Modeling and understanding students' off-task behavior in intelligent tutoring systems

  title={Modeling and understanding students' off-task behavior in intelligent tutoring systems},
  author={R. Baker},
  journal={Proceedings of the SIGCHI Conference on Human Factors in Computing Systems},
  • R. Baker
  • Published 29 April 2007
  • Computer Science
  • Proceedings of the SIGCHI Conference on Human Factors in Computing Systems
We present a machine-learned model that can automatically detect when a student using an intelligent tutoring system is off-task, i.e., engaged in behavior which does not involve the system or a learning task. [] Key Method We use this model in combination with motivational and attitudinal instruments, developing a profile of the attitudes and motivations associated with off-task behavior, and compare this profile to the attitudes and motivations associated with other behaviors in intelligent tutoring…

Figures and Tables from this paper

Automatic Detection of Off-Task Behaviors in Intelligent Tutoring Systems with Machine Learning Techniques

An extensive set of experiment results demonstrates the power of using multiple types of evidence, the personalized model, and the robust Ridge Regression algorithm to address data sparseness problem.

Learning to Identify Students' Off-Task Behavior in Intelligent Tutoring Systems

A machine learning model that can automatically identify off-task behaviors of students while using an intelligent tutoring system is proposed and a robust Ridge Regression algorithm is designed to estimate model parameters.

Differences Between Intelligent Tutor Lessons, and the Choice to Go Off-Task

Surprisingly, the best model predicting off-task behavior from lesson features contains only one feature: lessons that involve equation-solving.

Student Off-Task Behavior in Computer-Based Learning in the Philippines: Comparison to Prior Research in the USA

Background Off-task behavior can be defined as any behavior that does not involve the learning task or material, or where learning from the material is not the primary goal. One suggested path for

Predicting College Enrollment from Student Interaction with an Intelligent Tutoring System in Middle School

This paper predicts college attendance from detectors of specific aspects of student learning and engagement in the context of 3,747 students using the ASSISTment system in New England, producing detection that is both successful and potentially more actionable than previous approaches.

Predicting Individualized Learner Models Across Tutor Lessons

It is found that best-fitting student parameters trained on previous lessons do not directly transfer to new lessons; however, one can effectively predict the student parameters for the new lesson by using features derived from prior lessons, and prior to tutor text-reading transaction data.

Learning analytics in outer space: a Hidden Naïve Bayes model for automatic student off-task behavior detection

A novel machine learning model is described which automatically detects students' off-task behavior as students interact with a learning system, ASSISTments, based solely on log file data.

Design Recommendations for Adaptive Intelligent Tutoring Systems Learner Modeling ( Volume I ) 155 CHAPTER 14 ‒ Assessing the Disengaged Behaviors of Learners

This chapter considers the relative merits of knowledge engineering and data-mining approaches for this type of model, and focuses on the key validity concerns that must be addressed for these types of models to be used with confidence in a comprehensive framework such as GIFT.

Automatically Detecting a Student's Preparation for Future Learning: Help Use is Key

We present an automated detector that can predict a student’s later performance on a paper test of preparation for future learning, a post-test involving learning new material to solve problems

Analysis of students' actions during online invention activities - seeing the thinking through the numbers

Intelligent Tutoring Systems adapt to students' behavior and knowledge by tracing students' learning trajectories using a cognitive model of the domain using the large solution space in these tasks makes the model tracing approach very hard to design and implement.



Off-task behavior in the cognitive tutor classroom: when students "game the system"

Analysis of students who choose to game the system suggests that learned helplessness or performance orientation might be better accounts for why students choose this behavior than lack of interest in the material.

Designing intelligent tutors that adapt to when students game the system

Students use intelligent tutors and other types of interactive learning environments in a considerable variety of ways. In this thesis, I detail my work to understand, automatically detect, and

Detecting Student Misuse of Intelligent Tutoring Systems

A machine-learned Latent Response Model is presented that can identify if a student is gaming the system in a way that leads to poor learning and will be useful both for re-designing tutors to respond appropriately to gaming, and for understanding the phenomenon of gaming better.

Cognitive Tutors: Lessons Learned

The 10-year history of tutor development based on the advanced computer tutoring (ACT) theory is reviewed, finding that a new system for developing and deploying tutors is being built to achieve the National Council of Teachers of Mathematics (NCTM) standards for high-school mathematics in an urban setting.

Do Performance Goals Lead Students to Game the System?

It is found that the frequency of gaming the system does not correlate to a known measure of performance goals, and instead, gaming is correlated to disliking computers and the tutor.

Toward Tutoring Help Seeking: Applying Cognitive Modeling to Meta-cognitive Skills

A preliminary model of help-seeking behavior is presented that will provide the basis for a Help-Seeking Tutor Agent and found that students frequently avoided using help when it was likely to be of benefit and often acted in a quick, possibly undeliberate manner.

Engagement tracing: using response times to model student disengagement

This paper explores student disengagement and proposes an approach, engagement tracing, for detecting whether a student is engaged in answering questions, based on item response theory, which is sensitive enough to detect variations in student engagement within a single tutoring session.

Automatic Recognition of Learner Groups in Exploratory Learning Environments

The value of a data-based approach for recognizing learners as an alternative to knowledge-based approaches that tend to be complex and time-consuming even for domain experts, especially in highly unstructured ELEs are shown.

Informing the Detection of the Students' Motivational State: An Empirical Study

An empirical study is presented which provided a considerable amount of knowledge regarding motivation diagnosis and was formalised in order to create a set of motivation diagnosis rules that can be incorporated into a prototype tutoring system.

Integrating Affect Sensors in an Intelligent Tutoring System

This project augments an existing intelligent tutoring system (AutoTutor) that helps learners construct explanations by interacting with them in natural language and helping them use simulation