Acceso usuarios
Zhaoying Wan
Linguistic heritage is closely connected to social identity, yet many minority languages remain underrepresented in digital communication systems. As digital tools increasingly mediate access to information, education, and public communication, the lack of natural language processing resources for minority languages creates new forms of linguistic and digital inequality. This issue is particularly significant in multilingual societies, where uneven technological support can reinforce existing disparities in language visibility and communicative participation. This study attempts to answer the question whether AI language technologies can reverse the imbalance of the digital accessibility and visibility of minority languages, including Tibetan, Uyghur, Kazakh and Mongolian. The study used a mixed methods approach. In particular, the main methods were fine-tuning and evaluating neural machine translation and TTS, as well as participative surveys with different stakeholders including the language speakers, AI specialists and community representatives. The findings show that AI tools can reduce digital communication barriers for minority languages and support their use in online environments. The survey results further indicate that acceptance is not determined by technical performance alone, but also by users’ perceptions of cultural appropriateness, social relevance, and linguistic legitimacy. These findings add to the sustainable use of associated technologies, language policies and digital inclusion debates.
