The DevMine project started as a project in the context of a Big Data class given at EPFL by Christoph Koch for the master of science in computer science. The original concept originates from Robin Hahling and was about creating a database of developers and establishing developers profiles based on their open-source contributions. This database of developers would then be used to offer a service allowing people, typically recruiters from companies, to search for developers with a specific set of skills and ranked by best match based on a user query.
Since then, the idea has evolved in some areas. From the original 7 members, 3 have decided to continue working on DevMine and took it twice as a semester project. From the original Python implementation, nothing much remains. The API server was rewritten in Go, mainly because of poor performances, and several sub-projects have emerged to achieve DevMine original goal and provide a source code analysis framework.
DevMine is made possible thanks to the EPFL DATA lab of Christoph Koch which provides us with the necessary resources. Think about the fact that DevMine has to deal with millions of developers metadata and millions of source code repositories as well that occupy terabytes of data storage. Performing analysis requires a lot of computing resources too and all of this would be impossible at large scale without the lab.
Does this project sounds interesting to you? Or maybe you are curious about it or
want to be part of it? If so, feel free to contact one of the current project's
member, or, better, join us on IRC:
#devmine on Freenode.