Wide & Deep Learning for Judging Student Performance in Online One-on-one Math Classes

TL;DR

This paper presents a Wide & Deep learning framework that improves the automation of student performance assessment in online one-on-one math classes by effectively handling noisy data and providing accurate mastery predictions.

Contribution

The paper introduces a novel Wide & Deep model tailored for noisy classroom data, enhancing student mastery prediction accuracy in online math education.

Findings

The model outperforms existing methods on mastery prediction metrics.

It demonstrates robustness to noisy and limited data scenarios.

The approach is effective for fine-grained student performance assessment.

Abstract

In this paper, we investigate the opportunities of automating the judgment process in online one-on-one math classes. We build a Wide & Deep framework to learn fine-grained predictive representations from a limited amount of noisy classroom conversation data that perform better student judgments. We conducted experiments on the task of predicting students' levels of mastery of example questions and the results demonstrate the superiority and availability of our model in terms of various evaluation metrics.