Using maths to examine online communities and networks

The ROBUST (Risk and Opportunity management of huge-scale Business community cooperation) Project has examined how online communities work and how to come up with new methods to understand and manage the risk, data and policies associated with the business. Frequently, community forums depend on the contributions of particular members but individuals may become dissatisfied over time, collaborate less and eventually leave.

The team at Southampton has written one of the core modules that used advanced mathematical techniques, forecasting and operational research to predict what users of online communities will do in the future. "It may seem strange but human beings are predictable to a certain extent," says Jörg Fliege, Professor of Mathematics. "They develop habits, go online at regular times and carry out certain levels of activity. This can be detected and predicted."

Using a demonstrator application of ROBUST technology, the software provider SAP AG, one of the project partners, reveals insights into the SAP Community Network (SCN) - a community with more than two million members. The ROBUST demonstrator identifies the forums that are flourishing, those that will become unpopular and the underlying trends. Even in large networks this technology allows community managers to keep track of the situation and developments and to identify experts and influential users.

In another example, more than 600,000 employees of IBM use the company's Connections platform to exchange ideas and business-relevant information; ROBUST analyses how these business communities develop. The application identifies the branches of the company that are collaborating well and the users that play a central role in this process.