Dr. Eduardo Silva Alvarado, director of the Fundación Universitaria Iberoamericana (Ibero-American University Foundation, FUNIBER) in Guatemala, is participating in a study that has developed an innovative model for the automatic detection of hate speech on social media. This advance addresses the challenges inherent in identifying offensive content in multilingual and multimodal digital environments.
The rise of social media has facilitated global communication, allowing users to share information in various formats and languages. However, this expansion has also led to the spread of hate speech, which can manifest itself in text, images, and videos, and in multiple languages. Detecting this type of content is essential for maintaining safe and respectful digital environments.
Traditionally, hate speech detection systems have focused on monolingual analysis and a single type of content, mainly text. These approaches have significant limitations when faced with the linguistic diversity and multimodal nature of social media posts. The lack of tools capable of addressing these complexities has hindered the effective identification of offensive content in multilingual and multimodal contexts.
The study in question introduces a model that combines advanced natural language processing and image analysis techniques to detect hate speech in multiple languages and content formats. This integrated approach enables more accurate and efficient identification of offensive content in diverse digital environments.
To develop this model, the team collected a multilingual and multimodal dataset that includes text, images, and videos with content labeled as hate speech. Deep learning techniques were used to train the model to identify patterns associated with hate speech in different languages and content formats.
Relevant results
The results of the study show that the proposed model outperforms traditional approaches in detecting hate speech in multilingual and multimodal contexts. The model’s ability to simultaneously analyze text and images in multiple languages significantly improves accuracy and efficiency in identifying offensive content.
This advance has important implications for content moderation on social media platforms, enabling a faster and more effective response to the spread of hate speech. In addition, the model can be adapted to different cultural and linguistic contexts, making it a valuable tool for promoting safer and more inclusive digital environments.
If you would like to learn more about this study, click here.
To read more research, check out the UNEATLANTICO repository.
The Ibero-American University Foundation (FUNIBER) offers scholarships to study the Master’s Degree in Strategic Management in Information Technology. This program is designed for professionals who seek to understand and master the strategic IT process as part of business development, addressing topics such as organizational change management, business innovation, and technology project management. Don’t miss this opportunity to boost your career in IT. Enroll now in this master’s program and become a leader in the digital transformation of organizations.