Analysis of Lexical Competence in Spanish as a Foreign Language (ELE) with Gemini
AI-Based Automated Feedback
DOI:
https://doi.org/10.1590/1983-3652.2026.64168Keywords:
Spanish as a foreign language, Written expression, Vocabulary, Natural language processing, Artificial intelligenceAbstract
Vocabulary is an essential component in the development of written proficiency in foreign languages. This study examines the lexical ability of a group of students learning Spanish as a Foreign Language (ELE) at a language centre within a Spanish university. The aim of the study is to determine whether lexical knowledge enables students to be assessed according to the CEFR. To this end, the Gemini language model is used; it is asked to assess the students’ proficiency levels after being provided with lexical information from their texts extracted using the Textflow tool, and is also asked to provide feedback, with the aim of guiding students towards reaching the next level. Finally, Gemini’s ability to level students correctly is evaluated, and the educational quality of its feedback is analysed. In short, this study aims to assess the potential of large language models (LLMs) as a support for teaching and self-directed learning.
Downloads
References
ANTCONC. AntConc. [S. l.]: [s. n.], 2022. Disponible en: https://www.laurenceanthony.net/software/antconc/. Acceso en: enero 2026.
CANTERO ROMERO, María Victoria; MARTÍN-VALDIVIA, María Teresa; ORTIZ-COLÓN, Ana María y JIMÉNEZ-ZAFRA, Salud María. Medición de la habilidad escrita en español como lengua extranjera con inteligencia artificial generativa. RIED – Revista Iberoamericana de Educación a Distancia, Madrid, volumen 29, número 1, páginas 353-379, 2026. DOI: 10.5944/ried.45486.
CANTERO ROMERO, María Victoria; VALLECILLO RODRÍGUEZ, María Estrella; ORTIZ-COLÓN, Ana María y JIMÉNEZ-ZAFRA, Salud María. GPT-3.5 for ELE leveling in a university environment: a comparative study of zero-shot learning and fine-tuning. Procesamiento del Lenguaje Natural, Alicante, número 75, páginas 53-63, 2025.
CONSOLINI, Carla H. y KYLE, Kristopher. The relationship between L2 Spanish proficiency and features of written lexical and lexicogrammatical use. Applied Linguistics, Oxford, volumen 45, número 6, páginas 992-1009, 2024. DOI: 10.1093/applin/amae032.
CUBEROS VICENTE, María del Rocío. Indicadores léxicos de calidad textual en español nativo y no nativo. 2019. Tesis (Doctorado en Didáctica de las Ciencias, las Lenguas, las Artes y las Humanidades) – Universidad de Granada, Granada.
DE LA TORRE GARCÍA, Nuria. La complejidad lingüística en los niveles de competencia del MCER: el caso de la variedad verbal en la expresión escrita en ELE. Revista Nebrija de Lingüística Aplicada a la Enseñanza de las Lenguas, Madrid, número 20, 2016.
DEGRAEUWE, Jasper y GOETHALS, Patrick. LexComSpaL2: a lexical complexity corpus for Spanish as a foreign language. En: PROCEEDINGS of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). Torino: ELRA; ICCL, 2024. páginas 10432-10447. International Conference on Language Resources and Evaluation (LREC-COLING), 2024.
GIRAY, Louie. Prompt engineering with ChatGPT: a guide for academic writers. Annals of Biomedical Engineering, New York, volumen 51, número 12, páginas 2629-2633, 2023.
GOOGLE. Gemini API: Gemini models. [S. l.]: [s. n.], 2025. Disponible en: https://ai.google.dev/gemini-api/docs/models. Acceso en: oct. 2025.
HU, Xiaolin. Predicting CEFR levels in L2 oral speech, based on lexical and syntactic complexity. Asia Pacific Journal of Corpus Research, Incheon, volumen 2, número 1, páginas 35-45, 2021. DOI: 10.22925/apjcr.2021.2.1.35.
INSTITUTO CERVANTES. Plan curricular del Instituto Cervantes: niveles de referencia para el español. Madrid: [s. n.], 2006. Disponible en: https://cvc.cervantes.es/ensenanza/biblioteca_ele/plan_curricular/default.htm. Acceso en: oct. 2025.
JIMÉNEZ PÉREZ, Elena del Pilar; LÓPEZ MARTÍNEZ, Olivia y CORBALÁN BERNÁ, Javier. LECOLE algorithm: readability, readability, and comprehensibility in Spanish. XLinguae, Nitra, volumen 17, número 1, páginas 12-25, 2024. DOI: 10.18355/XL.2024.17.01.12.
KYLE, Kristopher y CROSSLEY, Scott. The relationship between lexical sophistication and independent and source-based writing. Journal of Second Language Writing, Amsterdam, volumen 34, páginas 12-24, 2016. DOI: 10.1016/j.jslw.2016.10.003.
LAUFER, Batia y NATION, Paul. Vocabulary size and use: lexical richness in L2 written production. Applied Linguistics, Oxford, volumen 16, número 3, páginas 307-322, 1995. DOI: 10.1093/applin/16.3.307.
ROBLES GARCÍA, Pablo. Competencia léxica y oportunidades de aprendizaje léxico en el aula de lenguas: una aproximación teórica. Lengua y Habla, Mérida, número 23, páginas 157-182, 2019.
RODRIGO YUSTE, Álvaro; MORENO ÁLVAREZ, Sergio; PÉREZ GARCÍA-PLAZA, Alberto; PEÑAS PADILLA, Anselmo; AGERRI, Rodrigo; FRUNS-JIMÉNEZ, Javier y SORIA-PASTOR, Inés. Overview of PROFE at IberLEF 2025: language proficiency evaluation. Procesamiento del Lenguaje Natural, Alicante, número 75, páginas 487-497, 2025.
SISTEMAS INTELIGENTES DE ACCESO A LA INFORMACIÓN (SINAI). Textflow. [S. l.]: [s. n.], 2025. Disponible en: https://github.com/sinai-uja/textflow. Acceso en: oct. 2025.
TEXT INSPECTOR. Text Inspector. [S. l.]: [s. n.], 2026. Disponible en: https://textinspector.com. Acceso en: enero 2026.
VINE JARA, Ana; FUENTES RIFFO, Karina; NEIRA MARTÍNEZ, Angie y POZA MOLINA, Carlos. Riqueza léxica en textos escritos por aprendientes de ELE: una contribución a las nociones específicas del Plan Curricular del Instituto Cervantes. Nueva Revista del Pacífico, Valparaíso, número 75, páginas 215-240, 2021. DOI: 10.4067/S0719-51762021000200201.
WILKINS, David A. Linguistics and language teaching. 1. edición. London: Edward Arnold, 1972.
Downloads
Published
Data Availability Statement
Los datos de investigación solo están disponibles previa solicitud.
Issue
Section
License
Copyright (c) 2026 María-Victoria Cantero-Romero, Ana María Ortiz-Colón, Salud María Jiménez-Zafra

This work is licensed under a Creative Commons Attribution 4.0 International License.
This is an open access article that allows unrestricted use, distribution and reproduction in any medium as long as the original article is properly cited.








