[3M1-GS-12] Human interface, education aid: Online education

[3M1-GS-12-05] An Interface for Browsing Measurement of Visual Intelligibility of Web Sites explaining Academic Concepts

〇Yohei Ohkawa¹, Kosei Soda², Shintaro Okada², Takehito Utsuro^2,5, Yasuhide Kawada³, Noriko Kando⁴ (1. AVILEN Inc., 2. University of Tsukuba, 3. Logworks Co., Ltd., 4. National Institute of Informatics, 5. RIKEN)

Keywords:Web sites explaining academic concepts, deep learning, visual intelligibility, ResNet, browsing interface

It often happens that learners who use Web pages as tools for studying
need to manually compare search results so that they can find beginner
friendly Web contents. This is simply because most Web contents detected
by existing search engines are not beginner friendly and are not
suitable for studying. Considering this, this paper especially
concentrates on the technique of measuring visual intelligibility of Web
pages explaining academic concepts. More specifically, this paper
proposes a Web interface that is quite useful when browsing across
several Web sites explaining academic concepts. As an example of
academic fields, we focus on the statistics field, and show that, with
the proposed Web interface, learners are able to easily select a Web
page explaining the given query academic concept of the statistics field
which has the most visually intelligible Web page layout among other Web
pages collected from several Web sites explaining statistics concepts.