaviour.…”

Collaborative Filtering on Data Streams

“…The first online algorithm applied to collaborative filtering was the Weighted Majority Prediction (WMP) [8], Delgado et al [9] extended this approach for multi-valued ratings. Papagelis et al [10] developed a method to incrementally update similarities among users, and Domingos et al [11] proposed VFDT, a system that allows the building of decision trees dynamically to mine data streams.…”