Usability and Learning Effect Evaluations of an Electrical Note-Taking Support System with Speech Processing Technologies

Abstract

This paper describes the evaluation of an electronic note-taking support system that works on a tablet/laptop computer in a real classroom environment. The distinguishing characteristic of the proposed electronic note-taking support system is the incorporation of a few state-of-the-art speech processing technologies. One is deep-neural-network-based automatic speech recognition for transcribing talks. Such transcriptions are very useful for not only note-taking but also recovering missing words. The other speech processing technology is keyword search in the recorded speech. We had already developed the system previously and evaluated it. In this study, we improved the previous support system from the viewpoints of usability and visibility of the system user interface. The improved system was evaluated in the subjective experiment in which listeners (students) who attended the real lectures. In the experiment, not only the usability of the system but also the learning effect for knowledge acquisition were evaluated. The experimental results showed that the improved note-taking support system improved usability and visibility compared to the previous system and the traditional note-taking method. Moreover, the students claimed that the system helped them acquire knowledge about the lectures they attended.