My Photo

« Is Authority Migrating? | Main | How to Improve Twitter Search »

March 13, 2009


Nigel Legg

I think automatic sentiment flagging is still very much in it's infancy, and I haven't come across any system that I can be sure is reliable. I've got a hunch that the requirement for getting sentiment tagging right for social media will lead to some major advances in computer understanding of language.

Kevin Murphy - Digital Experience Director

Thanks for kicking off this discussion. I'm one of the folks at Waggener who worked on this. You're absolutely right on seniment analyis being in its infancy. On this project we particlarly struggled with intepreting sentiment with such a small data set to work with (140 or fewer characters). We do hope that people find this a good monitoring tool to look at aggregate sentiment and identify interesting discussion points trends and related topics.

Kevin Murphy

Brendan Cooper


I checked out the sentimenting for the Jade Goody search today. At first it showed overwhelmingly negative, which I think could be accounted for instances where "Sad that Jade died" would be interpreted as negative even though, really it was positive.

However, at the end of the day (literally), it said sentiment was mixed - which did in fact reflect the opinion of articles and interviews on the subject. So, did it work? Yes, this time. I think it's an interesting approach and will be testing it out on other issues.


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