Abstract: Recent advances in AI-based adaptive technologies hold great promise for improving the quality and impact of computer science education. The recent surge in students calls for a robust and scalable solution to providing individualized pedagogical support for computer science students. Advanced learning technologies can make this possible by effectively modeling students' current programming knowledge and skills through conducting static and dynamic code analysis. Meanwhile, the emergence of programming language modeling techniques has brought significant opportunities for automated analysis of programs, including source code classification and code clone detection. However, those models cannot be readily applied to analyze students' generated codes. Modeling students' current state of knowledge and skills require significant adaptation of existing techniques to the specific characteristics of students' programs, including the abundance of errors and style issues and the particular needs of the educational domain. This talk presents current and future work that focuses on devising and adapting deep-learning-based code-representation approaches for student modeling in introductory computer science classrooms. It explores how AI-empowered learning technologies can play a central role in CS education for the next generation of computer science professionals.

Short Bio: Bita Akram is an Assistant Research Professor at the Department of Computer Science at North Carolina State University. She obtained her Ph.D. from North Carolina State University in 2019. She is the director of the AI-Assisted Learning lab. Her research focuses on adaptive learning technology for computer science education research and equity and diversity in computing fields.