Data Mining Streams of Social Networks, A Tool to Improve The Library Services

Abstract

The Groupware systems are a valuable source for disseminating information in contexts in which the participation of a group of people is required to perform a task. One such context is the Library, Archives and Documentation. The interactions among users and professionals in this area, who use tools such as Twitter, Facebook, RSS feeds and blogs, generate a large amount of unstructured data streams. They can be used to the problem of mining topic-specific influence, graph mining, opinion mining and recommender systems, thus achieving that libraries can obtain maximum benefit from the use of Information and Communication Technologies. From the perspective of data stream mining, the processing of these streams poses significant challenges. The algorithms must be adapted to problems such as: high arrival rate, memory requirements without restrictions, diverse sources of data and concept-drift. In this work, we explore the current state-of-the-art solutions of data stream mining originating from social networks, specifically, Facebook and Twitter. We present a review of the most representative algorithms and how they contribute to knowledge discovery in the area of librarianship. We conclude by presenting some of the problems that are the subject of active research.