[FILLED] PhD position, provenance function for Gephi

[kaigephi]

I am hoping this can also become a way to support the development of Gephi.

The research topic I have in mind will be something like "Interactive visualisation for provenance handling in Gephi". The simplest "provenance" form is like a history function that can be used for building a "undo" and "redo" for Gephi. However, provenance goes much further than that.

Provenance captures the entire history and intermediate results. Currently, most software only provides a "linear" undo/redo: if you go back to a previous state s0 and start doing something else, you can't go back to any state after s0 because you took a different branch from s0.

Provenance will be built to support "sense making": not only how to go back to a previous state, but also support people solving the problem they have at hand (i.e., finding the trends in the social network evolution).

This is supported by the MAKING SENSE project, which is a £2.3 million (US$ 3.7 million) UK project aimed at researching and developing the next generation visual analytics systems that support information analysts in making sense of large, mixed format, multi-sourced data sets of poor reliability (http://www.making-sense.org/).

The PhD studentships is available from the School of Engineering and Information Sciences at Middlesex University in London, UK. Middlesex University leads the UK Visual Analytics Consortium (UKVAC, http://www.eis.mdx.ac.uk/research/idc/UKVAC/Index.htm), which comprises Imperial College, University College London, Bangor University, and Oxford University. The University also has strong connections with leading US visual analytics research institutes such as National Visualization and Analytics Center (NVAC), Purdue University, and Georgia Institute of Technology.

The Studentships are offered on a full-time basis and cover a maintenance award, fee payments (at the UK/EU rate) and research support costs. The Studentships are for three years, subject to satisfactory progress. The maintenance award is currently £13,590 (about US$22,000) per annum and it is free of both tax and national insurance contributions.

For informal discussions about the topics, please contact me/Kai Xu (k.xu@mdx.ac.uk), at the Interaction Design Centre (http://www.eis.mdx.ac.uk/research/idc/).

[admin]

That is great!

And is a technological challenge by itself regarding the efficiency.

[kaigephi]

Yes, the amount of history/provenance data can be quite large and you don't want to slow the system down by monitoring, collecting and storing the history/provenance information.

It will also be a very interesting problem of how to visualise the history/provenance information, which is itself likely to be a network: nodes are the snapshots of the visualisation at different stages and the edges are the operations/manipulations that cause the changes between snapshots!