I was sitting around contemplating my MythTV setup the other day and it
occurred to me that with the number of deployed MythTV boxes, there
should be some really interesting "social" possibilities. Perhaps
this has already been discussed, but it seems to me that the open
nature of MythTV would make these relatively easy to add.<br>
<br>
The specific case I was thinking about was what to watch on TV.
TiVo has its suggestions capability and I do miss that. But
MythTV could do much better. Instead of looking at my viewing
habits and trying to guess what I'd like, what if you looked at my
viewing habits and then compared them with viewing habits of other
MythTV owners? If you were somehow able to compile all the
viewing information (in an opt-in, privacy friendly way of course),
then you could pretty easily identify patterns. So if I record
"Law & Order" and "Numb3rs" and it notices that most people who
record those shows also record "The Closer" then it would point that
out to me.<br>
<br>
This could be just like the online commerce approach of "people who bought this also bought...".<br>
<br>
This kind of capability wouldn't just have to rely on past data, it
could see into the future. It could recognize patterns in what
people with similar interests will record in the future.<br>
<br>
I wonder if commercial identification could work in a similar
way...combining information about what the various commercial
identifying algorithms identified along with information about what
parts of a program people fast-forwarded or skipped through to create a
nearly definitive map of a broadcast (of course, this requires that
time pass and some people have already watched the episode, but I'm
just brainstorming here).<br>
<br>
As I said, perhaps this has been discussed before, I don't know.
I thought it was an interesting idea and reflects many trends in how
social data is being handled (e.g. <a href="http://del.icio.us">http://del.icio.us</a>).<br>
<br>
If I had the time, I'd probably try and implement something
myself. But being overcommitted already with other obligations,
I'm afraid I have little to offer such a project except perhaps some
ideas, user feedback and encouragement. :-(<br>
<br>
--<br>
Mike<br>
<br>