How does adaptive learning work?

How does it work?

Knowledge Decomposition

A subject is broken down into hundreds or thousands of knowledge components
Learning content (videos, articles, examples, questions) are matched to each knowledge component.
A graph structure is used to relate each knowledge component.

Collect, Analyze, Adjust

Collect

Student data are collected through student interaction with the system.

diagnostic test
behavior (time taken, preferred content type)
achievements and mistakes

Analyze

Based on info collected, system analyzes student learning, skills, and select content, difficulty, and path accordingly.

Adjust

Using analyzed results, system adjusts the delivery, amount, difficulty and design of content, sequence, and assessment to fit student needs.

(EdSurge, 2016)

Where does AI come in.

Artificial Intelligence (AI) is used in adaptive learning systems to predict a student's grasp of a specific knowledge component. It is trained so that it can identify and predict what a student needs. It mimics a human teacher in selecting the right material and revisiting concepts that a student did not understand. AI constantly updates its model of a student's understanding and adjust parameters to provide the optimal learning path. AI requires data to train and become better, so the more the student uses the system, the better the AI algorithms can adjust to provide personalized learning.