A Distributional Semantics Model for Image Recommendation using Learning Analytics

Abstract

In Society 5.0, a global trend is to process educational big data in order to solve various social and educational problems. This study, leveraging the power of learning analytics intends to investigate an unsolved problem in the domain of image-based vocabulary learning, namely how to recommend context-specific appropriate images at the right time, right place and right context. The aim of this study is to develop an innovative image recommendation system that can recommend Feature-based Context-specific Appropriate Images (FCAI) by analyzing learners’ previous learning records. This paper introduces a distributional semantics model for image recommendation.