Track

Virtual Communities and Collaborations

Abstract

Many Online Social Networks do not generate sustainable revenues through advertising, even though active usage hasreached enormous scales. To enable more effective advertising strategies in Online Social Networks, it is essential to identifyusers who can affect a large number of friends, acquaintances, or other users in the network. In this context, especially users’future level of communication activity in the Online Social Network plays an important role. A highly active past, however,does not guarantee high levels of future communication activity. Thus, approaches for the prediction of users’ future level ofcommunication activity are needed. Therefore, we transfer a probability-based method that has been primarily developed toforecast purchasing behavior of customers to the context of users’ communication activity in Online Social Networks. Inaddition, we demonstrate the method’s applicability and suitability by using a publicly available dataset of Facebook.com.

Download

Share

COinS