Engaging or Zoning Out in Class: Automated Engagement Assessment with Unsupervised Clustering (2024 Spring)

Project Overview

This project involved automating the assessment of student engagement in STEM classrooms using eye-tracking data, clustering attention motifs with unsupervised learning methods.

Key Features

• Data Preprocessing: Processed eye-tracking data from over 25 STEM sessions to extract key features.
• Clustering with UMAP and K-Means: Applied UMAP for dimensionality reduction and K-Means for clustering to identify patterns of attention and zoning out.
• Attention Motifs: Discovered and analyzed motifs representing various levels of student engagement.