<![CDATA[Regional languages face challenges in digital representation, learning accessibility, and the availability of technological resources for educational applications. Existing digital language learning platforms often rely on static vocabulary retrieval with limited contextual understanding and adaptive learning capabilities. This study proposes SmartSasak Edu, an adaptive Natural Language Processing (NLP)-based mobile framework to support contextual Sasak language learning. The study employed a Design Science Research Methodology (DSRM) approach consisting of requirement analysis, framework development, NLP implementation, expert validation, and usability evaluation. The framework integrates Sasak language resources, NLP-based text processing, contextual retrieval mechanisms, adaptive learning features, and mobile interaction. Evaluation was conducted through NLP component testing, expert assessment, and user usability evaluation. The NLP evaluation achieved accuracy scores of 96.40% for text normalization, 95.80% for tokenization, and 91.70% for semantic retrieval, while contextual response generation obtained 89.60% expert agreement. Expert validation resulted in 93.00% linguistic validity and 91.40% technical feasibility. Usability evaluation involving 42 students using the System Usability Scale (SUS) produced a score of 84.25, indicating positive user perceptions of usability. The findings suggest that SmartSasak Edu provides an adaptive approach for supporting regional language learning through NLP-based educational technology and contributes to low-resource NLP and AI-supported language education research.]]>
Copyrights © 2026
