With the rise of AI in education, knowledge tracing (KT) has become important for modeling students’ knowledge from interaction data. However, existing methods still face three major challenges, including limited modeling of personalized exercise–concept relations, low robustness to noisy interactions, and inefficient training due to suboptimal data selection. To address these issues, we propose R 2 GCurL, a novel KT framework with two key designs. First, we recast KT as a graph classification problem and construct dynamic graphs from student responses, enabling the model to capture structural relations between exercises and concepts for more personalized KT. Second, we introduce a data-centric curriculum learning strategy based on dynamic graph entropy. Under our definition, pairwise dynamic graph entropy measures graph-transition continuity, where larger values indicate stronger structural similarity. Its sequence-level aggregation is used to derive a structure-aware difficulty signal for sample scheduling. On top of this, an RL-based scheduler further adapts batch selection based on model feedback and is especially beneficial under noisier and more unstable training regimes. Theoretical analysis shows that R 2 GCurL has lower computational complexity than existing graph-based KT models. Extensive experiments on five real-world datasets confirm its effectiveness, robustness, and generalizability, including as a plug-and-play enhancement for sequence-based KT models.

Sequential recommendation systems aim to predict the future behaviors of users based on their historical interactions. Despite the success of neural architectures like Transformer and Graph Neural Networks, these models often struggle with the inherent challenge of sparse data in accurately predicting future user behaviors. To alleviate the data sparsity problem, some methods leverage the contrastive learning to generate contrastive views, assuming the items appear discretely at the same time intervals and focusing on the sequence order. However, these approaches neglect the crucial temporal-aware collaborative patterns hidden within the user-item interactions, leading to a limited variety of contrastive pairs and less informative embeddings. The proposed framework,Temporal-awaregraph contrastive learning with theoretical guarantees for sequentialRecommendation (TagRec), integrates temporal-aware collaborative patterns with adaptive data augmentation to generate more informative user and item representations. TagRec employs a temporal-aware graph neural network to embed the original graph, then generates augmented graphs through the addition of interactions via latent user interest mining, the dropping of redundant interaction edges, and the perturbation of temporal information. Theoretical guarantees are provided that these augmentations enhance the graph’s utility. Extensive experiments on real-world datasets demonstrate the superiority of the proposed approach over the state-of-the-art recommendation methods.

Graph neural networks (GNNs) have shown ad-vantages in graph-based analysis tasks. However, most existing methods have the homogeneity assumption and show poor performance on heterophilic graphs, where the linked nodes have dissimilar features and different class labels, and the semantically related nodes might be multi-hop away. To address this limitation, this paper presents GraphRARE, a general framework built upon node relative entropy and deep reinforcement learning, to strengthen the expressive capability of GNNs. An innovative node relative entropy, which considers node features and structural similarity, is used to measure mutual information between node pairs. In addition, to avoid the sub-optimal solutions caused by mixing useful information and noises of remote nodes, a deep reinforcement learning-based algorithm is developed to optimize the graph topology. This algorithm selects informative nodes and discards noisy nodes based on the defined node relative en-tropy. Extensive experiments are conducted on seven real-world datasets. The experimental results demonstrate the superiority of GraphRARE in node classification and its capability to optimize the original graph topology.

Knowledge tracing models based on deep learning can achieve impressive predictive performance by leveraging attention mechanisms. However, there still exist two challenges in attentive knowledge tracing (AKT): First, the mechanism of classical models of AKT demonstrates relatively low attention when processing exercise sequences with shifting knowledge concepts (KC), making it difficult to capture the comprehensive state of knowledge across sequences. Second, classical models do not consider stochastic behaviors, which negatively affects models of AKT in terms of capturing anomalous knowledge states. This article proposes a model of AKT, called Enhancing Locality for Attentive Knowledge Tracing (ELAKT), that is a variant of the deep KT model. The proposed model leverages the encoder module of the transformer to aggregate knowledge embedding generated by both exercises and responses over all timesteps. In addition, it uses causal convolutions to aggregate and smooth the states of local knowledge. The ELAKT model uses the states of comprehensive KCs to introduce a prediction correction module to forecast the future responses of students to deal with noise caused by stochastic behaviors. The results of experiments demonstrated that the ELAKT model consistently outperforms state-of-the-art baseline KT models.

Active learning can select informative data for model training to reduce the amount of labelling efforts required. Because traditional active learning methods cannot be directly used for deep learning, researchers have proposed multiple deep active learning methods. However, none of the previous research efforts on deep active learning algorithms presents a specific framework for learning-to-rank tasks. In this work, we introduce a novel deep active learning framework based on Deep Expected Loss Optimization (DELO) for the answer selection task.

Given a question and a set of candidate answers, answer selection is the task of identifying the best answer, which can be viewed as a kind of learning-to-rank tasks. Learning to rank arises in many information retrieval applications, where deep learning models can achieve inspiring results. Training a deep learning model often requires large scale annotated data that are expensive and time-consuming to obtain. Active learning presents a promising approach to this problem by selecting more informative training data to reduce the amount of labelling efforts required. Because traditional active learning methods cannot be directly used for deep learning, researchers have proposed multiple deep active learning methods. However, none of the previous research efforts on deep active learning algorithms presents a specific framework for learning-to-rank tasks. In this work, we introduce a novel deep active learning framework based onDeepExpectedLossOptimization (DELO) for the answer selection task. It adopts a data acquisition function based on model uncertainty with Bayesian deep learning and the expected loss optimization. Moreover, a two-step batch-mode procedure, combining DELO and other data acquisition strategies is proposed to further improve the performance of active learning. Experimental results verify the effectiveness of the proposed framework.

Reinforcement Learning, a common framework for AI planing or decision making, is regarded as an effective framework fro planning students' learning sequences. However, previous research efforts rely on simulation data or students and few of them have verified effectiveness of their algorithms with teaching students in real educational environments. Also, the previous research is hard to using in real task such as online course or real classroom because of the strong assumptions and restrictions. In this paper, We propose a new deep reinforcement learning framework for instructional sequencing that can recommend students with personalized learning exercises in both MOOCs and classrooms. Based on our ATC model, is used to trace students learning process and Deep Reinforcement Learning Agent is used to induce efficient l earning sequences for students adaptively. Both simulation and experiment in classrooms confirm the effectiveness of our method.

With the advent of Massive Online Open Courses (MOOCs), the data scale of student learning behavior and knowledge mastery has significantly increased. In order to effectively and efficiently analyze these datasets and present on-the-fly intelligent tutoring to online learners, it is necessary to improve existing learning analytics tools in a parallel and automatic way. One of the most common tools is Bayesian Knowledge Tracing (BKT) that can model temporal progress of online learners and evaluate their mastery of course knowledge. Current implementation of BKT is mostly based on single machine, which leads to slow execution performance during its Expectation Maximization(EM) algorithm for parameter fitting. Although there are a few parallel implementations for EM algorithm, they don't support automatic initial BKT parameter tuning to ensure the correct convergence of the EM iteration. Therefore, this paper presents a new parallel BKT open source tool based on the Spark computational framework with the method of automatic tuning of initial parameters. This tool improves traditional knowledge tracing systems using a parallel EM algorithm with the capabilty of automatically choosing initial parameters. Experimental result demonstrates that our tool can achieve fast execution speed and greatly improve the accuracy of training parameters on both different sizes of simulated data and real educational data sets.

A large amount of data generated by students in online education can be used to improve the quality of education. The important task of online education is to estimate the student proficiency and the characteristics of educational components. We developed the T-BMIRT model: a temporal, multidimensional, IRT-based method for estimating the above parameters. The model added learning video parameters and modeled the student proficiencies over time as a random process, accounting for the student learning and forgetting process. And it was extended to multidimensional to estimate the educational components which contain multiple skills. So the model can describe the student learning trajectories in an online education system. In addition, we evaluated this model by predicting student next response to assessment, and found it is better than the IRT and temporal IRT models on each dataset we used, especially when the dataset contains learning videos interactions.

Knowledge Tracing (KT) is a standard model for inferring student knowledge mastery from their performance data. Generally, there are five parameters in KT model we need to estimate, they are transition parameters: learning and forgetting probability, emission parameters: guessing and slipping probability, and prior probability. KT model is widely used in students learning outcome modeling. However, it does not support multiple subskills modeling, for the model works by checking the historical observations at a specific skill. To overcome this drawback, this paper proposes three models using logistic regression over each step: KTLR-GS, extending guessing and slipping parameters; KTLR-LFID, introducing item difficult into the KT model while extending learning and forgetting parameters; KTLR-FP, extending both transition and emission parameters. Unlike previous methods our models discuss the efficiency of extending transition and emission parameters while relaxing the assumptions that subskills are independent. In the end we evaluate how well our models perform with comparison to LR-DBN, KT-IDEM and KT on two datasets: the open dataset ASSISTments, and one private dataset, student algebra dataset collected through our tutoring system. Our models outperform LR-DBN, KT-IDEM and KT on both datasets.