Legal Aspects of Recommender Systems in the Web 2.0: Trust, Liability and Social Networking

Abstract

Along with an increasing sophistication of products, services and contents migrating to the digital space, the Social Web has revolutionized the architecture of social relationships. In a social networking environment, the traditional legal analysis is challenged by new behavioral patterns likely to evoke certain privacy concerns. On the one hand, users are actively involved in the provision of relevant information intended to fuel personalization strategies and recommender systems. Therefore, the recommending activity becomes decentralized and users go from passive observer to active recommendation provider. On the other hand, the intense participation in social networks represents an infinite source of valuable information. Since like-minded individuals assemble and share interests and preferences, regular and repetitive social patterns in searching, browsing, selecting or buying offer a wealth of (shared) data about users and represent a valuable source of information. Both assumptions, linked to the Social Web, call for exploring the need to redefine the legal approach to recommender systems under new coordinates.

This paper aims to examine the development of recommender systems, considering the new patterns of social behavior, and tackle the emerging legal concerns aroused by the collection and process of personal and social information and the legal consequences of devising a decentralized recommendation model. As a conclusion, a model likely to achieve a balance between privacy risks and personalization advantages is proposed. Service providers are encouraged to devise technical and organizational measures intended to minimize data collection and guarantee a free and well-informed user consent giving by implementing opt-out mechanisms and deploying ambitious transparency and disclosure policies, enabling individuals to build their digital ego on a sound consent-directed basis.