Fake News Detection and Dataset Analysis in Ecuador’s 2025 Elections: A Comparative Study of Machine Learning Models with X Social Media Platform Data

Milagros Delgado; Britney Robalino Ramírez; Brithany Suárez; Aisha Gómez; Diego Almeida-Galarraga; Paulina Vizcaíno-Imacaña; Andrés Tirado-Espín

doi:10.2991/978-94-6239-803-0_2

<Previous Article In Volume

Next Article In Volume>

Fake News Detection and Dataset Analysis in Ecuador’s 2025 Elections: A Comparative Study of Machine Learning Models with X Social Media Platform Data

Authors

Milagros Delgado¹, Britney Robalino Ramírez¹, Brithany Suárez¹, Aisha Gómez², Diego Almeida-Galarraga³, Paulina Vizcaíno-Imacaña⁴, Andrés Tirado-Espín²^{, *}

¹School of Physical Sciences and Nanotechnology, Yachay Tech University, Hacienda San José s/n, San Miguel de Urcuquí, 100119, Ecuador

²School of Mathematical and Computational Sciences, Yachay Tech University, Hacienda San José s/n, San Miguel de Urcuquí, 100119, Ecuador

³School of Biological Sciences and Engineering, Yachay Tech University, San Miguel de Urcuquí, 100119, Ecuador

⁴Facultad de Ingenierías Digitales y Tecnologías Emergentes, Universidad Internacional del Ecuador (UIDE), Quito, 170411, Ecuador

^*Corresponding author. Email: ctirado@yachaytech.edu.ec

Corresponding Author

Andrés Tirado-Espín

Available Online 16 October 2026.

DOI: 10.2991/978-94-6239-803-0_2 How to use a DOI?
Keywords: Social media; medical pseudoscience; digital misinformation; public health; media literacy
Abstract: In a fragmented and polarized society, recognizing fake content is essential for safeguarding democratic processes. This problem was particularly relevant in the context of Ecuador’s 2025 Presidential Elections, where misinformation can significantly influence public opinion. This study evaluates the effectiveness of large language models (LLMs) in detecting fake news related to this electoral process on X (formerly Twitter). A dataset of 600 news samples was compiled. Three traditional machine learning models and five pre-trained Spanish-language LLMs were fine-tuned and compared. The analysis revealed that’Daniel Noboa’ was the most frequent term in both real and fake news, reflecting a strong social media presence during the campaign. Among the evaluated models, RoBERTuito achieved the best performance, reaching 90% accuracy. The results confirm the potential of transformer-based models for fake news detection in Spanish-speaking political contexts and highlight their contribution to strengthening public discourse and promoting electoral transparency.
Open Access: Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

<Previous Article In Volume

Next Article In Volume>

Volume Title: Proceedings of the VI International Conference on Communication and Applied Technologies 2026 (ICOMTA 2026)
Series: Atlantis Highlights in Social Sciences, Education and Humanities
Publication Date: 16 October 2026
ISBN: 978-94-6239-803-0
ISSN: 2667-128X
DOI: 10.2991/978-94-6239-803-0_2 How to use a DOI?
Open Access: Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Cite this article

ris enw bib

TY  - CONF
AU  - Milagros Delgado
AU  - Britney Robalino Ramírez
AU  - Brithany Suárez
AU  - Aisha Gómez
AU  - Diego Almeida-Galarraga
AU  - Paulina Vizcaíno-Imacaña
AU  - Andrés Tirado-Espín
PY  - 2026
DA  - 2026/10/16
TI  - Fake News Detection and Dataset Analysis in Ecuador’s 2025 Elections: A Comparative Study of Machine Learning Models with X Social Media Platform Data
BT  - Proceedings of the VI International Conference on Communication and Applied Technologies 2026 (ICOMTA 2026)
PB  - Atlantis Press
SP  - 3
EP  - 13
SN  - 2667-128X
UR  - https://doi.org/10.2991/978-94-6239-803-0_2
DO  - 10.2991/978-94-6239-803-0_2
ID  - Delgado2026
ER  -

download .riscopy to clipboard