Comments on: Latent Semantic Analysis in Ruby http://blog.josephwilk.net/ruby/latent-semantic-analysis-in-ruby.html on AI, The Web, Usability, Testing & Software process Wed, 27 Jan 2010 02:14:31 +0000 http://wordpress.org/?v=2.9.1 hourly 1 By: Chris http://blog.josephwilk.net/ruby/latent-semantic-analysis-in-ruby.html/comment-page-1#comment-1016 Chris Mon, 24 Nov 2008 14:22:52 +0000 http://www.joesniff.co.uk/?p=386#comment-1016 Very interesting article. I wonder, have you seen the Classifier gem ? It has a Latent Semantic Indexer, and relies on bindings to the GNU Scientific Library to do the heavy lifting. I've had no problems installing GSL on OS X, so that may be an interesting alternative to linalg's LAPACK bindings. Very interesting article. I wonder, have you seen the Classifier gem ? It has a Latent Semantic Indexer, and relies on bindings to the GNU Scientific Library to do the heavy lifting. I’ve had no problems installing GSL on OS X, so that may be an interesting alternative to linalg’s LAPACK bindings.

]]> By: Joseph Wilk http://blog.josephwilk.net/ruby/latent-semantic-analysis-in-ruby.html/comment-page-1#comment-1015 Joseph Wilk Mon, 24 Nov 2008 14:18:55 +0000 http://www.joesniff.co.uk/?p=386#comment-1015 Well LSA ultimately is a feature extraction technique, identifying relationships beyond mere structure. So its any application that would benefit from these extracted relationships. The obvious example is in search where it could help improve the quality of search results and related content. Projects like Lucene and Ferret make use of LSA. Here are some other applications where LSA has been applied: * Thesaurus * Finding synonyms * Categorization of documents * Improving OCR of Chinese * Identifying similarities in source code * Detecting certain Proteins Well LSA ultimately is a feature extraction technique, identifying relationships beyond mere structure. So its any application that would benefit from these extracted relationships. The obvious example is in search where it could help improve the quality of search results and related content. Projects like Lucene and Ferret make use of LSA.

Here are some other applications where LSA has been applied:
* Thesaurus
* Finding synonyms
* Categorization of documents
* Improving OCR of Chinese
* Identifying similarities in source code
* Detecting certain Proteins

]]> By: michaelr http://blog.josephwilk.net/ruby/latent-semantic-analysis-in-ruby.html/comment-page-1#comment-1010 michaelr Sun, 23 Nov 2008 21:31:32 +0000 http://www.joesniff.co.uk/?p=386#comment-1010 I wonder what use does it have? I wonder what use does it have?

]]> By: Glenn http://blog.josephwilk.net/ruby/latent-semantic-analysis-in-ruby.html/comment-page-1#comment-986 Glenn Sat, 22 Nov 2008 19:24:54 +0000 http://www.joesniff.co.uk/?p=386#comment-986 Great write up. I've been reading up on this for quite a while and looking to do a ruby implementation, the classifier gem seems to have become somewhat neglected. Have you tried it with much larger document sets (say a million)? I'm wondering what approaches are possible with such data in ruby without requiring massive amounts of available memory. Great write up. I’ve been reading up on this for quite a while and looking to do a ruby implementation, the classifier gem seems to have become somewhat neglected.

Have you tried it with much larger document sets (say a million)? I’m wondering what approaches are possible with such data in ruby without requiring massive amounts of available memory.

]]>