So the art is being built and tagged in a multisensory way? (Or at least lends itself to multisensory-aware tagging?) In these cases, time and location may not be central or obvious, so the 'movement' from one to a semantically close piece happens with no physical dimension grounding. Ok, so on Pandora, for instance, a significant amount of taxonomic work was done to facilitate matching. We won't have that, but we will have so many taxonomies to draw from we can use a "weight of the evidence" approach to mapping?
Here again, having a visualization of the movement from one set of points (essentially, the tags associated with the different indexes) to another would be (truly) insightful. I've got no idea what that looks like, and maybe it's not even representable visually, but it darn sure is going to be in the code we would have to write. Once we get past 5 dimensions, we'll drop into pure abstractions. Hmmm... did someone say, "The five senses?"
Build this, and 100 dissertations start in our names.
Like steve.museum or the wefeelfine.org data, in order to make taxonomic connections. Combine those tags with the geo-pathing information and the evoked response and it's a more powerful recreation of the moment.
Beyond the cultural and technical dissertations I think there is a commercial seed here - think about product/promotion placement. What can we learn that would increase the impact and therefore the ROI for advertising efforts?
Maybe we should talk to Bob Propst a little before we head down this path? ;)
Steve.museum reminds me of that game some phd put up on the internet that had people "competing?" to arrive at the same term for indexing an image. They played each other, the system could play them, and eventually it arrived at the top indexing terms, very specific, short, and very meaningful together. This is what steve is trying to do, I think, although they've got very sophisticated ways to try to capture these (and add semantics in the capture?). I like the idea.
wefeelfine.org is a mood-catcher! I really like the idea of capturing a certain set of words as emotions, as we would capture ?locations? ?sounds[music]? ?textures? with word sets that we identify as part of our own development and user interaction experience. there may be a really simple way to attack this by using the multi-sensory stuff as index types!!!
see the video from the Carnegie Mellon PhD Luis von Ahn about his ESP game (which Google bought and turned into Image Labeler) (see video below)
the ESPGame site is now re-directing to http://www.gwap.com/gwap/, which would be interesting for us to dig into a bit. GWAP stands for Games With A Purpose
Sometimes I think I might be better at seeing how various pieces of the puzzle fit together rather than starting from scratch! Of course I then think I need to go get my PhD too, and that ain't gonna happen anytime soon!
semantic closeness of art, movement
Kevin — Mon, 07/21/2008 - 00:23So the art is being built and tagged in a multisensory way? (Or at least lends itself to multisensory-aware tagging?) In these cases, time and location may not be central or obvious, so the 'movement' from one to a semantically close piece happens with no physical dimension grounding. Ok, so on Pandora, for instance, a significant amount of taxonomic work was done to facilitate matching. We won't have that, but we will have so many taxonomies to draw from we can use a "weight of the evidence" approach to mapping?
Here again, having a visualization of the movement from one set of points (essentially, the tags associated with the different indexes) to another would be (truly) insightful. I've got no idea what that looks like, and maybe it's not even representable visually, but it darn sure is going to be in the code we would have to write. Once we get past 5 dimensions, we'll drop into pure abstractions. Hmmm... did someone say, "The five senses?"
Build this, and 100 dissertations start in our names.
There may be some things we can leverage
christine — Mon, 07/21/2008 - 09:46Like steve.museum or the wefeelfine.org data, in order to make taxonomic connections. Combine those tags with the geo-pathing information and the evoked response and it's a more powerful recreation of the moment.
Beyond the cultural and technical dissertations I think there is a commercial seed here - think about product/promotion placement. What can we learn that would increase the impact and therefore the ROI for advertising efforts?
Maybe we should talk to Bob Propst a little before we head down this path? ;)
steve.museum, wefeelfine.org great places to start thinking
Kevin — Mon, 07/21/2008 - 23:53Steve.museum reminds me of that game some phd put up on the internet that had people "competing?" to arrive at the same term for indexing an image. They played each other, the system could play them, and eventually it arrived at the top indexing terms, very specific, short, and very meaningful together. This is what steve is trying to do, I think, although they've got very sophisticated ways to try to capture these (and add semantics in the capture?). I like the idea.
wefeelfine.org is a mood-catcher! I really like the idea of capturing a certain set of words as emotions, as we would capture ?locations? ?sounds[music]? ?textures? with word sets that we identify as part of our own development and user interaction experience. there may be a really simple way to attack this by using the multi-sensory stuff as index types!!!
Let's find texture-catchers! We've got location-catchers! Smell-catchers? [sounds fishy [sic]].
Google Image Labeler
christine — Tue, 07/22/2008 - 09:32at http://images.google.com/imagelabeler/
see the video from the Carnegie Mellon PhD Luis von Ahn about his ESP game (which Google bought and turned into Image Labeler) (see video below)
the ESPGame site is now re-directing to http://www.gwap.com/gwap/, which would be interesting for us to dig into a bit. GWAP stands for Games With A Purpose
Sometimes I think I might be better at seeing how various pieces of the puzzle fit together rather than starting from scratch! Of course I then think I need to go get my PhD too, and that ain't gonna happen anytime soon!
Post new comment