Bi-CLKT: Bi-Graph Contrastive Learning based Knowledge Tracing

Song, Xiangyu; Li, Jianxin; Liu, Qi; Wang, Zhao; Chen, Yunliang; Mian, Ajmal

doi:10.48550/arxiv.2201.09020

Cited by 3 publications

(3 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For such trends, graphical representation is harnessed through Graph Neural Networks (GNNs). There are multiple models that have been proposed in the field of graphical representation such as Graph-based knowledge Tracing (GKT) [53], Graph-based Interaction knowledge tracing (GIKT) [54], Bi-Graph Contrastive Learning based knowledge tracing [55] and Structure-based knowledge tracing (SKT) [56].…”

Section: ) Knowledge Tracing Methodsmentioning

confidence: 99%

AI-Based Personalized E-Learning Systems: Issues, Challenges, and Solutions

et al. 2022

View full text Add to dashboard Cite

A personalized e-learning system is effective in imparting enhanced learning to its users. As compared to a conventional e-learning system, which provides similar contents to each learner, a personalized learning system provides specific learning contents and assessments to the learners. Personalization is based on Artificial Intelligence (AI) based techniques in which appropriate contents for each learner are determined using the level of comprehension of the learner and the preferred modes of learning. This paper presents requirements and challenges for a personalized e-learning system. The paper is focused in elaborating four research questions, which are related to identifying key factors of personalized education, elaborating on state of the art research in the domain, utilizing benefits of AI in personalized education, and determining future research directions. The paper utilizes an extensive survey of current research papers in answering these questions. It provides a comprehensive review of existing solutions in offering personalized e-learning solutions. It also elaborates on different learning models and learning theories, which are significant in providing personalized education. It proposes an efficient framework, which can offer personalized elearning to each learner and identifies significant directions for future research. The paper is beneficial for academicians and researchers in understanding the requirements of such a system, comprehending its methodologies, and identifying challenges which are needed to be addressed.

show abstract

Section: ) Knowledge Tracing Methodsmentioning

confidence: 99%

AI-Based Personalized E-Learning Systems: Issues, Challenges, and Solutions

et al. 2022

View full text Add to dashboard Cite

show abstract

“…For example, Liu et al [32] used KNN-XGBoost to predict missing values, which played an essential role in the task of detecting transmission line risks in smart grids. Song et al [33] proposed a Bi-CLKT to track students' learning, which helps education departments to formulate systematic learning plans for students. However, most of the above work is based on the traditional singlemachine computing mode, which is not good at processing big data as the amount of data in the actual task increases rapidly.…”

Section: B Distributed Deep Learning Computing Frameworkmentioning

confidence: 99%

Object Detection in Large-Scale Remote Sensing Images With a Distributed Deep Learning Framework

Liu

Yan

et al. 2022

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

With the accumulation and storage of remote sensing images in various satellite data centers, the rapid detection of objects of interest from large-scale remote sensing images is a current research focus and application requirement. Although some cutting-edge object detection algorithms in remote sensing images perform well in terms of accuracy (mAP), their inference speed is slow and requires high hardware requirements that are not suitable for real-time object detection in large-scale remote sensing images. To address this issue, we propose a fast inference framework for object detection in large-scale remote sensing images. On the one hand, we introduce α-IoU Loss on the YWCSL model to implement adaptive weighted loss and gradient, which achieves 64.62% and 79.54% mAP on DIOR-R and DOTA test sets, respectively. More importantly, the inference speed of the YWCSL model reaches 60.74 FPS on a single NVIDIA GeForce RTX 3080 Ti, which is 2.87 times faster than the current state-of-the-art one-stage detector S 2 A-Net. On the other hand, we build a distributed inference framework to enable fast inference on large-scale remote sensing images. Specifically, we save the images on HDFS for distributed storage and deploy the pretrained YWCSL model on the Spark cluster. In addition, we use a custom partitioner RankPartition to repartition the data to further improve the performance of the cluster. When using 5 nodes, the speedup of the cluster reaches 9.54, which is 90.80% higher than the theoretical linear speedup (5.00). Our distributed inference framework for large-scale remote sensing images significantly reduces the dependence of object detection on expensive hardware resources, which has important research significance for the wide application of object detection in remote sensing images.

show abstract

“…Second, these methods depend highly on data augmentation methods requiring many hyperparameter tunings. On the other hands, based on graph contrastive learning [27], Bi-CLKT [21] proposes graph-based CL. However, as GKT [15], constructing graph structures are computationally expensive and too slow to apply to large-scale educational platforms like MOOC, which requires inferring the correct probabilities according to the conditions of all items or skills (see Fig.…”

Section: Contrastivementioning

confidence: 99%

SAICL: Student Modelling with Interaction-level Auxiliary Contrastive Tasks for Knowledge Tracing and Dropout Prediction

Park¹,

Kim²,

Kwan³

et al. 2022

Preprint

View full text Add to dashboard Cite

Knowledge tracing and dropout prediction are crucial for online education to estimate students' knowledge states or to prevent dropout rates. While traditional systems interacting with students suffered from data sparsity and overfitting, recent sample-level contrastive learning helps to alleviate this issue. One major limitation of sample-level approaches is that they regard students' behavior interaction sequences as a bundle, so they often fail to encode temporal contexts and track their dynamic changes, making it hard to find optimal representations for knowledge tracing and dropout prediction. To apply temporal context within the sequence, this study introduces a novel student modeling framework, SAICL: student modeling with auxiliary interactionlevel contrastive learning. In detail, SAICL can utilize both proposed self-supervised/supervised interaction-level contrastive objectives: MilCPC (Multi-Interaction-Level Contrastive Predictive Coding) and SupCPC (Supervised Contrastive Predictive Coding). While previous samplelevel contrastive methods for student modeling are highly dependent on data augmentation methods, the SAICL is free of data augmentation while showing better performance in both self-supervised and supervised settings. By combining cross-entropy with contrastive objectives, the proposed SAICL achieved comparable knowledge tracing and dropout prediction performance with other state-of-art models without compromising inference costs.

show abstract

Bi-CLKT: Bi-Graph Contrastive Learning based Knowledge Tracing

Cited by 3 publications

References 27 publications

AI-Based Personalized E-Learning Systems: Issues, Challenges, and Solutions

AI-Based Personalized E-Learning Systems: Issues, Challenges, and Solutions

Object Detection in Large-Scale Remote Sensing Images With a Distributed Deep Learning Framework

SAICL: Student Modelling with Interaction-level Auxiliary Contrastive Tasks for Knowledge Tracing and Dropout Prediction

Contact Info

Product

Resources

About