Apprentice Learning

This research aims to develop a unified theory of learning from instruction and to provide a framework for building machines that are natural to teach. Central to this work is the Apprentice Learner Architecture, which faclitates model development and testing, and the Natural Training Interaction Framework, which outlines the space of natural training interactions.

The biggest output of this line of work to date is my thesis, which describes the core research thrust and its applications for tutor design, behavior predictions, and theory testing. Recently, I have been making exciting progress towards developing HCI theory to support the construction of machines that are natural to teach.

Naturally Teachable Machines

MacLellan, C.J., Harpstead, E., Marinier III, R. P., Koedinger, K.R. (2018). A Framework for Natural Cognitive System Training Interactions. Advances in Cognitive Systems, 6, 177-192. (pdf)

Sheline, R. & MacLellan., C.J. (2018). Investigating Machine-Learning Interaction with Wizard-of-Oz Experiments. In Proceedings of the NeurIPS 2018 Workshop on Learning by Instruction. (pdf)

Behavior Prediction and Learning Theory Testing

MacLellan, C.J., Harpstead, E., Patel, R., Koedinger, K.R. (2016). The Apprentice Learner Architecture: Closing the loop between learning theory and educational data. In Proceedings of the 9th International Conference on Educational Data Mining. Raleigh, NC: International Educational Data Mining Society. (Exemplary Paper Award) (pdf)

Tutor Design

MacLellan, C.J., Koedinger, K.R., Matsuda, N. (2014) Authoring Tutors with SimStudent: An Evaluation of Efficiency and Model Quality. In S. Trausan-Matu, K. E. Boyer, M. Crosby, & K. Panourgia (Eds.), Proceedings of the 12th International Conference on Intelligent Tutoring Systems (pp. 551-560). Switzerland: Springer International. doi: 10.1007/978-3-319-07221-0 (pdf)