Who are the most attacked on networks? Study identifies the target of hate

The analysis, based on almost 10 million messages collected from social networks such as X, Facebook and Spanish digital media websites, reveals that 70% of hate messages are directed mainly at politicians, women, immigrants and the LGTBI+ community. With messages that not only encourage hostility towards vulnerable groups, but, in many cases, “seek to impose polarizing narratives on public opinion.”

A monitor that classifies messages according to their toxicity

Hatemedia has developed Hate Monitor, an algorithm-based tool that identifies and classifies messages according to their level of toxicity. This monitor not only detects hate speech in real time, but also offers an early warning system to prevent harmful messages from going viral. Its database, with more than 7,000 terms in Spanish, is “the most advanced in the detection of hate in the Spanish language”, according to the UNIR, and has an accuracy of 90%.

It is also capable of classifying messages according to their type and intensity. Regarding their theme, 35% promote political hatredanother 35% reflects xenophobic, misogynistic or sexual orientation hatredand the remaining 30% is of a generalized naturethe latter being more frequent on social networks. Regarding intensity, 63% of the messages generate a climate of media hostility, while 37% promote explicit violence, including threats.

“This is an innovative contribution,” says Julio Montero, co-director of the Project with Elías Said-Hung, “because until now the detection models are focused on identifying whether or not hate exists, but they do not analyze the types of hate.” specific attacks against certain vulnerable groups (for misogynistic, racist, xenophobic, sexual, ideological reasons, etc.) nor the intensity.

The idea of ​​the project is that this tool be accessible to all mediaprofessionals in the sector and public and social agents, to help detect and implement active strategies that prevent the presence of hate expressions in the digital media environment.

This interdisciplinary project brings together experts in linguistics, computer science and social sciences from various Spanish universities. Its goal is not only to combat digital hate, but also educate and raise awareness about its impact. Initiatives like these represent a crucial step to address a challenge that, as researchers warn, is reaching worrying levels in our society.

The profile of the “haters”

The study also offers a preliminary view of the profile of the so-called “haters”, these individuals who take advantage of social networks and the comments section of the news media to promote hate messages. According to the findings, these users They often act in a coordinated mannerwith strategies designed to repeat and amplify ideas, seeking to influence the public narrative.

Its main objectives They include political actors, journalists and the media themselves, whom they attack through expressions of hate that are based on racial, ethnic, political, misogynistic or sexual reasons. This behavior not only affects the vulnerable groups that these actors represent, but also attempts to undermine their role as key figures in the democratic debate and the transmission of information.

The researchers emphasize that hate on networks not only threatens the groups attacked, but also the democratic pillars. According to Julio Montero, “Hate is a cancer of democracy” that deteriorates mutual respect and rational dialogue, essential to resolve differences in a plural society. For his part, Elías Said emphasizes that although hatred is manifested in comments published in the media, it does not mean that they promote it as institutions.