Video-based learning assistant scheme for sustainable education

Recently, owing to the development of information technology, there have been many changes in learning activities and equipment. For instance, compared with traditional textbook-based learning, e-learning has freed learners from spatial-temporal restrictions, and other educational media are replacing textbooks. Specifically, educational videos have become very popular, as they are effective at conveying their content or meaning. Furthermore, their accessibility has been improved because of the emergence of platforms that enable the sharing and delivery of such videos. However, when learners use such platforms, they tend to face difficulty in finding videos they want, as these platforms have a lot of videos. They also experience the inconvenience of performing additional searches because of the lack of information relevant to the video. To solve these problems, we propose a learning assistant system for educational video platforms. Our system helps learners find videos based on their readability level and contents. Moreover, it provides learners with useful information such as keywords and named entities for further learning activities. To demonstrate the effectiveness, we implemented a prototype system on a website dealing with speech videos and performed various experiments. By reporting some results, we show that our scheme can be used to implement sustainable education.