e-space
Manchester Metropolitan University's Research Repository

    Enhancing Educational Equity: A Native Language Identification Approach for Tailoring Linguistic Support and Inclusive Curricula

    Sarwar, Raheem, Teh, Pin Shen ORCID logoORCID: https://orcid.org/0000-0002-0607-2617, Fayyaz, Muhammad Asad Bilal ORCID logoORCID: https://orcid.org/0000-0002-1794-3000, Sabah, Fahad, Hassan, Muhammad Umair and Hassan, Syed Mustafa (2025) Enhancing Educational Equity: A Native Language Identification Approach for Tailoring Linguistic Support and Inclusive Curricula. In: RIIFORUM 202 (Springer Proceedings in Complexity series), pp. 213-239. Presented at Research and Innovation Forum 2024, 10 April - 12 April 2024, Ravello, Italy.

    [img]
    Preview
    Accepted Version
    Available under License In Copyright.

    Download (587kB) | Preview

    Abstract

    This paper introduces a solution for native language identification (NLI) on texts written in English, French, and German, offering diverse applications in education. NLI research provides insights into students’ linguistic backgrounds, enabling educational institutions to customize materials, assignments, and assessments for individual needs. By identifying students who may benefit from language support, institutions can develop targeted language-specific curricula. Understanding students’ native languages also helps educators incorporate relevant cultural references and create a more inclusive learning experience. Furthermore, NLI can guide the creation of targeted training for educators, equipping them with strategies to address language-specific challenges and foster effective communication in diverse classrooms. The proposed NLI approach analyzes text samples in non-native languages, providing a robust solution that captures language usage and production patterns across documents and languages. The approach is supported by three new corpora in German, French, and English and has shown superior performance compared to existing state-of-the-art NLI methods and pre-trained language models like DistilBERT, mBERT, multilingual DeBERTa, and XLM-RoBERTa. This enhanced NLI model contributes to improved cross-cultural communication within academic communities, fostering a more inclusive and supportive environment for students and faculty alike.

    Impact and Reach

    Statistics

    Activity Overview
    6 month trend
    3Downloads
    6 month trend
    8Hits

    Additional statistics for this dataset are available via IRStats2.

    Altmetric

    Repository staff only

    Edit record Edit record