11 December 2008
The first public Gephi release has been successfully launched in October 2008 and has already been tested by dozens of people. As we are motivated to continue the project we are now thinking about a 2009 roadmap. Continually inspired by users comments and our own experience, we want to share our ideas with the community and create a group dynamic.
At this time we are pretty sure to release a 0.6 beta2 soon but this article is about guidelines for next versions, from 0.7 to final 1.0. Plans may change however, we always expect to implement new experimental features based on current research. The software tends to be both a flexible experimental platform and a robust application fulfilling daily needs.
We lead the project to three main directions: Explore networks, Create maps and Organize your work. Although some specific topics are already in development and will be explained later in following articles letâ€™s have a look first at the main goals for 2009.
More mature the project is, more the architecture should be flexible and durable. Giving a real power to plugin development will for sure provide the best chance from Gephi project to fulfill community and emerging projects.
Open the possibility of editing data imported in Gephi, through a data laboratory or annotation. Moreover, have search features to explore and mine network data.
Performance, scalability and flexibility
Continue the work on Dynamic Network Analysis (DNA), integrate network dynamics in visualization and manipulation. Think about tools for tracking burst phenomena, contagion or percolation
As for spatialisation algorithm, create a development platform for clustering and classification algorithm. Besides of the huge interests of clustering algorithms themselves, this area will open the possibility of hierarchical (multi-level) networks and the support of larger networks.
Many works can be done to improve vectorial (SVG) export, with more settings and freedom but a PDF exporter has also to be done and a regular screenshot PNG if necessary. Needless to say more filters and modules has to be developed, they will bring better link analysis and help to build richer maps.
User Interface Design, Usability and Ergonomics
We attach great importance to UI and easy-to-use software. The first feedback we had shows they are still more to do in that. Besides many tiny improvements can be done everywhere, the global rigor of the application has to be enhanced.
As we are developing dynamic features, we also expand the GEXF (Graph Exchange XML Format) specification into the community and keep explaining concepts. In order to make this work circulate, libraries able to create/export GEXF must be developed.
As I said before this may not be all but it’s already enough for keeping our motivation at the maximum level. I hope these guidelines will also motivate again the community to participate, help of any kind is definitively needed.