[gsoc] Review needed --- BGL extension : community detection algorithm
Hi, Thanks for Robert's advise. I've posted my proposal on 04/05 but still didn't receive any reply including email or comment. I will be very appreciated if any one can give me any feedback about my proposal. The link is here: http://www.google-melange.com/gsoc/proposal/review/google/gsoc2011/sunbow/1 And the abstract: Implement new library based on Blondel's method and Newman's method providing community detection functionality into Boost libraries. Through this library, users can get their input direct/indirect weighted/unweighted graph segmented based on the criterion that the nodes within the same community have a more significant relationship than those between different communities. Sorry to bother others. Thanks, Xu Sun
I will be very appreciated if any one can give me any feedback about my proposal.
I'd like to see more detailed proposals regarding the design and implementation of the described algorithms. I can't tell from the proposal if you've spent any serious amount of time looking at the BGL and considering how the algorithms might be implemented in or adapted to the library's style. Andrew
On Wed, Apr 20, 2011 at 12:43 AM, Andrew Sutton <asutton.list@gmail.com> wrote:
I will be very appreciated if any one can give me any feedback about my proposal.
I'd like to see more detailed proposals regarding the design and implementation of the described algorithms. I can't tell from the proposal if you've spent any serious amount of time looking at the BGL and considering how the algorithms might be implemented in or adapted to the library's style.
+1 I gave it a read through and from what I gather from the wording in the proposal that it doesn't look like how BGL implements its generic algorithms and data structures. If you can dive a little deeper into what the algorithms are and what the data structure you're proposing would look like -- and whether you're going to implement concrete Graph models -- then it would be easier for us reading through the proposal to see whether the direction you would want to take would be "acceptable". Cheers! -- Dean Michael Berris http://about.me/deanberris
participants (3)
-
Andrew Sutton -
Dean Michael Berris -
Xu Sun