The article seemed really weak on details. I have done some work on text analysis from a couple of different major camps so I can read between the lines a bit,

The article seemed really weak on details. I have done some work on text analysis from a couple of different major camps so I can read between the lines a bit, but for the HN audience it would be nice to point to a more technical write-up, including links to academic papers and patent filings that this is built on. Is that kind of write-up available? A comparison with the approach IBM is taking with Watson might also be informative. Thanks