Trust-Aware Clustering Collaborative Filtering: Identification of Relevant Items

Abstract

Abstract. Identifying a customer profile of interest is a challenging task for sellers. Preferences and profile features can range during the time in accordance with current trends. In this paper we investigate the application of different model-based Collaborative Filtering (CF) techniques and in particular propose a trusted approach to user-based clustering CF. We propose a Trust-aware Clustering Collaborative Filtering and we compare several approaches by means of Epinions, which contains explicit trust statements, and MovieLens dataset, where we have implicitly defined a trust information. Experimental results show an increased value of coverage of the recommendations provided by our approach without affecting recommendation quality. To conclude, we introduce a tool, based on recommender systems, able to assist merchants in delivering special offers or in discovering potential interests of their customers. This tool allows each merchant to identify the products to suggest to the target customer in order to best fit his profile of interests.

Advances in Clustering Collaborative Filtering by means of Fuzzy C-means and trust

“…Analyzing the results of Poste Italiane dataset (see Table 5), we note the highest value of coverage with a comparable RMSE value. In other words, the number of provided suggestion increases from 66.67% of Slope One algorithm, to 86.67% of IFCCF, reaching the highest value with iTRACCF, thus providing an evidence, based on real customers, of the preliminary findings introduced in Birtolo et al (2012).…”

“…It is preliminary introduced in Birtolo, Ronca, and Aurilio (2012) and considers a User Clustering by means of K-means algorithm (User-based Fuzzy Cmeans imply a computational effort without any proved benefits) in the User/item Clustering step, implements Pearson Correlation as similarity measure of two users and introduces a trust index between two users (the User/item Information step).…”

A flexible framework for context-aware recommendations in the social commerce domain

“…Adding products to the shopping cart, "Association rules-based" suggestions are available. Finally, in the last case, once customers are registered in a social-Commerce platform, social network information can further improve the quality and number of suggestions [5]. In particular, the main advantage derived from Social-Commerce information is related to the adoption of trust between users in the recommendation system so that trust-aware recommendation is available.…”

“…Indeed, different cultures, different habits and different preferences complicate this problem and data are often unavailable [12]. Different proposed approaches [13,16,15,5] try to address this challenge. In particular, in [16] the trust is explicitly defined as a trust network which correlated two different users and is propagated to all the users in order to predict, how much an user could trust every other user; while in [5], the concept of trust is evaluated by means of shopping cart of the customers, so that users with a similar desired product are trustworthy.…”

“…In detail, the Trust-aware algorithm [5] merges a Clustering CF and a Clustering-based version of the TaRS proposed by Massa [16]. The Trust-aware algorithm, named Trust-aware Clustering CF (TRACCF), has been introduced in [5] and evaluates the prediction of a rating p of item i for the user u as expressed by Eq. 4.…”