Towards Interpretable Deep Learning Models for Knowledge Tracing

Lu, Yu; Wang, DeLiang; Meng, Qinggang; Chen, Penghe

doi:10.1007/978-3-030-52240-7_34

Cited by 27 publications

(12 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our results demonstrate that all explainability methods can derive interpretable motivations behind student success predictions, confirming the similar yet coherent observations made by [8] for the knowledge tracing field. However, while there was some agreement regarding the top features across the five explainability methods, key differences across methods emerged when we considered the exact importance scores (RQ1).…”

Section: Discussionsupporting

confidence: 86%

“…However, there exists only a handful of papers focusing on explainability in the field of machine learning for education. For example, [8] examined the inner workings of deep learning models for knowledge tracing through layer-relevance propagation. Other researchers [9] experimented with traditional machine learning models for student success prediction and implemented local explanations with LIME for transparency in the best performing model.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Evaluating the Explainers: Black-Box Explainable Machine Learning for Student Success Prediction in MOOCs

Swamy¹,

Krco²,

Marras³

et al. 2022

Preprint

View full text Add to dashboard Cite

Neural networks are ubiquitous in applied machine learning for education. Their pervasive success in predictive performance comes alongside a severe weakness, the lack of explainability of their decisions, especially relevant in humancentric fields. We implement five state-of-the-art methodologies for explaining black-box machine learning models (LIME, PermutationSHAP, KernelSHAP, DiCE, CEM) and examine the strengths of each approach on the downstream task of student performance prediction for five massive open online courses. Our experiments demonstrate that the families of explainers do not agree with each other on feature importance for the same Bidirectional LSTM models with the same representative set of students. We use Principal Component Analysis, Jensen-Shannon distance, and Spearman's rank-order correlation to quantitatively cross-examine explanations across methods and courses. Furthermore, we validate explainer performance across curriculum-based prerequisite relationships. Our results come to the concerning conclusion that the choice of explainer is an important decision and is in fact paramount to the interpretation of the predictive results, even more so than the course the model is trained on. Source code and models are released at http://github.com/epfl-ml4ed/evaluating-explainers.

show abstract

Section: Discussionsupporting

confidence: 86%

Section: Introductionmentioning

confidence: 99%

Evaluating the Explainers: Black-Box Explainable Machine Learning for Student Success Prediction in MOOCs

Swamy¹,

Krco²,

Marras³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Interpretable forms of knowledge retrieved from knowledge tracing models has a concrete domain of applicability in the educational environment [Liu et al 2021]. For example, in the BKT model, the knowledge estimates which are updated in the algorithm process for each student in the data can be used directly to estimate the strength and weakness of the students during the learning process [Lu et al 2020]. BKT have achieved better results than PFA in some studies [Raposo et al 2020].…”

Section: Problem Definitionmentioning

confidence: 99%

Towards Interpretability of Attention-Based Knowledge Tracing Models

Souza²,

Bernardino

et al. 2022

Anais Do XXXIII Simpósio Brasileiro De Informática Na Educação (SBIE 2022)

View full text Add to dashboard Cite

Knowledge Tracing (KT) models based on attention mechanisms have demonstrated in literature the capability to predict student performance more accurately than previous models in some datasets. However, they fail to directly infer student knowledge. In this paper, we apply a proposed extension already seen in KT literature in order to infer latent knowledge to these models. We apply the extension to four different attention-based KT models, to investigate whether these models can better infer the knowledge outside the learning system than previous models. We find that attention-based models can generate better knowledge estimate correlations with student’s scores than the previous models.

show abstract

“…Te key to intelligent learning diagnosis is to mine the learner's potential learning state, including their mastery state of knowledge points and learning ability, then predict the learner's performance on particular learning tasks. For this pivotal problem, two types of solutions are widely developed at present: psychometrics-based methods [8][9][10][11][12][13][14][15] and deep learning-based methods [16][17][18][19][20][21][22][23]. First, psychometrics-based methods appeared earlier and have been developed for more than 40 years.…”

Section: Introductionmentioning

confidence: 99%

A Unified Interpretable Intelligent Learning Diagnosis Framework for Learning Performance Prediction in Intelligent Tutoring Systems

Wang

Yan

Zeng

et al. 2023

International Journal of Intelligent Systems

View full text Add to dashboard Cite

Intelligent learning diagnosis is a critical engine of intelligent tutoring systems, which aims to estimate learners’ current knowledge mastery status and predict their future learning performance. The significant challenge with traditional learning diagnosis methods is the inability to balance diagnostic accuracy and interpretability. Although the existing psychometric-based learning diagnosis methods provide some domain interpretation through cognitive parameters, they have insufficient modeling capability with a shallow structure for large-scale learning data. While the deep learning-based learning diagnosis methods have improved the accuracy of learning performance prediction, their inherent black-box properties lead to a lack of interpretability, making their results untrustworthy for educational applications. To settle the abovementioned problem, the proposed unified interpretable intelligent learning diagnosis framework, which benefits from the powerful representation learning ability of deep learning and the interpretability of psychometrics, achieves a better performance of learning prediction and provides interpretability from three aspects: cognitive parameters, learner-resource response network, and weights of self-attention mechanism. Within the proposed framework, this paper presents a two-channel learning diagnosis mechanism LDM-ID as well as a three-channel learning diagnosis mechanism LDM-HMI. Experiments on two real-world datasets and a simulation dataset show that our method has higher accuracy in predicting learners’ performances compared with the state-of-the-art models and can provide valuable educational interpretability for applications such as precise learning resource recommendation and personalized learning tutoring in intelligent tutoring systems.

show abstract

Towards Interpretable Deep Learning Models for Knowledge Tracing

Cited by 27 publications

References 13 publications

Evaluating the Explainers: Black-Box Explainable Machine Learning for Student Success Prediction in MOOCs

Evaluating the Explainers: Black-Box Explainable Machine Learning for Student Success Prediction in MOOCs

Towards Interpretability of Attention-Based Knowledge Tracing Models

A Unified Interpretable Intelligent Learning Diagnosis Framework for Learning Performance Prediction in Intelligent Tutoring Systems

Contact Info

Product

Resources

About