Replies to This Discussion

Permalink Reply by Vincent Granville on June 11, 2010 at 12:10pm

Hi Greg -- See the discussion on "Classification of items into product categories", at http://www.analyticbridge.com/forum/topics/classification-of-items-....

• ▶ Reply

Permalink Reply by Greg Sundberg on June 11, 2010 at 12:45pm

Thanks Vincent. Yes, I had reviewed that but I'm looking for a tool currently deployed/available. Much appreciated.

• ▶ Reply

Permalink Reply by Ralph Winters on June 12, 2010 at 6:56pm

I can think of 3 Open Source products which use Naive Bayes algorithm. Weka, Rapidminer, and Gate.

-Ralph Winters

• ▶ Reply

Permalink Reply by Greg Sundberg on June 12, 2010 at 9:40pm

Ralph,
Thanks for the information. Very helpful.

In the course of researching your suggestions I came upon this list of text mining, natural language processing, and classification tools. Listing it here as a breadcrumb for others in the future.

Thanks again,
Greg

• ▶ Reply

Permalink Reply by Vincent Granville on June 20, 2010 at 11:12am

On a different note, I am looking for a very large list of products, something like 1 million product names (or even much more), to create PPC keyword campaigns. Product names are usually good keywords to purchase (depending on your goal).

Greg, do you know where I could find very large product lists?

Thank you,
Vincent

• ▶ Reply

Permalink Reply by lisa jenkins on July 25, 2012 at 9:55am

has anyone actually used weka on a large dataset.  It seems to throw outOfMemory exceptions (aka heap) no matter how much memory your machine has and  regardless of how much heap you specify for it from the command line.  It seems the heap settings do not flow to things like the evaluation classes.  Using it has been a very frustrating experience.

Same is true for RWeka....

• ▶ Reply