diffeRS: A Mobile Recommender Service

Abstract

Abstract-Thanks to advances in mobile technology, modern mobile devices have become essential companions, assisting their users in attaining their daily tasks. It will not be long before these devices will become recommending companions, advising users about what data (e.g., restaurants) and what services (e.g., podcast channels) they may enjoy in the local area at the present time. Because of the very nature of the items (both data and services) being suggested (i.e., location dependent and mobile with respect to the consuming user), recommendations cannot be computed on central servers and then pushed to the users. Rather, a novel decentralised mobile recommender service will have to be developed and deployed; instead of relying on global knowledge about users' profiles, such service will have to exploit the wisdom of local communities to compute recommendations. Moreover, because of resource limitations of mobile devices, the algorithms it will employ will have to be computationally light. In this paper, we propose diffeRS, a totally decentralised mobile recommender service specifically designed for pervasive environments. diffeRS crafts a virtual view of the local community's preferences, by exchanging users' profiles via radio technology (e.g., Bluetooth) during periods of colocation. Profiles are stored locally and recommendations are computed using a lightweight algorithm.As our experimental evaluations demonstrate, diffeRS achieves an accuracy and coverage that are comparable to those of centralized recommender systems in use today. .

Can Matrix Factorization Improve the Accuracy of Recommendations Provided to Grey Sheep Users?

“…The accuracy of the recommendations provided to difficult users is not improved when only difficult users are used in the training set. This conclusion is contradictory to the one presented in (Del Prete and Capra, 2010). (Bobadilla et al, 2012) have proposed to take into account the singularity of preferences of users in the evaluation of the similarity between two users.…”

“…Several techniques have been proposed to perform the identification of GSU in RS. Most of them exploit the properties of the ratings of the resources: Abnormality (Del Prete and Capra, 2010) evaluates to what extent the ratings of a user is distant from the average rating of the resources he/she rated, Abnormality-CRU (Gras et al, 2015) includes the variance of the ratings for each resource, whereas DILikelihood (Gras et al, 2016) relies on the distribution of the ratings, etc.…”

“…These works either simply split the set of users into normal and abnormal users (Del Prete and Capra, 2010), and then learn a model dedicated to each set, or rely on clustering (Ghazanfar and A., 2014), or design new similarity measures between users (Bobadilla et al, 2012).…”

A robust reputation-based location-privacy recommender system using opportunistic networks

“…To alleviate the negative effects of people's privacy concerns about providing their data to a centralised recommender [42], in contrast to the existing works, we investigate the feasibility of deploying a location-privacy recommender in a decentralised fashion. In addition, compared with other proposed decentralised recommender systems [13,24,33], we demonstrate the vulnerability of decentralised recommender to sampling attacks, as explained in the next subsection, and propose a reputation scheme that mitigates the effectiveness of such attacks.…”

Interest Aware Location-Based Recommender System Using Geo-Tagged Social Media

“…User location history-based: These are recommenders that use user location history either through the ratings they give to locations they visited or through their check-in history. For example [17][18][19][20][21][22] provided personalized recommendations for locations taking into account other user's rating using Collaborative Filtering (CF) models. Hence, this improved the quality of recommendation by ignoring poorly-reviewed locations that could match a user profile in the user profile-based recommenders.…”