[mythtv] 'recommendations' section

[Kirby Vandivort]

On Sun, Sep 14, 2003 at 12:32:40PM -0400, Michael J. Pedersen wrote:
> On Sun, Sep 14, 2003 at 08:54:58AM -0500, Kirby Vandivort wrote:
> > Right now, the code is purely based on what one person is interested in.
> > For me, it is doing 'ok' right now.  It's giving a bit too much emphasis
> > to some words, so I have to go back in and rein that back a bit.  I had
> > a version during testing that seemed to be giving better answers, so
> > I'm going to go back and find that version and tweak the current numbers.
> 
> Here's a thought to ponder from the email world. Consider that any TV
> show, from a user's perspective falls into one of two categories: "Want
> to watch" and "Don't want to watch". Now, this idea may very well suck,
> but take a cue from email: Good email and spam. A very hot topic there
> is Bayesian filters, and they seem to work reasonably well. Maybe
> there's a way to do something similar for tv shows?
> 
> It may work, it may not, it may be the worst idea you've ever heard, but
> I thought I'd pass it along.

http://www.gossamer-threads.com/perl/mailarc/gforum.cgi?post=67532;search_string=bayesian;guest=1177793&t=search_engine#67532

:)

It's not the worst idea i've ever heard.  As I was trying to implement
such a thing, I ran into a roadblock, though..  It's not quite the same
as email.  For email, you have a corpus that can clearly be defined as
'this is spam' or 'this is not spam'.  We don't really have an equivalent
thing in (current) myth.  You have a 'this is a good show' and 'these
are other shows'.  We can't really say accurately that every show that is
on that wasn't recorded is something that the user doesn't want to see.

So, another plan that I have for helping with this is a rating system.
I'm currently thinking each show gets rated 1-10 (possibly with a slider
of some sort that you can set).  This gets translated into the database as
a new table and then this table could be used to do the bayesian filtering
(as well as better normal recommendation calculations).  But, I haven't
had time to work on that yet.  If someone else wanted to add that it
would definitely be something that my code could take advantage of.

-- 

Kirby Vandivort                      Theoretical and Computational Biophysics 
Email: kvandivo at ks.uiuc.edu          3051 Beckman Institute
http://www.ks.uiuc.edu/~kvandivo/    University of Illinois
Phone: (217) 244-5711                405 N. Mathews Ave
Fax  : (217) 244-6078                Urbana, IL  61801, USA