Review Article

A Systematic Review of Deep Learning Approaches to Educational Data Mining

Table 2

Summary of EDM tasks, approaches, datasets, and types of datasets. “Specific” means that the dataset has been created for a specific study, and “General” means that it has been used in different publications.

Task ReferenceDatasetType

Predicting student performance, achievement of learning outcomes or characteristicsLin and Chi, 2017 [11]ITS PyreneesSpecific
Zhang et al., 2017 [49]ASSISment and OLI datasetsGeneral
Kim et al., 2018 [26]UdacitySpecific
Lalwani and Agrawal, 2017 [14]Funtoot datasetSpecific
Okubo et al., 2017 [24]Information Science Course datasetSpecific
Guo et al., 2015 [23]High schools datasetSpecific
Sharada et al., 2018 [22]ASSIStment 2018General
Wang et al., 2017 [12]Code course datasetSpecific
Tang et al., 2016 [21]Kaggle Automated Essay ScoringGeneral
Bendangnuksung and P., 2018 [20]Kaggle Students’ Academic Performance datasetGeneral
Mao et al., 2018 [15]ITS Pyrenees and ITS CordilleraSpecific
Wilson et al., 2016 [50]ASSISTment 2009-2010, KDD Cup 2010 and ITS KnewtonGeneral
Wilson et al., 2016 [16]ASSISTment 2009-2010 dataset, KDD Cup 2010 dataset and ITS KnewtonGeneral
Khajah et al., 2016 [17]Assistment 2009-2010 dataset, virtual student dataset, and data from Spanish and Engineering coursesGeneral and Specific
Xiong et al., 2016 [18]ASSISTments 2009-2010 datasetGeneral
Wang et al., 2017 [53]KDD Cup 2015 datasetGeneral
Kim et al., 2018 [27]UdacitySpecific
Montero et al., 2018 [13]ASSISTment 2009-2010 dataset, KDD Cup 2010 dataset and ITS Woot MathGeneral and specific
Piech et al., 2015 [10]Virtual student dataset and Assistments 2009-2010 datasetGeneral
Singh et al. 2018 [54]Kaggle Automated Essay ScoringGeneral
Alam et al., 2018 [25]Kaggle Students’ Academic Performance datasetSpecific
Yeung and Yeung, 2018 [19]ASSISTment 2009, ASSISTment 2015, ASSISTment Challenge, Statics2011, Simulated-5Specific

Detecting undesirable student behaviorsAung et al., 2018 [36]YouTube videos of school classroomsSpecific
Sharma et al., 2016 [34]StyleX dataset (multimedia)Specific
Teruel and Alemany, 2018 [29]ASSISTment 2009-2010 dataset and KDD Cup 2015General
Fei and Yeung, 2015 [28]--
Whitehill et al., 2017 [31]HarvardX MOOCsGeneral
Wang et al., 2017 [30]Code course datasetSpecific
Min et al., 2016 [33]Game-based virtual learning environment Crystal IslandSpecific
Tato et al., 2017 [37]French corpusSpecific
Yang et al., 2018 [35]Videos collected in unconstrained environmentsSpecific
Xing and Du, 2018 [32]Canvas project management MOOCSpecific

Generating recommendationsWong, 2018 [39]Student transcript recordsSpecific
Abhinav et al., 2018 [38]Learner’s profile dataSpecific

EvaluationAkram et al., 2018 [44]problem-solving dataset from game-based learning environmentSpecific
Zhang et al., 2016 [42]Short answers from ITS CordilleraSpecific
Taghipour and Ng, 2016 [41]Kaggle Automated Essay ScoringGeneral
Zhao et al., 2017 [40]ASSISTment 2009-2010 and Kaggle Automated Essay ScoringGeneral
Alvarado et al., 2018 [43]Short-answer question dataset from biology courseSpecific
Choi et al., 2017 [45]PODS datasetSpecific
Sales et al., 2018 [46]2015 ASSISTments Skill Builder DataGeneral