[Chandler-dev] [MVA] Next steps after SoC project

[Xun Luo]

Hi Philippe,

Sorry for being away from the project for a week. The new semester starts
and my advisor is instructing a graduate-level Human-Computer Interaction
class, so I was assigned to collect papers for class reading, as well as
preparing 2-3 group projects till this Friday. Luckily I am not a TA for
this class so there will be no further assignments after this:).

I am however be able to do 1) and 2) this week. I have read about the new
parcel structure described in
http://lists.osafoundation.org/pipermail/chandler-dev/2006-March/005552.htmland
think it provides enough information for the egg based thing. I will
follow-up with people (Makku and Viksit) to make sure these are done this
week.

(There is no 3)? )

For 4), I think we should definitely include MDP as one of the standardized
tools.

I will start doing 5) next week. Also thank you very much for the help all
across summer during GSoC!

Bear,
I will start a separate thread on list so we could discuss incorprating this
package to, say, /util or /tool.

Viksit,
Making the the parcel to run is pretty automatic. The only issues are: 1)
it's based on 0.7a1, so a 0.7a1 installation is needed. 2) I am not
branching the whole 0.7a1 tree but just the modified files. So there is a
manual step to checkout then overwrite these files. With delivery for
0.7a4this should be fixed. Besides these issues, it is usable as same
as the
original feed parcel. Also there is an installed one on Haleakala.

Best,
Xun

On 8/30/06, Viksit Gaur <vik.list.chandler at gmail.com> wrote:
>
> Hi everyone,
>
> > - a working parcel to make experiments
> > - a set of data performed on Slashdot feeds showing that auto tagging
> > after a learning phase is feasible
> > - a path to incorporate PCA into Chandler (using MDP)
>
> Awesome. I was wondering if the alpha1 (or 3?) code is usable at this
> point, so that I might be able to play around with it? I remember Xun
> talking about some steps needed to make the code run, and from what I
> gather from Philippe, things weren't totally automated, and a number of
> steps were needed to run them.
>
> >
> > I've been discussing with vikSIT on IRC what should be the next steps.
> > Eventually, the goal will be to have auto tagging turned on (optionally)
> > on the trunk but we won't even see that in Chandler before we have a
> > tagging feature in (planned for alpha6). In the meantime, there's a set
> > of short term things we should do:
>
> Ah. Couldn't we do it using some hard coded tags anyway? These tags
> needn't be applicable to all the other Chandler data, and can instead be
> just dummy data.
>
> > 2- update the parcel to be running against the 0.7alpha4 trunk code:
> > right now, it runs against 0.7alpha1 which is kind of old. I know that
> > the egg stuff sort of threw a wrench in your plan but we should get pass
> > that hurdle. May be vikSIT or other contributors (Markku?) could help
> > there.
>
> Yep, that might be fun. I'm not sure what the general egg problem was,
> to be honest - could I have some more information about it?
>
>
> > 5- make the relevant changes in the MVA project to call the MDP library:
> ..
> > that we consider sandbox/xluo as the official repo for MVA.
> Sounds good.
>
> > BTW, just as a roll call, who on this list would be interested to
> > contribute to this project moving forward?
>
> Me, for one :)
>
> Cheers,
> Viksit
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.osafoundation.org/pipermail/chandler-dev/attachments/20060830/0ef40020/attachment.htm