I have been working on a method to tell people something without
spamming everyone. It's sort of like reddit, but the submitter of a
link is responsible for targeting it to people who will probably like
To get things started, I have submitted many links into the system,
including many siggraph papers:
To see how a particular paper is targeted, click "targeting". For
example, here is the targeting for the SIGGRAPH 2006 "Manga
Note how targeting URLs are used to suggest the paper to people. For
example, the paper is targeted to people who liked an
optimization-based colorization method.
People who are targeted can click "more like this" or "fewer like this"
to build their interest profile. This in turn determines your
submission score. Careful targeting is important because your
submission score determines your submission's rank among personalized
recommendations for targeted people.
Now, perhaps you wrote this paper or like the research direction, so
you would like to promote it as well. Moreover, you might believe that
you can do a better job targeting it so more interested people will see
In that case, click on "evolve" to change its targeting, title, and/or
tags. Note that all instances of the link will remain in the system,
so in fact, many people could try to promote this particular paper in
different ways. Click "family" to see all instances of the link.
To receive personalized recommendations, take a look at the links under
"My Targeted Links". If you would like to receive more personalized
recommendations, give your preferences for some of the links under "My
In any case, I plan to add many research papers to the system. I hope
that people will evolve my submissions and add papers of their own. I
think this can result in an interesting way to promote research of
interest to you and receive well targeted recommendations.
Here are other papers I submitted:
The vast majority of these are not targeted at all, so I would
appreciate help in that regard, particularly if some of those papers