Co-embeddings for Student Modeling in Virtual Learning Environments

A Systematic Review of Deep Learning Approaches to Educational Data Mining

“…Using DL techniques, they obtained significantly better performance than traditional machine learning methods for all three definitions of dropout: participation in the final week, last week of engagement, and participation in the next week. References [29,30] defined dropout as a binary classification problem. Reference [30] combined different DL architectures in a bottom-up manner, selecting three attributes from the dataset as an input.…”

“…The results showed that the proposed model could achieve comparable performance to approaches relying on feature engineering performed by experts. Reference [29] optimized a joint embedding function to represent both students and course elements into a single shared space. The results indicated that coembeddings were able to capture the latent causes involved in dropout, outperforming other disjoint and not embedded representations.…”

“…This is exactly the aim of ASSISTment (https://sites.google.com/site/assistmentsdata/) [47,48]. This dataset is used in many papers to predict student performance [10,13,16,18,19,22,29,46,49,50]. It is composed of a series of mathematics exercises offered to middle-school students through the ASSISTment platform (https://www.assistments.org/), including information such as assignment and user identification, whether the answer is correct on the first attempt or not (a binary flag indicating if the student completed the exercise correctly), the number of student attempts on a problem, answer type, etc.…”

“…Unfortunately, it seems that the data is no longer available. This dataset was used in [29,53]. The largest dataset for the analysis of student dropout was presented in [31].…”

“…The symbol "−" represents approaches that do not compare DL with traditional machine learning techniques. Instead, they present comparisons of different DL architectures [19,29,35,45,50], comparisons of different hyperparameters for the same DL architecture [31,46], or proposals not evaluated yet [39].…”