In this extended abstract we use a new type of network, which we call the mention network, for the task of dominance prediction of employees of the Enron corporation given their emails. We show that the network formed using "who mentions whom to whom" links out-performs the traditionally used Enron email network. We present a comprehensive set of experiments to conclude that organizational dominance is best predicted by capturing the number of people mentioned to a person.
Unfortunately, ACM prohibits us from displaying non-influential references for this paper.
To see the full reference list, please visit http://dl.acm.org/citation.cfm?id=2640065.