InSuggest: Del.icio.us Recommendations Reborn
Recommendations based on your personal tastes are the holy grail for many services on the web. Yahoo-owned social bookmarking service Del.icio.us has been one of the most compelling opportunities for recommendation technology but to date that opportunity has been missed. The troubled in-house recommendation feature at Del.icio.us hasn't been replaced and 3rd party services have had a very hard time meeting the scaling challenge.
Enter InSuggest for Bookmarks -now offering bookmark recommendations based on your Del.icio.us archive. The recommendations come so fast that it's hard to imagine they are good ones, but after some testing they look quite good to us.
How to Use it
InSuggest for bookmarks is very simple. You enter your Del.ico.us username, it looks at your archive of bookmarks and then recommends other similar pages you might like to bookmark. You can filter by one or multiple tags you've used. Up to 20 tag filter options are provided but you can enter any tag you've used in your account.
You can run anyone's Del.icio.us username through InSuggest and get recommendations, not just your own.
The interface is very nice, it's one of the best uses of Ma.gnolia's Thumbshot.org that I've seen yet, and the whole thing feels fairly smooth. In fact, it almost feels too smooth. The recommendations come to you very, very quickly. No where on the site, or in response to our email inquiry so far, can we find an explanation of how it works.
Despite that, it does seem to work well. There are a limited number of ways to parse Del.ico.us data, though, and we wouldn't be surprised if there's just a touch of caching going on. It's hard to say, but the end result is good. InSuggest developer Dennis Gustafsson was elected "Engineering Hero" by The Swedish Association of Graduate Engineers last year, according to search blog Pandia. So someone's seen behind Gustafsson's work and liked it.
Continued below.
Our Recommendations
Beyond some clarity around process and perhaps basic instructions on use, there are a few other things we'd like to see from InSuggest. The first is a feed for future recommendations. The display is in Javascript so we haven't been able to scrape it yet. We'd also love to see a Greasemonkey script for displaying InSuggest recommendations on top of the Del.icio.us bookmarking popup, archive page and item pages.
Other features that would be nice would be the option to input a Ma.gnolia username instead of just del.icio.us, tooltips to display full item titles that are too long for the basic display and the ability to exclude particular domains from future recommendations. Some sort of user feedback to inform recommendations should be doable.
Finally, the biggest fish in the pond when it comes to Del.icio.us recommendations is user recommendations, not just item level ones. We'd like to see other users be recommended, ideally with those who tend to find items of interest earliest privileged on the list.
That may be too much to ask for, though. It's hard to say. Feeds and user recommendations are the kinds of gifts that keep on giving, though, and are far more compelling than one-off recommendations.
For now, though, we think InSuggest for bookmarks is worth checking out. You could very well discover some things there that are just the kind of thing you've been looking for.
Share
Comments
Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts
That's very cool and very fast - even with my ~1000 bookmarks. I'd love to be able to get RSS feeds of recommendations on a tag by tag basis.
Ma.gnolia has nothing to do with thumbshots.org. I think you might be thinking of thmbnl.com?
For me, 3 of the first 5 recommended URLs were ones I'd already bookmarked, and the other recommendations were moderately interesting (it doesn't seem to handle "via:username" tags very well). But recommendation systems seem terribly difficult to get right, especially on unwieldy sets of data like bookmarks, and so this is a fun thing to play with. A person also recently made a little python script that generates recommended del.icio.us users.
Interesting. If you want to try something similar that you can subscribe to, check out my Delicious Recommendations feed...
http://joelaz.com/post/21608641/delicious-recommendations
It's just a hack, but it offers some similar functionality in RSS format.
The recommendations feature is the key feature that gives Twine an advantage. Of course, a recommender is not based on semweb alone (and I don't believe that Twine's recommendation engine is purely semweb-based -- at least I hope it isn't).
And there's still a serious problem with Del.icio.us: It requires manual tagging. This is where Twine shines. The real-world will NEVER add tags. OK, maybe a tag or two on a rare occasion, but not multiple tags on a consistent basis. The geeks, yes; the masses, no.
What is needed is auto-tagging (preferably with a heavy semweb component), plus a recommender for items, topics (think of twines), and people. At this point, Twine is clearly leading this race.
What I wonder is what Dave Beckett is planning to do with Yahoo's properties. If he can add a semweb layer to Del.icio.us and Yahoo! Groups, tie them together, perform some magic with Pipes, add a recommender, then he'd have a helluva service offering: The best in both social bookmarking and online threaded discussions. Wouldn't this pretty much kick everyone else in the teeth? (Hmmm ... maybe Yang should hold out for more $$$ after all?)
At the SemTech conference, the one speaker from Google was a semweb naysayer. Fair enough. But it seems like lying is SOP at Google, i.e., they often deny what they're up to and who they're targeting. (Google has less transparency than the People's Liberation Army.) So I have to believe that Google also has something in the works and/or Twine/Radar Networks, Qitera, ..., are positioning their bizdev efforts toward being acquired by Google.
(BTW, Marshall, I've seen a live demo of Qitera. It's real. The Economist article was on target. But it's still in alpha, not ready yet for a private beta. So the article you wrote on Qitera was factually correct.)
My biggest complaint about Twine is that I get very little out of it compared to what I put into it. For example, for every 100 items I add, I may find 1 item that I wouldn't have found otherwise. OK, maybe it's really 30 inputs to 1 unique output, but this is still a lousy ratio. This may or may not have anything to do with Twine per se, and may have a lot more to do with the fact that I have very good ways for keeping up, primarily through multiple FeedHub feeds scouring over 1,000 RSS feeds. And most of what I've found on Twine has been "fun" stuff, rarely anything that helps me professionally -- and what I want (and need) is a service that will help me professionally.
Yet, Twine is still the best solution on the market, much better than Del.icio.us, Ma.gnolia or Digg. I'm eagerly awaiting for Twine's major upgrade that will be released this October. Until then, I'm reducing my activity on Twine -- for the reason stated above. (Plus, I have too many other irons in the fire this summer, with limited cycles for Twine.)
The race to watch: Twine vs. Yahoo! And let's not forget what Google may have up its sleeve. And for purely social bookmarking with a semweb flavor, keep an eye on Qitera, too. (At this moment, Qitera's social bookmarking feature seems superior to Twine's social bookmarking feature, but life isn't static for either player, so we'll have to see how things look in Q4/Q1 (2009).) Personally, I'm biased in favor of Twine: I've been the most active Twinerian throughout their private beta (and I'm still in the top ten or so), plus they have a great team!! But it's hard to argue against Yahoo! when it comes to social bookmarking and online discussion groups. So, what is Dave Beckett really up to?
Thanks for this article: I found InSuggest really useful!
Quick comment specifically on InSuggest: How can anyone read it? Stylish (to some), perhaps. But with its black background and its predominantly orange lettering, I find it nearly impossible to read (depending upon what is being displayed).
On the plus side, the recommendations seem pretty good. A collaborative filtering approach, I guess. But I'm pretty sure Twine has (or will) integrate collaborative filtering into its recommendation engine. Also, there are many ways to do collaborative filtering, so InSuggest is just one example of how this can be done.
Overall, I'm impressed with InSuggest.
Insuggest is what I have been waiting for and as I said on my blog all that work of tagging and bookmarking pages on Del.icio.us now seems useful.
I would agree with David that if the design was a little more web 2.0ish it would have helped though that would not stop me from using Insuggest.
Thank you for sharing the information.