Pages Menu
Categories Menu

Posted by on Jun 8, 2012 in Social Networks |

That would be the face of Twitter

That would be the face of Twitter

When I read the I remembered a friend telling me psychoanalyst social networking site was an excellent field to analyze patterns of user behavior, some funny, some controversial and many with negative and depressive tendencies. Seeing “the face of Twitter” recreated by this scientific test checked what I meant by its conclusions.

Professor Nello Cristianini Artificial Intelligence, University of Bristol, working on automated analysis of genomic data and content of the media to play on an animated face feelings of the users of the microblogging social network, and results are surprising, a cyclothymic mood constantly changing according to the spread or point of view applied to the treatment of news.

Using seven servers to analyze nearly a thousand daily news agencies, the system analyzed about 50 items per minute to accumulate a total daily equivalent of 200 books of 350 pages.

The same technique of analysis of this content is to form a pattern reduced to a simple story that contains a simple subject, verb and object. Initially, these figures were used to measure the behavior of the media during the U.S. presidential race.

When trying to do the same with Twitter were found with two levels of complexity, speed building tweets (estimated at 400 million per day) and the complexity of analyzing each one individually. After a time and a minimum flow of tweets, able to detect patterns.

To test this theory of mood and give a face to the social network, researchers analyzed about 500 million user-generated tweets British in the last three years. Each state of mind known assigned one or more words and phrases associated.

The results showed exaggerated increases in some positive moods for dates such as Christmas, New Year and Valentine’s Day, which also revealed that users generally positive predominate. But the most surprising Furon bleak and negative reactions to announcements of October 2010 the reduction in public spending (a factor that directly impacts the job losses) and during the riots of August 2011, when cities were burned by youth groups.

After obtaining these results, the investigators decided to make them simpler to read incorporating an animated face that change their expression patterns according to the most intense during the content analysis. Is it a tool for future leaders in the population can test before taking decisions? How much could benefit or harm us as individuals who are familiar with the general mood through the analysis of these patterns of behavior? Ah, yes, the face of Twitter is this:

Link: Revealed: The face of Twitter (geeksaresexy)

Tags: , , , ,