My Photo

« Half-lifing the Planet | Main | Attenex Update »

March 01, 2006


Niall Cook

Matthew, you hit on the single biggest problem in automated sentiment monitoring - the fact that sentiment is relative to the views of the reader. What is good news for one person/company/country may be bad news for another. So how can an automated system ever get it right?

Matthew Hurst

Niall - I'm not sure this is the biggest problem. I suspect that the real issue is how you interpret and report the results. If I read 'I liked Super Size Me', the author expresses the appreciation of the movie. To McDonalds, the company being criticised, this is not a good thing. However, by reporting the number of people who like the movie (positive sentiment), they can judge the *negative* impact to their brand.

The other aspect of mining textual data for emotional content (affect analysis) aims to capture the emotional state of the author - this is actually easier to do independent of the pov problem.

The comments to this entry are closed.

Twitter Updates

    follow me on Twitter

    March 2016

    Sun Mon Tue Wed Thu Fri Sat
        1 2 3 4 5
    6 7 8 9 10 11 12
    13 14 15 16 17 18 19
    20 21 22 23 24 25 26
    27 28 29 30 31    


    Blog powered by Typepad