Comments on: Microsoft and Information Retrieval http://www.scienceforseo.com/information-retrieval/microsoft-and-information-retrieval/ a bridge between worlds Sun, 08 Aug 2010 04:54:41 +0000 hourly 1 By: CJ http://www.scienceforseo.com/information-retrieval/microsoft-and-information-retrieval/comment-page-1/#comment-1087 CJ Fri, 03 Jul 2009 14:27:24 +0000 http://www.scienceforseo.com/?p=1091#comment-1087 Thank you matey!<!-- Touched by JuLiA --> Thank you matey!

]]> By: Daily Links for Friday, July 3rd, 2009 http://www.scienceforseo.com/information-retrieval/microsoft-and-information-retrieval/comment-page-1/#comment-1085 Daily Links for Friday, July 3rd, 2009 Fri, 03 Jul 2009 12:05:16 +0000 http://www.scienceforseo.com/?p=1091#comment-1085 [...] Microsoft and Information Retrieval | Science for SEO [...]<!-- Touched by JuLiA --> [...] Microsoft and Information Retrieval | Science for SEO [...]

]]> By: Guy Valerio http://www.scienceforseo.com/information-retrieval/microsoft-and-information-retrieval/comment-page-1/#comment-1084 Guy Valerio Fri, 03 Jul 2009 11:50:49 +0000 http://www.scienceforseo.com/?p=1091#comment-1084 Try this one: http://bit.ly/kEuZf<!-- Touched by JuLiA --> Try this one:
http://bit.ly/kEuZf

]]> By: Guy Valerio http://www.scienceforseo.com/information-retrieval/microsoft-and-information-retrieval/comment-page-1/#comment-1083 Guy Valerio Fri, 03 Jul 2009 11:38:14 +0000 http://www.scienceforseo.com/?p=1091#comment-1083 >but seem to have lost the slides I 'Binged' (???) for them and have possibly found them, though Powerpoint won't let me view them. http://bit.ly/qXM7L<!-- Touched by JuLiA --> >but seem to have lost the slides

I ‘Binged’ (???) for them and have possibly found them, though Powerpoint won’t let me view them.
http://bit.ly/qXM7L

]]> By: Guy Valerio http://www.scienceforseo.com/information-retrieval/microsoft-and-information-retrieval/comment-page-1/#comment-1082 Guy Valerio Fri, 03 Jul 2009 10:50:36 +0000 http://www.scienceforseo.com/?p=1091#comment-1082 >The image search is phenomenal I agree - I've always thought the UX of the image search from "Live" had been better. I do remember an interesting talk from Nick Craswell of MSFT's research group in 2007 regarding it but seem to have lost the slides. Guy<!-- Touched by JuLiA --> >The image search is phenomenal

I agree – I’ve always thought the UX of the image search from “Live” had been better. I do remember an interesting talk from Nick Craswell of MSFT’s research group in 2007 regarding it but seem to have lost the slides.

Guy

]]> By: CJ http://www.scienceforseo.com/information-retrieval/microsoft-and-information-retrieval/comment-page-1/#comment-1077 CJ Thu, 02 Jul 2009 23:34:00 +0000 http://www.scienceforseo.com/?p=1091#comment-1077 I love it too! It shakes things up.<!-- Touched by JuLiA --> I love it too! It shakes things up.

]]> By: CJ http://www.scienceforseo.com/information-retrieval/microsoft-and-information-retrieval/comment-page-1/#comment-1076 CJ Thu, 02 Jul 2009 23:33:17 +0000 http://www.scienceforseo.com/?p=1091#comment-1076 I think it's exactly what you said. It's important to think laterally about problems in computing. I'm always surprised how many times common sense lets me down! The solution is usually more subtle. I guess that if a site about Tea links to your blog post about Import tax, it would be related thematically but not from a vocab point of view so there's that to consider as well.<!-- Touched by JuLiA --> I think it’s exactly what you said. It’s important to think laterally about problems in computing. I’m always surprised how many times common sense lets me down! The solution is usually more subtle. I guess that if a site about Tea links to your blog post about Import tax, it would be related thematically but not from a vocab point of view so there’s that to consider as well.

]]> By: Joe Hall http://www.scienceforseo.com/information-retrieval/microsoft-and-information-retrieval/comment-page-1/#comment-1072 Joe Hall Thu, 02 Jul 2009 16:41:35 +0000 http://www.scienceforseo.com/?p=1091#comment-1072 What I find most interesting about this is the "Site relationship model". I am having a hard time understanding why a site that isn't related to a target site would carry more relavant anchor text. It goes by simple logic that a site that is relevant to the target site would "understand" said topic better and would thus apply better quality anchor text. Are they assuming that a site that is less relevant would have better anchor text because they would be introducing general concepts to their audiance that might not otherwise be familiar? Anyways, excellent post!<!-- Touched by JuLiA --> What I find most interesting about this is the “Site relationship model”. I am having a hard time understanding why a site that isn’t related to a target site would carry more relavant anchor text. It goes by simple logic that a site that is relevant to the target site would “understand” said topic better and would thus apply better quality anchor text.

Are they assuming that a site that is less relevant would have better anchor text because they would be introducing general concepts to their audiance that might not otherwise be familiar?

Anyways, excellent post!

]]> By: Disa Johnson http://www.scienceforseo.com/information-retrieval/microsoft-and-information-retrieval/comment-page-1/#comment-1071 Disa Johnson Thu, 02 Jul 2009 16:09:34 +0000 http://www.scienceforseo.com/?p=1091#comment-1071 At the Bing launch, there were several things I noted that they are taking advantage of for improved IR. There was also eye candy (mouse hover instant video playback, no click!). Still, I think machine learning, tweaking to better handle being a target after critical mass (assuming they get there with that) they will have a dangerous search engine. I love that!<!-- Touched by JuLiA --> At the Bing launch, there were several things I noted that they are taking advantage of for improved IR. There was also eye candy (mouse hover instant video playback, no click!). Still, I think machine learning, tweaking to better handle being a target after critical mass (assuming they get there with that) they will have a dangerous search engine. I love that!

]]>