How do things for Rupert and James Murdoch in Twitter view? How not too well, you might think. According to Twitter sentiment app engine moved their appearance most negative analysis - how you below, the results of the search for the Hashtag # hackgate and "Murdoch will see".
The idea of sentiment analysis is that you take a particular piece of text and find out, from the words used, and how they are put together, whether they express positive or negative feelings are about a certain phrase. It is a complex but promising research - especially for brands that know what to say to people about it on social networks. "They love us or they hate us on Twitter/Facebook/where?" That's what the marketing people need to know.
So here we go for "Murdoch", given the fact that they were pretty busy on Tuesday... Red is negative, green is positive.
Twitter sentiment analysis for 'Murdoch': mostly negative
And now for "# hackgate" on Twitter:
Sentiment analysis for "# hackgate" Twitter: negative
But.... Michael Brunton-Spall, one of the developers of the guardian, emphasizes the weakness in the analysis. He did a search on the phrase "News of the World" and concludes that it is "basically useless":
He says: "from the first page-"
His conclusion: "Tweets are too small to analyze semantically, and the surrounding context is not specified, so it pretty much is an impossible task."
But he has good news: "nice [Devexpress] Although."
Then again.... If you are looking for Wendi Deng, a one-woman's rapid reaction force...
Twitter sentiment analysis for 'Wendi Deng': mostly positive
Hmm, mostly positive... That seems to reflect many peoples reactions to their intervention. To who else a Twitter sentiment analysis engine offer?
The idea of sentiment analysis is that you take a particular piece of text and find out, from the words used, and how they are put together, whether they express positive or negative feelings are about a certain phrase. It is a complex but promising research - especially for brands that know what to say to people about it on social networks. "They love us or they hate us on Twitter/Facebook/where?" That's what the marketing people need to know.
So here we go for "Murdoch", given the fact that they were pretty busy on Tuesday... Red is negative, green is positive.
Twitter sentiment analysis for 'Murdoch': mostly negative
And now for "# hackgate" on Twitter:
Sentiment analysis for "# hackgate" Twitter: negative
But.... Michael Brunton-Spall, one of the developers of the guardian, emphasizes the weakness in the analysis. He did a search on the phrase "News of the World" and concludes that it is "basically useless":
He says: "from the first page-"
' I really hope that news of the world-9/11 victims not hack' phones. I would be very excited. ""is negative."
"' Mrs. Murdoch joined the Pieman ass." Bad news of the world has also disappeared. What a headline.'"is negative."
"'@ _staralfur_ for the news of the world with headlines such as the works should.." Oh wait: p'"is positive."
His conclusion: "Tweets are too small to analyze semantically, and the surrounding context is not specified, so it pretty much is an impossible task."
But he has good news: "nice [Devexpress] Although."
Then again.... If you are looking for Wendi Deng, a one-woman's rapid reaction force...
Twitter sentiment analysis for 'Wendi Deng': mostly positive
Hmm, mostly positive... That seems to reflect many peoples reactions to their intervention. To who else a Twitter sentiment analysis engine offer?
No comments:
Post a Comment