News for hackers. Stuff that matters.

A comment on Show HN: Comments by Top HN Posters Analysed by IBM's Watson User Modelling API.

Re:Show HN: Comments by Top HN Posters Analysed by IBM's Wat... by agilebyte on Saturday December 13, 2014 @04:07PM (#8746416)
I think that IBM Watson expects a general text no? I admit to having done some personality classification on OKC profiles (around 200) and found that disproportionately to the average population, there were a lot of people that were classified as caring, helpful, social, popular individuals.

It fits the medium. Just like one would expect HN comments to be full of more "head-y" discussions.

[ Parent | on HN ]
Re:Show HN: Comments by Top HN Posters Analysed by IBM's Wat... by kolinko on Saturday December 13, 2014 @11:38PM (#8747457)
I didn't find such expectation in the docs, and someone else posted this, where it says that researchers from IBM were using their tech to analyse tweets:

http://www.economist.com/blogs/economist-explains/2013/05/ec...

I'd assume that this might be the same algorithm, but I'm not really sure of that.

[ Parent | on HN ]