Finding relevant people in online social networks
Year: 2013
Type: dissertation
Abstract: The objective of this thesis is to develop novel techniques to find relevant people in Online Social Networks (OSN). To that end, we consider different notions of relevance, taking the point of view of the OSN providers (like Facebook) and advertisers, as well as considering the people who are trying to push new ideas and topics on the network. We go beyond people's popularity, showing that the users with a lot of followers are not necessarily the most relevant. Specifically, we develop three algorithms that allow to: (i) compute the monetary value that each user produces for OSN provider; (ii) find users that push new ideas and create trends; and (iii) a recommender system that allows advertisers (focusing in local shops, like restaurants or pubs) to find potential customers. Furthermore, we also provide useful insights about users' behavior according to their relevance and popularity, showing - among other things - that most active users are usually more relevant than the popular ones. Moreover, we show that usually very popular users arrive late to the new trends, and that there are less popular, but very active users that generate value and push new ideas in the network.
Author: Diego Sáez-Trumper
Language: English
Subfield: Statistical and Nonlinear Physics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2228244391
