A Platform for Image Recommendation in Foreign Word Learning
Mohammad Nehal Hasnine, Brendan Flanagan, Masatoshi Ishikawa, Hiroaki Ogata, Kousuke Mouri, Keiichi Kaneko · Kyoto University Research Information Repository (Kyoto University) · 2019
This paper introduces a platform for image recommendation that can be used in informal learning of foreign words. The platform is based on a distributional semantics model (DSM) that is designed to recommend Feature-based Context-specific Appropriate Images (FCAIs) for representing a word. This technology is for a context-aware ubiquitous learning system that captures ubiquitous learning logs from various learning scenarios. This paper briefly discusses the data capturing tool, methods of employing learning analytics for ubiquitous learning logs analysis, natural language processing techniques applied for wordbank creation, and image embedding methods employed for feature analysis, development of an algorithm that determines the most appropriate FCAI images, and related scientific issues.