Re: Academic Research: Recommending similar bugs

Hello all,

Thank you very much for the feedback. I welcome any comment on this subject.

On Sep 15, 2014 at 10:00 AM, Mike Hoye wrote:
> Certainly as a community manager I'm interested -
> the prospect of automating, even partially, a
> next-good-bug search is pretty compelling from a
> community-building and contributor-fostering sense.

I hope NextBug could help with that. The experiments show reasonable
results so far.

On Sep 15, 2014 at 11:22 AM, Andre Klapper wrote:
> I highly appreciate when academia reaches out to developers.
> There seems to be a huge gap between research and getting stuff
> implemented upstream.

My research group also shares these believes. We are always trying to apply
our research to help developers.

On Sep 15, 2014 at 11:22 AM, Andre Klapper wrote:
> There's been dozens of research papers (and some prototypes) on
> duplicate detection in Bugzilla and I assume you're well aware of them.

I am aware of the papers you mentioned and I am considering theses works to
improve our current prototype.

Basically, our current prototype is indeed more similar to Sun et al. (2011)

On Sep 15, 2014 at 11:36 PM, Gervase Markham wrote:
> Is the code for this extension available anywhere?
> I think it would be good for us to see and evaluate your prototype.

Yes, we have a prototype tool already running. Please look at: <
http://aserg.labsoft.dcc.ufmg.br/nextbug/ >

Of course, we are working hardly now to improve this first prototype. In
fact, by next week we plan to have a new version where the developers can
customize filters to select recommendations on just specific types of bugs
(e.g, bugs of a certain severity, bugs not assigned to anyone, etc).

Thanks again for all the comments.

- Henrique Rocha
Computer Science PhD Student
Applied Software Engineering Research Group  (
http://aserg.labsoft.dcc.ufmg.br/)