The Inclusive Glossary: An Embedded, Interactive Approach to Accessible and Inclusive Learning

To assist students in engineering and related STEM disciplines, we report on the motivation, design, implementation, and evaluation of the Inclusive Glossary, a novel embedded interactive educational tool. The Glossary explains technical terms when the student encounters new terms in video and written content. The Glossary was motivated by two equally-important factors. Firstly, to add American Sign Language (ASL) signing of technical terms as a first-class, inclusive educational outcome, and within the normal learning environment of university students. Secondly, to help mitigate the ongoing readiness-to-learn effects due to the lowered learning outcomes from the 2020-22 COVID-19 pandemic and inequity in students' prior high school education experiences. The Glossary takes a strong inclusive design stance; for all students there are valuable context-specific just-in-time learning opportunities to address “Knowledge-Gaps” that create barriers to learning the current topic of study. It also enables ASL signers to learn the growing and evolving corpus of engineering, physics and computer science signs. The Glossary's design and implementation is introduced from three perspectives: ASL, Universal Design for Learning (UDL), and Active Learning. ASL - a complete natural language with its own unique grammar and terms - is the first and primary language of some students who are Deaf or Hard of Hearing (DHH). The principles of UDL promote a user-configurable design that provides multiple forms of modality, engagement and interactivity. Scholastic research into Active Learning suggests student-initiated knowledge-seeking actions, when embedded into video-based and text-based learning experiences, improve learning outcomes and reduce the difficulty or perceived difficulty of a course. The Glossary is implemented as a web application that uses an automated workflow to efficiently find, download, and index domain-specific terms, definitions, and explanations in two primary languages, English and ASL, in text and video form. The automated workflow extracts domain terms from both the audio transcription and visual text from video content. Definitions and explanations of the glossary terms in English and ASL are automatically curated from open web-sources with zero or minimal instructor time required. Explanations in different lengths are provided for students with different interest levels, learning needs, and attention spans. ASL video entries are provided in three sign forms; as an isolated sign, a sentence definition, and an example usage. Students can view both English and ASL explanations. By embedding the Glossary into ClassTranscribe, we describe the user interface comprised of i) A glossary appendix inside the course notes, ii) Web page popups in the video player, and iii) An online gallery page to browse, edit, and search for glossary terms of the course. The extraction efficiency, precision and recall of the system were evaluated using a corpus of 300 candidate domain-specific terms automatically extracted from 8 videos. For English entries, 241 (80.3%) glossary items had a corresponding English explanation available. For ASL entries, 39 (13.0%) glossary items had a corresponding ASL definition available, and 20 (6.7%) items had ASL sign, example and definition available. Preliminary results suggest this is a promising educational technology that has the potential to help all students thrive in their engineering disciplines.