Knowledge Inference: Estimating how much the student knows
Submitted by Amar Lalwani (@amar1707) on Saturday, 30 April 2016
Very high student-teacher ratios, lack of infrastructure and other socio-economic issues have affected quality and accessibility of education significantly. Moreover, Education can also benefit from the potential and promises of technology (particularly AI), which has already transformed our lives in many aspects. An Intelligent Tutoring System (ITS) is a computer system which enables learning in some domain of study. By this abstract definition, an ITS is supposed to possess knowledge about the domain, knowledge about the learner and knowledge about the teaching strategies.
Enhancing student knowledge being the primary goal of education, it is important to be able to measure the student knowledge. If we can measure it, we would know whether we are making it any better. Of course, if we can measure it, we can make automated pedagogical decisions and also inform instructors and other stakeholders about it.
This talk is about techniques to model students’ changing knowledge state during the process of skill (knowledge) acquisition, also known as Knowledge Tracing. This model enables the system to maintain the estimate of the probability that the student has mastered the concepts. Based on these probability estimates, the system individualizes the learning path and provides assistance as necessary.
The talk will mainly focus on the following topics:
i) Modelling of Knowledge, KC (Knowledge Component) Models
ii) Bayesian Knowledge Tracing (BKT): A classical approach to Knowledge Inference and Tracing
iii) Variants of BKT
iv) Other models like PFA(Performance Factor Analysis), IRT(Item Response Theory) and other ML models which use all the unutilized features.
Responsible for Research and Development of the algorithms to make funtoot (Intelligent and Adaptive personal tutor for K-12 Education) intelligent and powerful using the tools of AI & ML.