Local, Relevance, and Japan!

We launched our MSN Search engine on February 1st and have remained hard at work to make it better every day.  Yesterday, we released some improvements we hope you’ll like.  Much of this work is targeted directly at feedback that our blog readers and other customers have given us, so please keep the comments coming!  We have a long way to go to achieve our vision of the perfect search engine – and we can’t do it without your continued feedback and support..

Our number one user request was to improve and expand upon our local search offering.  The ‘Near me’ feature we launched in February was missing something really important – the ability to look up yellow page and white page numbers.  Our new local search beta integrates maps, yellow pages, white pages, and web results. It also integrates satellite images from TerraServer to provide you with an overhead shot of where you want to go. This makes it much easier to find a local pizza place or art museum! You can access this beta by heading to our local search page or by entering a query term with local data (such as a zipcode or city, state) in our mainline search box.

Our team focuses obsessively on the relevance of the web results.  We want to get the most relevant result in the top position all of the time.  The ranker we released in February served us well, but had some flaws that we weren’t happy about.  In collaboration with Chris Burges and other friends from Microsoft Research, we now have a brand new ranker.  The new ranker has improved our relevance and perhaps most importantly gives us a platform we think we can move forward on quicker than before.  This new ranker also is based on technology with an awesome name – it’s a neural net, which we internally call “RankNet”.  For you Star Trek fans out there – I keep thinking of the android Data powering our search engine.
 
Pictured below is an illustration of the relevance for the query “pbs evolution videos” over time. With the oldest on top, you can see that over the past few weeks we’ve brought high quality results into the top positions and moved the official PBS page to position 1.  The “correct” page is highlighted in green below and you can see it move into position #1.

Aided by our new ranker, we were able to produce particularly big relevance gains in Japan. The queries tsutaya (A Multimedia Entertainment Shopping Company), win2000, and kat-tun (a music band) all showed noticeable improvement.  The feedback we’ve received so far has been very positive so we think we are on the right track.

You’ll notice that we’ve expanded our sports instant answers too. You can now get instant answers for your favorite professional baseball teams and players. By using MSN Search, you can see that Ichiro’s batting average is back over .300 – whew.

We’ve also added a number of new operators that will enable you to narrow your search to exactly what you’re looking for. We’ve added FileType:, one of the most asked for operators, which restricts documents to a particular filetype. InAnchor:, InURL:, InTitle:, and InBody: are now available to find keywords in a particular part of the document, or in anchor text pointing to a document. We’ve augmented the Link: keyword that finds documents that link to a particular page with LinkDomain:<domain>, which finds documents that point to any page in a domain. Finally, we’ve added a new experimental operator called Contains:. Contains: returns documents that contain hyperlinks to documents with a particular file extension; for example, contains:wma returns documents that contain a link to a WMA file.

We hope you like these new improvements – please leave us a comment to tell us what you think.

 Ken Moss, MSN Search General Manager