(Copying the list with some general comments on purpose.)

To the extent you have general ideas to discuss (or specific bugs to
work through) I think you're welcome to ask on the list and see how
much mileage you get out of volunteered info.

For example I can tell you that I think 99% of the problem you're
trying to solve is just to define what it means for two users or items
to be similar. The best corner of the project to look at for that are
clustering / classifications "vectorization" support. That leads to
one form of answer: if you can turn the things into vectors, you can
compute a similarity.

If you're looking for consulting I bet someone could come out of the
woodwork, privately -- might say where you're based geographically.

Sean

On Thu, Oct 20, 2011 at 7:45 PM, Eric Hilberg <[email protected]> wrote:
> Apologies in advance for the non-technical post, but hopefully this is the 
> right audience for it.
>
> We're looking to get in touch with someone (or someones) who's interested in 
> helping a young startup accelerate our work with Mahout.  We're looking at 
> content rather than user or item based approaches, and would love to get in 
> touch with anyone with experience using Mahout for this.  If you are, or know 
> of someone who is, interested and experienced in Mahout and looking for a 
> challenging project, please do get in touch.  In fact, even if you're not 
> looking for a new project, we'd still love to talk to you!  For the sake of 
> saving the entire list from having to look at responses, please email me 
> directly rather than reply to the list.
>
> Best,
>
> Eric
> [email protected]
>
>

Reply via email to