Alleviating the new user problem in collaborative filtering by exploiting personality information

Abstract

Esta es la versión de autor del artículo publicado en: This is an author produced version of a paper published in: In this paper we analyze viable solutions to the new user problem in collaborative filtering that are based on the exploitation of user personality information: (a) personality-based collaborative filtering, which directly improves the recommendation prediction model by incorporating user personality information; (b) personality-based active learning, which utilizes personality information for identifying additional useful preference data in the target recommendation domain to be elicited from the user; and (c) personality-based cross-domain recommendation, which exploits personality information to better use user preference data from auxiliary domains which can be used to compensate the lack of user preference data in the target domain. We benchmark the effectiveness of these methods on large datasets that span several domains, namely movies, music and books. Our results show that personality-aware methods achieve performance improvements that range from 6% to 94% for users completely new to the system, while increasing the novelty of the recommended items by 3% to 40% with respect to the non-personalized popularity baseline. We also discuss the limitations of our approach and the situations in which the proposed methods can be better applied, hence providing guidelines for researchers and practitioners in the field.

Context-Aware Personalization Using Neighborhood-Based Context Similarity

“…Therefore, despite being excellent proposals however, they have not sufficiently addressed the cold-start problem. Thus, the new user problem remains a challenge with no unique solution that can be applied in any domain or situation [1]. In this paper, we demonstrate how similarity between contexts of new users and those of other users can be used to minimize the user-based cold-start problem with or without explicit rating information from users as proposed in previous solutions [10-12, 14, 16-22, 29].…”

“…They are confused because of the unprecedented volume of multimedia content that is now available for online access via Internet and multimedia enabled devices. In addition, mobile users are overwhelmed because the content being pushed to them is often not relevant to their current contextual situations [1][2][3][4][5]. Therefore, the delivered content in most cases does not match their preferences.…”

“…The cold-start problem arises when personalization systems lack adequate knowledge of either the new user's preferences or of new items for providing relevant suggestions [6,7]. This often is due to the system's inability to gather sufficient information about users and the items they have preferred in the past [1]. Users normally prefer different content in different contexts.…”

“…To address this kind of cold-start problem, different techniques based on context-aware personalized recommendations have been proposed, assisting mobile users to obtain content according to their contextual preferences [1,2,[9][10][11][12][13][14][15][16][17][18][19][20][21][22][23]. Figure 1 illustrates a typical architecture of a context-aware personalized recommendation system, exemplifying the system being proposed in this article.…”

“…More recently, Fernández-Tobías et al [1] proposed an interesting solution to mitigate the cold-start problem based on (a) exploitation of personality information in a personalitybased CF, to improve the recommendation prediction model, (b) personality-based active learning, utilizing personality information to identify additional useful preference data, elicited from users, in the target recommendation domain and (c) personality-based crossdomain recommendation that exploits personality information to better use preference data from auxiliary domains to compensate for the lack of user preference data in the target domain. This solution however, in addition to explicitly relying on rating information from users, did consider user's contextual information as an enabler for addressing cold-start problem.…”