tag:blogger.com,1999:blog-6569681.post1436269490113427565..comments2024-03-24T10:38:16.997-07:00Comments on Geeking with Greg: Designing search for re-findingGreg Lindenhttp://www.blogger.com/profile/09216403000599463072noreply@blogger.comBlogger5125tag:blogger.com,1999:blog-6569681.post-72990342131132531122010-03-18T12:30:18.453-07:002010-03-18T12:30:18.453-07:00In a paper I wrote I proposed something I called q...In a paper I wrote I proposed something I called query cacheing - i.e., cache queries or better the results for some time so you do not have to run them again - helps to keep the load low when you spawn queries across multiple systems.<br /><br />Good post keep it up<br /><br />regards<br /><br />swsswshttps://www.blogger.com/profile/05525969285085271353noreply@blogger.comtag:blogger.com,1999:blog-6569681.post-50733507746197380012010-03-17T03:30:14.804-07:002010-03-17T03:30:14.804-07:00Great post, Greg! Thanks for sharing. (long time r...Great post, Greg! Thanks for sharing. (long time reader first time commenter)<br /><br />Was about to suggest that second paper you mentioned in your earlier post in '07 for additional reading :) (Its possible I may have discovered that paper on your blog ;P). I found the stat that 40% of all queries are re-finding queries pretty surprising when I first came across it.<br /><br />As an aside, am also a co-founder of Infoaxe (thanks for the shoutout AM! :)). <br /><br />One thing we found fascinating as an application of Web History search was that it allowed people to be lazy while searching. We have a widget which displays results from your Web memory/history alongside results from Google. For eg. if I have been to your blog once before, I can just type "Greg" into Google and Infoaxe's top result will be your blog. A lot of our users use Infoaxe for this sort of implicit personalization and regionalization (for eg. instead of "siam royal palo alto" you could type in just "siam royal") being less specific than they would otherwise need to be. We were surprised since we imagined most of the utility for a Web History Search engine would be to find that arcane page that you knew you should have bookmarked. :)Jonathan Siddharthhttps://www.blogger.com/profile/07788292704272921873noreply@blogger.comtag:blogger.com,1999:blog-6569681.post-25930168612722838202010-03-16T14:40:50.809-07:002010-03-16T14:40:50.809-07:00Infoaxe is not a new idea, AM. See, for an early ...Infoaxe is not a new idea, AM. See, for an early example, "<a href="http://research.microsoft.com/apps/pubs/default.aspx?id=64328" rel="nofollow">Stuff I've Seen</a>", also out of Microsoft Research.<br /><br />But searching over a cached copy of your browsing history also was a the seed of a few startups a decade ago in the dot-com boom, as well as being a common feature in the many desktop search engines back when that was a hot idea several years ago. And, of course, there are the gazillion web bookmark services out there, all of which do things fairly similar to Infoaxe.<br /><br />The hard part with this idea, beyond that the space is extremely crowded with also-rans, is getting people to download something. That's quite a hurdle. Needs a lot of obvious value to make it worth it beyond a small initial crowd of early adopters (who will try anything once).Greg Lindenhttps://www.blogger.com/profile/09216403000599463072noreply@blogger.comtag:blogger.com,1999:blog-6569681.post-12696100219932003412010-03-16T14:05:16.820-07:002010-03-16T14:05:16.820-07:00You may want to take a look at: www.infoaxe.com. A...You may want to take a look at: www.infoaxe.com. According to the website:<br /><br />"Infoaxe is a Search Engine for your Web Memory"Unknownhttps://www.blogger.com/profile/10896547115861688406noreply@blogger.comtag:blogger.com,1999:blog-6569681.post-10669007072029220602010-03-16T12:04:16.516-07:002010-03-16T12:04:16.516-07:00This is brilliant.
Amazon has been doing somethin...This is brilliant.<br /><br />Amazon has been doing something related for years: tracking the items you have browsed. It is one of the great features that set Amazon apart.<br /><br />A big problem I face with Google Scholar is that often, weeks ago, I found some exciting paper, only to "lose it". And I'm left thinking... "hmmm, there is a paper about this idea somewhere, I found it once... who wrote it?... I can't remember..."<br /><br />In fact, I thought about writing my own application that would track the papers I read specifically so that it could "remind me" that they are related to what I'm doing right now.Daniel Lemirehttp://www.daniel-lemire.com/en/noreply@blogger.com