Exploration of communication networks from the Enron email corpus

The Enron email corpus is appealing to researchers because it is a) a large scale email collection from b) a real organization c) over a period of 3.5 years. In this paper we contribute to the initial investigation of the Enron email dataset from a social network analytic perspective. We report on how we enhanced and refined the Enron corpus with respect to relational data and how we extracted communication networks from it. We apply various network analytic techniques in order to explore structural properties of the networks in Enron and to identify key players across time. Our initial results indicate that during the Enron crisis the network had been denser, more centralized and more connected than during normal times. Our data also suggests that during the crisis the communication among Enron’s employees had been more diverse with respect to people’s formal positions, and that top executives had formed a tight clique with mutual support and highly brokered interactions with the rest of organization. The insights gained with the analyses we perform and propose are of potential further benefit for modeling the development of crisis scenarios in organizations and the investigation of indicators of failure.