Abstract
This study examines the development of sociocultural competence among international linguistics students learning English as a second foreign language through global simulation supported by Natural Language Processing (NLP) and speech recognition technologies. Sociocultural competence is defined as the capacity to communicate effectively and appropriately across cultural contexts and is assessed through indicators such as intercultural awareness, context-sensitive language use, communicative adaptability, and pragmatically appropriate choices. The instructional intervention integrates NLP techniques including text classification, semantic similarity analysis, and language modeling to generate real-time feedback on the cultural appropriateness of learners’ responses. Speech recognition technology is employed to analyze oral production and to support immediate corrective guidance across varying proficiency levels. Simulation tasks are structured around authentic intercultural dilemmas and role-based interactions, with methodological precautions taken to reduce cultural bias and stereotyping. A quasi-experimental design (n = 32) compares technology-enhanced simulation with conventional instruction. Data are collected through validated assessment rubrics, pre- and post-test measures, and interaction logs to quantify changes in sociocultural performance. Findings indicate measurable improvement in students’ intercultural communicative competence in the experimental group, suggesting that AI-supported simulation can strengthen both linguistic accuracy and sociocultural appropriateness beyond traditional classroom practices. The study also addresses ethical considerations related to data privacy and speech processing. Although limited by sample size and institutional scope, the research offers a replicable framework for integrating computational tools into linguistics education and highlights directions for broader, cross-contextual implementation.