November 10, 2006
There is a trend in recommender systems that I think is extremely interesting: systems are starting to explain themselves. The first place I noticed this was at Amazon in their personal recommendations section, at the bottom of a given suggestion:
In this case, Amazon recommended Moon Palace because I had rated another book by Paul Auster. This makes perfect sense, namely I rated something by an author, so the system recommended other books by the same author. The second place this popped up was at the new social music service iLike. Every time a user views another user’s profile, the system calculates a compatibility score based on how similar your favorite artists are, as shown here:
 In this case, I share interest in the bands ESG, TV on the radio, et al. with this user, so our compatibility is high. When I share more popular artists like Miles Davis or Bob Dylan, my compatibility score is lower. This makes sense since rarer bands suggest a closer connection. Last.fm has added a similar feature called Taste-o-meter.
What’s interesting about these examples is not the algorithm, some augmented form of collaborative filtering, but rather in the way that the algorithm explains itself to the user. Many years ago, with the likes of Firefly and CDNow showing off the power of recommender systems, this sort of behavior would have been considered crazy. Showing to users elements of how your algorithm works? What if they reverse engineer it and copy your methods and copy your system and steal all your users?!
Not likely. For most intents and purposes, recommender systems are within wiggling distance of each other. Netflix is holding a contest to see if theirs can be improved, offering a cool $1M to anyone who can show a 10% gain over their current algorithm. While the current leaderboard shows the best contenders at a 4% gain over the original algorithm, Netflix does not expect people to make the 10% gain necessary anytime soon, suggesting the contest could run until 2011. But companies like Amazon and iLike are making improvements through the way that these algorithms are explained.
Explanation creates understanding, and understanding leads to trust.
What if all systems started to take this approach? We mostly assume that search providers keep their ranking algorithms in a 6-foot safe behind a wall of lasers, but at the same time Google is starting to release more information about PageRank through various systems. Someday we might have search results that explain themselves, while keeping the special sauce away from SEO geeks and spammers. Imagine if a top search result said “This result is first because: your search term was in the title, the author is a well known writer, and the host is a reputable newspaper.” I would probably say “that makes sense,” and in turn I would trust that system even more.
Originally posted by cameron from overstated, remediated by yatta on Nov 10, 2006 at 3:56 PM
|
unmediated.av:
The Weekly Show
drawing from extrastruggle.
We've been having a back channel conversation amongst the trackers at unmediated about how/whether to update the way in which we aggregate, present, and make useable the content on the site, in light of all the various aggregators, digg and its clones, and role model group blog sites that we all consume/use/hate/love. Since we all primarily support open media movements and the freedom of bits and so forth, and with all of us being busy with our primary projects, we are looking for ways to make getting content on the site easier and more streamlined, while making it obvious that we are presenting other sources content. With the availability of open API's for just about any type of media aggegration literally getting past the saturation point, and mashups taking every possible form, we are wondering, is it time to take a step back, or a step forward with how/what we do at umediated? In the course of my surfing today, i found this new site, Boxxet Which just might be the straw that breaks the camel's back in how we all perceive the current mix and match nature of the web as it now stands. What's different about Boxxet from other aggregators and mashups like the newest entry popurls, (which aggregates digg, slashdot, reddit, newsvine, tailrank, and flickr) is that Boxxet is a Website generator. Thats right, just pop in all the urls u want to aggregate (and WHAT from them) choose how u want to format it, plug in the url that u want it to be accessed at... and whammo: Your own site with everyone elses content, and all thats left to do is decide whether googleplex or yahooza is going to be the source of your linklove revenue. And if u have on older domain that u plug this into...well, we all know how the pageranking with search engines work by now. It used to be that u had to have a bit of code knowledge to make all this stuff work. Eyebeam's Re-blog engine which powers this site was not a simple undertaking at the time that Michael Frumin and Michael Migurski put it all together... a half a year before Marc Broadband-mechanicked the term Reblog as his latest buzzword before casting his attention on the ourmedia-meme. (kudo's, kudo's) But now, with the cut and paste mentality of webculture that we at unmediated have helped create, the pace at which people are remixing and repurposing code is accelerating at a rate similar to the curve that we saw with pro-sumer desktop video... almost anyone can do it. I have this sinking feeling in my gut that we will arrive sooner than later at the same existential threshold that the film studios and record labels are squirming under to our joyful cries of "die, dinosaurs, die!". What i am wondering, is how long until my hero of the open-information movement, Cory Doctorow, and the rest of our pals at BB will tolerate re-aggregation and repurposing of his content, (now that he is investing so much more time at the site) before he (or any of one us) screams, "FOUL!" Stewart Butterfield over at Flickr is dealing with this beast at the moment...and i have to admire the dryness with which he states, "I loaded the FlickrCentral pool and firefox got up to using 240mb of ram before dying. So that's not a great user experience, but it's really terrible for Flickr. If it catches on and you don't limit it, we'll have to cut you off :\" Sure, Stewart, blame it on the user experience and firefox. ;) I admire your candor, and personal attention/approach to what has become one of the hottest new BRANDS in Web 2.0 ...that u still have time to be personal and all flickr-fuzzy even after being acquired, but I am sure that your jeans feel like they're fitting a bit tighter all of a sudden. Pretty soon, I expect, a lot of us bell-bottomed infornistas are going to wake up in a similar pair of Jordaches. I'm curious which of us will cut the inseams and sew in another totally different material to keep our style,and which of us will claim that now that we're wearing skintight jeans ("they're really really comfortable...REALLY! You think i should get a pair of Reeboks to go with 'em?"), that the manufacture of bell-bottoms should be forbidden. I point this all out in good humour only to illustrate a point: The times, they are('nt) a changin'>, and Cory just might wake up one day soon in his magic kingdom, and say "Hey, man, where'd all my whuffie go? And he's going to have no choice but to join Walt's pinstripesuits in pushing for copyright extension. It's a pill i hope he (and we) never have to swallow. So i pose the question to our community readers: How do you see unmediated-Are we crossing the boundaries in how we repurpose content? Would you like to see more editorializing? Narrower/Broader scope? Are we a repository of information that you come back to use, or just part of your daily information addiction? Let us know... I, for one, would like to have an idea about what pair of jeans to wear this year ;) michael
Featured Project
Berkeley Conference: Online Video and the Future of Television - Friday, September 30, 2005
This one-day conference brings together archivists, educators, technologists, entrepreneurs, producers, legal experts, and investors to explore the enormous promise offered by the availability of online video and television content. Demonstrations and interactive panel discussions will highlight new video technologies, services, legal issues, and economic models. Participants from diverse – and until now, largely disconnected – specialties will be especially encouraged to interact.
del.icio.us/tag/unmediated
[+]
About unmediated
unmediated is a group blog that tracks the tools, processes,
and ideas being used to decentralize media production and distribution.
archives
November 2006
| Sun |
Mon |
Tue |
Wed |
Thu |
Fri |
Sat |
| |
|
|
1 |
2 |
3 |
4 |
| 5 |
6 |
7 |
8 |
9 |
10 |
11 |
| 12 |
13 |
14 |
15 |
16 |
17 |
18 |
| 19 |
20 |
21 |
22 |
23 |
24 |
25 |
| 26 |
27 |
28 |
29 |
30 |
|
|
|
flickr/tag/
citizenmedia
[+]
|
