Based on the research a web based tool Stakesource, has been developed. This will send out email to a list of potential stakeholders, allowing them to rate their colleagues.
The talk is about identifying people important to a software project. While the description of the technique suggests automated extraction of information. Soo Ling described a manual process where staff and students were interviewed at UCL and paper forms filled in. People were asked who else was important. I had assumed that the information would be derived automatically from existing online information. As an example LinkedIn and similar systems which relate people to people might be used.
One limitation with any such process is that the project manager first has to nominate the list of potential stakeholders. An example Soo Ling used was for an access card system for UCL. While asking staff and students on campus about the system, the developers forgot to ask external library users disenfranchising a large number of interested people.
It seems to me that with large scale systems with millions of potential users and therefore stakeholders, this method will not scale, even with the use of an automated tool. It might be better to concentrate on getting input online from anyone concerned.