Google has just integrated music search, called “Google Discover Music“, into its search results – in US only so far. Partnering up with imeem, lala, myspace (iLike), Pandora, Rhapsody as well as the major music record labels, Google is striving to help users find and listen to music in just a few clicks. As much as 2 out of the top-10 searches are music related, which really suggests that many people are looking for music at Google.
Google “similar images” search is now out of beta (Cnet article), and it looks very promising at first try. The problem with the traditional image search is that it relies on keywords to be found either in image filename or in the text surrounding the image on the web page. With the new feature, a real image similarity comparison is done and can therefore bring back many images that was previously hard or impossible to find.
After doing a normal image search, many of the results will let you click “find similar” (see illustration), bringing similar images. I tried it with the pyramids in Egypt which worked well. Searching for “laptop” brings up many different brands, and clicking similar on a MacBook really brings up other macbooks, on a Dell brings up other Dells (mostly) and on a Vaio brings up other Vaio’s.
I remember that FAST had similar capabilities even 5 years ago, which was piloted in a few showcases and select customer installations. For instance it was utilized for various Police departments to reveal child pornography on the internet. Now Google brings the same kind of technology to the masses, making search one step easier for all of us. Bravo!
Google runs a web page called AllForGood which helps people find opportunities to volunteer in various organizations in their neighbourhood. In the beginning, the search in the site was updated from Google’s crawlers crawling several volunteer webpages.
However, when designing a tighter integrated, more real-time search, they turned to Apache Solr. On their blog they say
“…our search engine is now powered by SOLR, an incredible open source project that will allow us to provide higher quality and more up-to-date opportunities.”
What a super testimony for Solr’s strenths and maturity!
Some people are perhaps more comfortable learning through a book than through Wiki pages and README files.
For years we’ve had books about Lucene, and now comes the first book about Apache Solr as well, titled Solr 1.4 Enterprise Search Server.
The book is written by David Smiley and Eric Pugh and is available for pre-order from Amazon. We’ll post a review once we’ve read it. Until then, read more about the book from the authors at Amazon.
Update: Packt Publishing is sending me a copy of the book, and if you order it from this link I’ll even get a share 
The Norwegian IT newspaper Digi.no today writes that CIA invests in Lucid Imagination through the company In-Q-Tel. This is a great confirmation that the Apache Lucene/Solr search engine is high-end software of key value also for intelligence use.
That would come as no surpise to us who have followed Lucene for some time. And perhaps for intelligence in particular, they do not want to be too heavily dependent upon closed source code. In their press release, In-Q-Tel they say that they want to secure advanced access to Lucene/Solr technology, and state that Lucene/Solr is one of the fastest growing search solutions over the last three years.
This is good news for all users of Lucene/Solr search, also in Norway where Cominvent is Lucid Imagination’s partner. It gives even more credibility to open source search, and makes this platform grow even faster. Time for Norwegian intelligence to follow?