23th September 2024

RefutES

Automatic Generation of Counter Speech in Spanish

 Description

Nowadays, the use of social networks has become an everyday activity for many people, as they share their day-to-day lives, interact with other users and express their opinions and feelings on various topics. Despite this familiarity with social platforms, we often overlook the impact we as users have on these platforms and the impact social networks have on us. Moreover, the anonymity and freedom of expression of these platforms have incited the emergence of inappropriate behaviours in the virtual environment. Among these behaviours is the dissemination of hate messages among users who use their posts to belittle, insult or harm others based on personal characteristics such as gender, race, religion and ideology, among others. Unfortunately, this form of communication can have harmful consequences and generate negative psychological effects on users, especially in young people, such as anxiety, cyberbullying and even extreme cases of suicide.

The increase in the dissemination of offensive messages on social networks has led the United Nations to develop a Strategy and Plan of Action against Hate Speech, where hate speech is defined as "any type of verbal, written or behavioural communication that attacks or uses derogatory or discriminatory language against a person or group based on who they are, i.e. their religion, ethnicity, nationality, race, colour, descent, gender or other identity factor.

The strategy most commonly used so far to combat offensive messages is the removal of the message or temporary blocking of users from the social network. However, this strategy can cause a perception of censorship and limitation of freedom of expression in the users who write these messages. This perception causes resistance to attitude change in the users who write these messages. In addition, censorship does not address the underlying motives of hate crimes and misses the opportunity to educate users. For these reasons, a new strategy called counter-narrative generation, is explored. This strategy seeks the elaboration of a response that negates the offending message. The counter-speech can be argued to provide accurate and truthful information about the offending message (known as refutation), or non-argued, simply rejecting the idea of the offending message (known as rejection). With this counter-speech we try to challenge the stereotypes generated by the offensive message, offering an alternative and constructive perspective to foster empathy, understanding and tolerance among users, thus promoting a more inclusive and respectful online environment.

RefutES is a novel task on the refutation of hate-speech messages directed to different targets of offensive. In this task, participants must be able to generate a response to the offensive message in Spanish (Vallecillo-Rodríguez, Montejo-Raéz, y Martín-Valdivia 2023). The response should be reasoned, respectful, non-offensive, and contain information that is specific and truthful. For this first edition, the offended targets are: disabled, Jews, LGBT+, migrants, Muslims, people of colour, women and other groups.

If you want to participate in the RefutES@IberLEF2024 shared task, please fill this form. Once you are registered, you can ask any questions through the Google Group of the shared task RefutEs@IberLEF2024.

Participants will be required to submit their runs and their submissions will be evaluated on the test partitions for the corresponding corpora. To submit their runs, a form will be available on date. 

Important: The files attached to the form must be named like "{TeamName}-run{X}-{typeFile}.{format}", where: 

Participants will be required to submit their runs and are asked to describe their systems in paper submissions. We encourage participating teams to highlight the real contribution of their systems in identifying successful approaches along with failed attempts and findings on how to advance in more performant solutions. This description must contain the following details:

 Contact

If you have any specific question about the RefutES 2024 task, we may ask you to let us know through the Google Group RefutEs@IberLEF2024.

For any other questions that do not directly concern the shared task, please contact with Estrella Vallecillo-Rodríguez.

Estrella 

Vallecillo-Rodríguez

  University of Jaén 

mevallec@ujaen.es

Email

This work has been partially supported by Project CONSENSO (PID2021-122263OB-C21), Project MODERATES (TED2021-130145B-I00) and Project SocialTox (PDC2022-133146-C21) funded by MCIN/AEI/10.13039/501100011033 

RefutES at IberLEF 2024

SINAI Research Group

Twitter: @NLP_SINAI