Towards a Deep Learning approach to regularise discourse of collaborative learner

Collaborative learning is a method of education in which a group of learners solves a particular task. A collaborative setting encourages learners to take a more active role in knowledge construction. However, when they communicate on a virtual platform such as a chat platform, it is important that they can refer to each other correctly so that they can improve their learning activities with the help of each other, but learners can be sidetracked, which retards their learning progress. To address this issue, this thesis practiced text classification approaches to regularize the conversation betweenlearners so they could refer to each other correctly. The dataset was collected from a focus group experiment designed for students in the Educational Technology Department at Saarland University. The report gives a clear idea of how the collected dataset has been coded and validated with the help of intercoder reliability measurements. After data preprocessing, state-of-the-art data augmentationtechniques such as spelling, insertion, substitution, and synonym augmentation are applied. The thesis examines various neural network models to identify the best model for the dataset. Among them, Bidirectional Encoder Representations from Transformers (BERT) provides the best performance with an accuracy of 0.94 and a 0.17 loss value for the augmented preprocessed dataset, where recurrent neural network models tend to overfit. In the evaluation part, a summary of performance matrices is shown, and to evaluate the model, a new dataset with similar data is generated with the help of the OpenAI API Key. The BERT model is able to classify 960 responses out of 1005, where both recurrent neural networks are classified less than 200. The thesis also discussed the issue of model poisoning so that when the model is updated, it can tackle the unclassified responses. Finally, a simple demo of how this BERT model is used to regularize the discourse of two collaborative learners is presented with the help of the Jupyter interface.