Machine Learning in Education
Quoting my advisor Prof Beverly Woolf, this field explores theories about learning and builds software that delivers differential teaching. The teaching response is adapted to student needs after reasoning about domain knowledge, student mastery and pedagogy. I feel a strong connection to this field as it is the perfect union of my computer science skills, my vision to make education openly available and the need to empower every underprivileged child in the world. These tutors have a great potential to customize their tutoring strategy to the needs of each student, which is difficult to do in a traditional educational setting.
Current Areas of Focus
Studies suggest that an emotional state of a student interacts with his/her engagement and learning. An accurate detector of negative emotional states would allow for a variety of improved interventions in a tutoring system for disengagement repair. Most of the work on automatic affect detection relies on the data from several sensors. Practically, it would be challenging to deploy these sensors in an actual classroom. I aim to build a cost effective solution in context of learning. I am also interested in exploring ways to respond to the student needs meaningfully by personalizing the choice of next pedagogical action.
Culturally sensitive tutors enhance learning because culture enables us to create a balance between academic abstractions and sociocultural realities. As Prof Geneva Gay realized, adapting culturally familiar styles and content in teaching and incorporating social experiences into curriculum and instruction will increase a student’s ability to engage with a learning environment and hence improve student performance. My work should aid the children in any country attain academic achievement without compromising their ethnic and cultural identity.