Our goal is to use machine learning and natural language processing to understand when & how FAQs are asked
and how to respond with useful feedback.
In most tutoring sessions, kids are not articulating their confusion into verbally-spoken questions. They don't have a script of words to ask for each question--they simply don't understand! In most cases, it is the job of the human tutor to lead the student to better understanding of the concepts, one problem at a time.
By recording and studying how good tutors interpret both student behavior and mathematical language, we can being to begin to identify common patterns, and write a script to address frequently asked questions.