No abstract

No abstract

As distributed online communication becomes increasingly common, and audiences for live online presentations grow larger, the ability to receive meaningful feedback from audience members who are distant and distributed becomes a necessity. To this end, we have built upon previous work to create a tool that is capable of providing real time feedback to an online presenter about the engagement level of the audience. The tool makes inferences by using computer vision and machine learning techniques to analyze the faces of audience members.

No abstract

The Online Course Tool for Adaptive Learning (OCTAL) is an adaptive exercise system that customizes the progression of question topics to each student. By creating a concept dependency graph of topics in a course and modeling a student's knowledge state, the tool presents questions that test knowledge within a student's zone of proximal development. We intend OCTAL to be a formative assessment tool that is not tied to any specific course by providing language-agnostic questions on computer science concepts. While the tool will be generalizable for many courses, our first prototype includes a concept map and question set for UC Berkeley's introductory computer science course, CS10: The Beauty and Joy of Computing. Using the tool, we will launch an experiment in the spring to investigate metacognitive improvements in the identification of knowledge gaps by presenting online course material in a nonlinear fashion.