Learning to Optimize Feedback for One Million Students: Insights from Multi-Armed and Contextual Bandits in Large-Scale Online Tutoring

• URL: http://arxiv.org/abs/2508.00270v1
• Date: Fri, 01 Aug 2025 02:30:29 GMT
• Title: Learning to Optimize Feedback for One Million Students: Insights from Multi-Armed and Contextual Bandits in Large-Scale Online Tutoring
• Authors: Robin Schmucker, Nimish Pachapurkar, Shanmuga Bala, Miral Shah, Tom Mitchell,
• Abstract summary: We present an online tutoring system that learns to provide effective feedback to students after they answer questions incorrectly.<n>Using data from one million students, the system learns which assistance action to provide for each question to optimize student learning.<n>We evaluate the resulting MAB policies in 166,000 practice sessions, verifying significant improvements in student outcomes.
• Score: 1.3184846764437286
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We present an online tutoring system that learns to provide effective feedback to students after they answer questions incorrectly. Using data from one million students, the system learns which assistance action (e.g., one of multiple hints) to provide for each question to optimize student learning. Employing the multi-armed bandit (MAB) framework and offline policy evaluation, we assess 43,000 assistance actions, and identify trade-offs between assistance policies optimized for different student outcomes (e.g., response correctness, session completion). We design an algorithm that for each question decides on a suitable policy training objective to enhance students' immediate second attempt success and overall practice session performance. We evaluate the resulting MAB policies in 166,000 practice sessions, verifying significant improvements in student outcomes. While MAB policies optimize feedback for the overall student population, we further investigate whether contextual bandit (CB) policies can enhance outcomes by personalizing feedback based on individual student features (e.g., ability estimates, response times). Using causal inference, we examine (i) how effects of assistance actions vary across students and (ii) whether CB policies, which leverage such effect heterogeneity, outperform MAB policies. While our analysis reveals that some actions for some questions exhibit effect heterogeneity, effect sizes may often be too small for CB policies to provide significant improvements beyond what well-optimized MAB policies that deliver the same action to all students already achieve. We discuss insights gained from deploying data-driven systems at scale and implications for future refinements. Today, the teaching policies optimized by our system support thousands of students daily.

Related papers

• Personalised Feedback Framework for Online Education Programmes Using Generative AI [0.0]
  This paper presents an alternative feedback framework which extends the capabilities of ChatGPT by integrating embeddings. As part of the study, we proposed and developed a proof of concept solution, achieving an efficacy rate of 90% and 100% for open-ended and multiple-choice questions.
  arXiv  Detail & Related papers  (2024-10-14T22:35:40Z)
• Active Fine-Tuning of Multi-Task Policies [54.65568433408307]
  We propose AMF (Active Multi-task Fine-tuning) to maximize multi-task policy performance under a limited demonstration budget.<n>We derive performance guarantees for AMF under regularity assumptions and demonstrate its empirical effectiveness in complex and high-dimensional environments.
  arXiv  Detail & Related papers  (2024-10-07T13:26:36Z)
• Reduced-Rank Multi-objective Policy Learning and Optimization [57.978477569678844]
  In practice, causal researchers do not have a single outcome in mind a priori. In government-assisted social benefit programs, policymakers collect many outcomes to understand the multidimensional nature of poverty. We present a data-driven dimensionality-reduction methodology for multiple outcomes in the context of optimal policy learning.
  arXiv  Detail & Related papers  (2024-04-29T08:16:30Z)
• Evaluating and Optimizing Educational Content with Large Language Model Judgments [52.33701672559594]
  We use Language Models (LMs) as educational experts to assess the impact of various instructions on learning outcomes. We introduce an instruction optimization approach in which one LM generates instructional materials using the judgments of another LM as a reward function. Human teachers' evaluations of these LM-generated worksheets show a significant alignment between the LM judgments and human teacher preferences.
  arXiv  Detail & Related papers  (2024-03-05T09:09:15Z)
• Differentiating Student Feedbacks for Knowledge Tracing [28.669001606806525]
  We propose a framework to reweight the contribution of different responses based on their discrimination in training.<n>We also introduce an adaptive predictive score fusion technique to maintain accuracy on less discriminative responses.
  arXiv  Detail & Related papers  (2022-12-16T13:55:07Z)
• Explainable Action Advising for Multi-Agent Reinforcement Learning [32.49380192781649]
  Action advising is a knowledge transfer technique for reinforcement learning based on the teacher-student paradigm. We introduce Explainable Action Advising, in which the teacher provides action advice as well as associated explanations indicating why the action was chosen. This allows the student to self-reflect on what it has learned, enabling generalization advice and leading to improved sample efficiency and learning performance.
  arXiv  Detail & Related papers  (2022-11-15T04:15:03Z)
• Off-policy Reinforcement Learning with Optimistic Exploration and Distribution Correction [73.77593805292194]
  We train a separate exploration policy to maximize an approximate upper confidence bound of the critics in an off-policy actor-critic framework. To mitigate the off-policy-ness, we adapt the recently introduced DICE framework to learn a distribution correction ratio for off-policy actor-critic training.
  arXiv  Detail & Related papers  (2021-10-22T22:07:51Z)
• Building a Foundation for Data-Driven, Interpretable, and Robust Policy Design using the AI Economist [67.08543240320756]
  We show that the AI Economist framework enables effective, flexible, and interpretable policy design using two-level reinforcement learning and data-driven simulations. We find that log-linear policies trained using RL significantly improve social welfare, based on both public health and economic outcomes, compared to past outcomes.
  arXiv  Detail & Related papers  (2021-08-06T01:30:41Z)
• Discovering an Aid Policy to Minimize Student Evasion Using Offline Reinforcement Learning [2.2344764434954256]
  We propose a decision support method to the selection of aid actions for students using offline reinforcement learning. Our experiments using logged data of real students shows, through off-policy evaluation, that the method should achieve roughly 1.0 to 1.5 times as much cumulative reward as the logged policy.
  arXiv  Detail & Related papers  (2021-04-20T21:45:19Z)
• Deep Discourse Analysis for Generating Personalized Feedback in Intelligent Tutor Systems [4.716555240531893]
  We explore creating automated, personalized feedback in an intelligent tutoring system (ITS) Our goal is to pinpoint correct and incorrect concepts in student answers in order to achieve better student learning gains.
  arXiv  Detail & Related papers  (2021-03-13T20:33:10Z)
• Generative Inverse Deep Reinforcement Learning for Online Recommendation [62.09946317831129]
  We propose a novel inverse reinforcement learning approach, namely InvRec, for online recommendation. InvRec extracts the reward function from user's behaviors automatically, for online recommendation.
  arXiv  Detail & Related papers  (2020-11-04T12:12:25Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.