(Subscribe to our Today's Cache newsletter for a quick snapshot of top 5 tech stories. Click here to subscribe for free.)
Researchers at the University of Vermont have developed a tool, called Hedonometer, to gauge the level of happiness of large populations online in real-time. It measures the online sentiment and mental health situation of people posting content.
The tool uses data from social media to measure how people present themselves to the outside world. At present, it uses Twitter as a source, but will soon expand to other sites as well, Hedonometer's website noted.
The name Hedonometer comes from the term coined in the 1800s, which refers to a psychophysical machine used to register the height of pleasure experienced by an individual. The tool's algorithm is described in a study titled 'Temporal Patterns of Happiness and Information in a Global-Scale Social Network: Hedonometrics and Twitter'.
To quantify the level of happiness, the team merged 5,000 frequently used words from four sources — Google Books, New York Times articles, Music lyrics and Twitter's messages, to produce a list of 10,000 unique words.
Using Amazon's Mechanical Turk service, each of these words was scored on a nine-point scale that ranges from sadness to happiness. The team updates the list to include words as they trend online, it stated. The list is available for download.
The hedonometer currently measures Twitter's API. The stream reflects 10% random sampling of the roughly 500 million messages posted to the service daily. Words in messages are thrown into a large bag containing roughly 200 million words per day. This bag is then assigned a happiness score based on the average happiness score of the words contained within.
Also read | Why an online search for Shah Rukh Khan could put you at risk
The team chose Twitter as a source as it provides a massive pool of real-time data. Its user base also represents a strong social signal. The team is also involved in projects that uses the hedonometer to characterise happiness variations with respect to geography, network topology, demographics, and socio-economic data.
The team puts out graphs to portray public sentiment. For instance, the day of Osama Bin Laden's death has a low score on the scale as the hedonometer captured tweets having negative language like 'death', 'war', and 'terrorist'.
Lately, the protests against police brutality in May scored 5.6 on the hedonometer, making it the saddest day recorded by the tool so far. The tool picked up words like 'anger', 'no', 'violence', 'terrorist', and 'racist' to assign the score.
The team hopes to incorporate other emotions like anger, fear, and surprise into the tool soon.
