1. Objective Knowledge tracing model tries to find parameters of the Bernoulli distribution: $p_{t+1} = P(c_{t + 1} = 1| e_{1: t + 1}, c_{1:t+1})$ $c_{t+1} \sim Bernoulli (p_{t+1})$ $e_t \in {1, ... N}$ is skill, $c_{t}$ is correctness at time step t.

2. Architecture

Knowledge Query Network consists of 3 components:

1. Knowledge Encoder The student responses are converted in knowledge state.

2. Skill Encoder Skill is converted into skill vector.

3. Knowledge State Query These two vectors are passed to knowledge state query to predict the probability of correctness.