Inferring Contexts From Facebook Interactions: A Social Publicity Scenario

• Published in: IEEE transactions on multimedia Vol. 15; no. 6; pp. 1296 - 1303
• Main Authors: Servia-Rodriguez, Sandra, Fernandez-Vilas, A., Diaz-Redondo, R. P., Pazos-Arias, J. J.
• Format: Journal Article
• Language: English
• Published: IEEE 01-10-2013
• Subjects: data mining; NLP; social contexts; social publicity; Social web
• ISSN: 1520-9210; 1941-0077
• DOI: 10.1109/TMM.2013.2265168

The great acceptation of the Social Web has converted social networks, blogs and wikis in almost perfect advertising mediums. However, many of the current social publicity strategies do not exploit all the potential of these mediums, since they obviate users' online life: the social contexts in which they are involved. Our proposal to reverse this situation is a model to infer users' social contexts by the application of several Natural Language Processing (NLP) and data mining techniques over users' interaction data on Facebook. We take advantage of both Facebook and Groupon APIs to provide a deployment scenario in which knowing users' social life allows ads to target the most potential customers, which is beneficial for both companies and possible customers.