Although in a spoken conversation, especially in person, identifying sarcasm can be simple, and does not involve many complications, in the case of social networks, instant messaging or email, it is evident that the thing is a little more hard. For this reason, a few days ago we knew that researchers from the University of Central Florida had developed a sarcasm detector in social networks , through the use of artificial intelligence (AI).
Now, we have learned that researchers from the University of Neuchâtel, in Switzerland, and the Norwegian University of Science and Technology, in Norway, have discovered the possibility of tracking mood trends in large groups of people by attending to their publications. on Twitter .
In an article published in the specialized journal Royal Society Open Science , the researchers describe their analysis of large amounts of Twitter data from users who tweet regularly from different densely populated counties in the United States.
Different studies published previously have already shown that people’s moods could be linked, in some way, to different factors such as drowsiness and hunger, as well as their circadian rhythm. All of these elements would have a huge impact on how they feel.
Now, scientists have wondered if people could also experience cyclical moods based on other factors, such as having the “obligation” to go to work.
Analyzing more than 25 million tweets …
In order to find an answer, they analyzed 25 million tweets written by users in some highly populated counties in the United States, and theorized that certain tweets, such as “I want to do this or that” or “I do not want to do this other thing” could offer clues to how happy or unhappy a person is at different times of the day or week. Likewise, they also noted and took into account which emojis users used to express how they felt.
Once analyzed, the researchers were able to find some interesting patterns . For example, they noticed that, in general, users seemed to be less happy on Sunday afternoons, staying with this feeling until the following Thursday, then maintaining a higher level of happiness until, again, the following Sunday arrived.
This pattern led the researchers to conclude that people are not very happy about going to work , as they tend to experience a drop in mood before their weekends end, long before Monday arrives. And subsequently, the mood picks up again as the weekend approaches.
Similarly, the group of users who seemed to feel depressed as work began each day began to feel better as it was time to go home.
It is, as we can see, a very interesting study, because it offers the possibility for scientists to analyze and discover what the mood of hundreds of thousands of people is like through the messages they publish on social networks throughout days or weeks. And, in addition, all this in a totally anonymous way.
