langreiter.com plain, simple: 2005-12-13

LANGREITER.COM plain, simple

START

INDEX

2005-12-13

CREATED BY chris • LAST EDITED BY chris 6919 days AGO

Very interesting: Twoday in Zahlen.

2005-12-13-twoByTwo

Announcing the Alexa Web Search Platform

I vaguely remember^* Larry Page or Sergey Brin talking about how they'd want Google to be some kind of Spacelab for web search data, allowing people to perform experiments and large-scale analyses on the data they collected. Somewhere along the way to the fabled AdKingdom, that vision got lost; all that remained was a crippled API.

Probably the reasons weren't all bad: Creating a distributed processing platform at that scale is a formidable task, opening it up to some kind of public opens up more than a few boxes Pandora would've been proud to take care of (as Greg Linden comments: "This is either incredibly bold or absurdly foolish.").

What kinds of applications (apart from a myriad of, er, "vortals") will emerge on that new platform and whether it's a particularly cost-effective way of building those of course remains to be seen (the CPU year coming in at $8640); if I had a killer search algorithm at hand I'd be quite reluctant to hand it over to Amazon. Nevertheless, it's an extremely interesting initiative and might just put Amazon somewhere close to relevancy in the search sector for the ~~first~~ second time (Search Inside the Book was great as well, A9 however ...).

Also, I should learn to formulate English sentences without killing a whole rainforest of parentheses.

^* And it wasn't even a fata morgana: "Another goal we have is to set up a Spacelab-like environment where researchers or even students can propose and do interesting experiments on our large-scale web data.", in The Anatomy of a Large-Scale Hypertextual Web Search Engine

2005-12-13-plalexa