The mood of the Internet
Recovered from the New Year? Moodgrapher detected, not unexpectedly, a large peak in the 'drunk' mood on New Years Day. After that the usual weekend peak in 'drunk' remained below the usual level. Time is on the x-axis and the y-axis shows the deviation compared to the usual observations (left) and the fraction of the messages with the 'drunk’ mood (right).
Internet diaries, blogs, are an increasingly common phenomenon on the Internet. The opinions and experiences of tens of millions of people can be followed every day. Researchers from the Pionier project 'Computing with Meaning' think that subjective aspects of blogs, such as moods, play in important role in determining which people look at and read the blogs. So Gilad Mishne, Krisztian Balog and project leader Maarten de Rijke have developed MoodViews. MoodViews follows the moods of 2 million bloggers from around the world. Each day the programs pick up about 150,000 blog messages.
The clearly measurable responses to worldwide events suggest that these instruments pick up the global mood. And not only those of bloggers. De Rijke has already been approached by journalists, marketing specialists, psychologists and bankers. The Moodteam has something interesting to offer to all of these groups.
At present MoodViews has three components. Moodgrapher monitors the global mood of bloggers. These data are converted into overview graphs. Moodteller uses language technology to predict the mood on the Web. The latest component, Moodsignals, records unusual peaks in mood levels. After this it uses search engine technology to find an explanation for these mood swings.
Various other components of MoodViews are still under development. For example, Moodspotter will be launched this summer. This instrument will be able to search the moods associated with certain persons, locations or products.
Maarten de Rijke's research is part of his Pionier project 'Computing with Meaning'. The Pionier programme of NWO ran from 1989 to 2002 and was aimed at highly experienced researchers of a professorial level.
Link: http://moodviews.com
Source: NWO
At present MoodViews has three components. Moodgrapher monitors the global mood of bloggers. These data are converted into overview graphs. Moodteller uses language technology to predict the mood on the Web. The latest component, Moodsignals, records unusual peaks in mood levels. After this it uses search engine technology to find an explanation for these mood swings.
Various other components of MoodViews are still under development. For example, Moodspotter will be launched this summer. This instrument will be able to search the moods associated with certain persons, locations or products.
Maarten de Rijke's research is part of his Pionier project 'Computing with Meaning'. The Pionier programme of NWO ran from 1989 to 2002 and was aimed at highly experienced researchers of a professorial level.
Link: http://moodviews.com
Source: NWO
» Next Article in Technology: In Brief: Softbank head wants to head Vodafone Japan

Rating: 4.9
Bookmark
Save as PDF
Print
Email
Blog It
Stumble It!


PhysOrg Forum
Video
Editorials
Free Magazines
Free White Papers
Newsletter
Advanced Search
Goto Archive
Suggest a story idea
Send feedback