To detect presumed suicidal ideas in texts from users of social networks such as Facebook and Twitter, a multidisciplinary group of scientists from UNAM is advancing in a computational linguistics project.
The research is led by Gerardo Sierra Martínez, head of the Linguistic engineering group of the Institute of Engineering (II), and Patricia Andrade Palos, Postgraduate academic of the Faculty of Psychology (FP) of the UNAM.
One of the emerging ways of expressing an intention to commit this act, said Gerardo Sierra, is through the platforms and analogous spaces that exist on the Internet. It expresses itself verbatim on these sites through discussion and, in the worst case, promotion.
For this reason, he continued, it is necessary to know the dynamics of its expression that are typical of these virtual environments and use methods such as language analysis, in order to develop detection tools that contribute to preventive work.
The project seeks to find linguistic characteristics that are identified and processed to detect risk, which would allow detecting people who hypothetically wish to attack their person.
A count and comparison of the lexicon was carried out in groups of Facebook and Twitter users (they are confidential, therefore their identity is ignored, because their profiles were not accessed) against random texts of other topics..
The establishment of a linguistic difference that exists between the people who point out some presumed risk, and the one who speaks of any other common thing was achieved.
“How was it achieved? Through a count of words that were grouped into different linguistic and psychological categories; among these, there is the one that users at risk talk about themselves, always in the first person, they do not use the plural, nor the ‘we,’ or ‘you,'” he said.
“I dropped out of school, pretended to have a job and wrote a suicide note”: the risks of video game addiction
The sentences with some presumed suicidal ideation may contain: “I feel like this”; “I am thinking”; “Why is this happening to me?; “It has happened to me…”. Concepts such as “crying”, “despair”, “loneliness”, “frustration”, “depressed”, “pessimistic”, are also integrated.
Likewise, categories of words that show anxiety, anguish, sadness or death, but are inescapably accompanied by the “I”, indicated the head of Linguistic Engineering of the II of the UNAM.
“In short, three sets of different texts were analyzed, whose content was about depression and suicide, on the one hand; and on the other, about random topics. The analysis between these yielded convincing results that there are significant linguistic differences that are a sign of suicide risk,” said Gerardo Sierra.
The results of the project, explained the researcher, are unprecedented for the country and for the Mexican Spanish language.
However, it is necessary to continue with research that confirms and expands the data of this first approach to the phenomenon, in order to have conclusive elements of the use of language for the detection of cases at risk.