An AI-Powered Interactive Mobile Application for Multilingual Conference Accessibility: Design, Deployment, and Field Evaluation

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Global academic conferences often face accessibility challenges for multilingual participants, especially during live sessions where real-time comprehension is important. This study presents the design, deployment, and evaluation of an AI-powered multilingual mobile application developed to support conference accessibility. The application integrates real-time speech translation, agenda management, venue navigation, and cultural information in a single mobile platform. It was deployed during the International Conference on Information Technology (ICIT 2025) and evaluated with 200 participants across four supported languages: Arabic, English, French, and Spanish. Using a design-based research (DBR) approach, the evaluation combined objective performance metrics with user-reported usability measures. The system achieved 92.5% translation accuracy, a word error rate (WER) of 7.2%, and system uptime of 99.1%. The usability results showed high participant satisfaction, with a mean score of 4.6 out of 5 and strong internal reliability (Cronbach’s α = 0.91). Open-ended feedback indicated that participants valued translation usefulness, ease of use, and the integration of conference services, while also suggesting additional language support and offline functionality. The findings suggest that integrating AI-based multilingual translation with mobile conference services can help reduce language barriers and improve participant experience in international academic events.