Context-Aware Multilingual Learning Based on Spatiotemporal Knowledge Base and Dynamic Conversation: A System Architecture

Most of the foreign students studying abroad in Taiwan may not speak Mandarin. It is important for them to learn Mandarin for their daily life. Although universities in Taiwan provide Mandarin classes, we design this system to overcome limitations of time and space occurring in traditional classroom. The system architecture provides contents in a smart phone application with multilingual versions of user interfaces, word banks and conversations. The contents are marked with spatiotemporal tags, therefore they become context-aware. The system will provide appropriate contents according to the learner's location and the time. The content of conversation are generated dynamically from a spatiotemporal knowledge base according to the learner's response to the previous progress of the conversation. The learning history is recorded and tracked via a cloud server, so that the system can provide just-in-time and just-in-place contents for the learners.