Example-Based Learning in Computer-Aided STEM Education

摘要

Providing personalized and interactive education (as in one-on-one tutoring) remains an unsolved problem in standard classrooms. The arrival of MOOCs, despite being an opportunity for sharing quality instruction with a large number of students, exacerbates the problem with an even higher student-to-teacher ratio. Recent advances in computer science can be brought together to rethink intelligent tutoring, with the phenomenal rise of online education making this investment very timely. This article has summarized recently published work from different areas of computer science, including programming languages, artificial intelligence, and human-computer interaction. It also reveals a common thread in this interdisciplinary line of work, namely the use of examples as an input to the underlying algorithms (for intent understanding), as an output of these algorithms (for generating the intended artifact), or even inside these algorithms (for inductive reasoning). This may enable other researchers to apply these principles to develop similar techniques for other subject domains. This article should inform educators about new advances to assist various educational activities, allowing them to think more creatively about curriculum and pedagogical reforms; for instance, these advances can enable development of gaming layers that take computational thinking into K-12 classrooms.