In the field of recommender systems, graph neural networks (GNNs) have been extensively applied to collaborative filtering to generate personalized recommendations for users. To solve the problem of lack of observed data and contrasting interactions during representation learning, graph contrastive learning as an effective self-supervised learning (SSL) technique is presented to obtain augmented user and item representations. Nevertheless, most self-supervised approaches to generate recommendation either disrupt the graph structure or node embeddings through random augmentations or introduce augmented SSL information from biased data through heuristic methods. To overcome these challenges, we propose a learnable view augmentation model for collaborative filtering (LACF). Specifically, our framework embeds parameterized learnable view generators layer by layer into the automatic augmentation strategy, thus dynamically optimizing the adaptive augmented views of users and items through the backpropagation of weight gradients. In addition, LACF introduces a multiscale learning strategy that guides the view generator with layer-wise aware optimization and graph-level adaptive augmentation, enabling joint learning of representations with topological heterogeneity and semantic similarity from integrated viewpoint, achieving superior view augmentation. Extensive experiments on realworld datasets demonstrate that our LACF outperforms state-of-the-art baselines. In-depth analysis confirms the advantages of LACF in resistance against noise disturbances, alleviating data sparsity, and improving training efficiency.

In the field of recommender systems, self-supervised learning has become an effective framework. In response to the noisy interaction behaviors in realworld scenarios, as well as the skewed distribution influenced by data sparsity and popularity bias, graph contrastive learning has been introduced as a powerful self-supervised method in collaborative filtering (CF) to learn enhanced user and item representations. Despite their success, neither heuristic manual enhancement methods nor the use of final node representations to construct contrastive pairs are sufficient to provide effective and rich self-supervised signals to regulate the training process. Therefore, the learned representations of users and items are either fragile or lack heuristic guidance. In light of this, we propose the Hierarchical multiview graph contrastive learning framework HMCF, which leverages the message passing mechanism at the layer level to introduce different granularity levels of view augmentation using supervised signals, thus better enhancing the CF paradigm. HMCF leverages rich, high-quality self-supervised signals from different granularity views for accurate contrastive optimization, helping to alleviate data sparsity and noise issues. It also explains the hierarchical topology and relative distances between nodes in the original graph. Comprehensive experiments on three public datasets shows that our model significantly outperforms the state-of-the-art baselines.

Mining learner preferences and needs from individual learning behavior data is a critical task in course recommendation systems. While graph-based models have shown efficacy in capturing pairwise relationships between learners and courses, they often overlook the complex higher-order interactions involving learners, courses and teachers that are essential for accurate recommendations. To address this limitation, we propose a novel Hypergraph Convolutional Network for Course Recommendation (HCNCR) framework, designed to model these higher-order interactions effectively. Our approach constructs course and learner hypergraphs based on course attributes and learner similarity relations, respectively. By employing hypergraph convolution, we capture the intrinsic higher-order relationships within these hypergraphs. Additionally, we utilize graph convolutional layers on the learner-course bipartite graph to integrate embeddings derived from hypergraphs, achieving comprehensive representations of both learners and courses. Extensive experiments conducted on real-world datasets demonstrate that HCNCR significantly outperforms existing state-of-the-art methods in course recommendation tasks.

Identifying the epistemic emotions of learner-generated reviews in massive open online courses (MOOCs) can help instructors provide adaptive guidance and interventions for learners. The epistemic emotion identification task is a fine-grained identification task that contains multiple categories of emotions arising during the learning process. Previous studies only consider emotional or semantic information within the review texts alone, which leads to insufficient feature representation. In addition, some categories of epistemic emotions are ambiguously distributed in feature space, making them hard to be distinguished. In this article, we present an emotion-semantic-aware dual contrastive learning (ES-DCL) approach to tackle these issues. In order to learn sufficient feature representation, implicit semantic features and human-interpretable emotional features are, respectively, extracted from two different views to form complementary emotional-semantic features. On this basis, by leveraging the experience of domain experts and the input emotional-semantic features, two types of contrastive losses (label contrastive loss and feature contrastive loss) are formulated. They are designed to train the discriminative distribution of emotional-semantic features in the sample space and to solve the anisotropy problem between different categories of epistemic emotions. The proposed ES-DCL is compared with 11 other baseline models on four different disciplinary MOOCs review datasets. Extensive experimental results show that our approach improves the performance of epistemic emotion identification, and significantly outperforms state-of-the-art deep learning-based methods in learning more discriminative sentence representations.

In the discussion forums of massive open online courses (MOOCs), cognitive processing (e.g., insight, certain) is considered an essential factor that can affect learners' learning outcomes, but the relationship between them has not been thoroughly investigated. Especially the dynamic nature of cognitive processing is still a significant research gap. In this study, we proposed an unsupervised topic model named Temporal Cognitive Topic Model (TCTM) to automatically classify cognitive processes and obtain the conditional probability with topics over time. The results indicated that completers had more active and timely cognitive engagement as time went on and tended to use certain cognitive words to discuss the topics related to the examination and certificates, which showed that they had explicit learning goals and plans. Non-completers often used exclusive cognitive words to discuss some off-task content that pointed out a distractive learning process. Using the model, teachers can capture learners' dynamic cognitive states and associated topics to improve teaching methods and increase course completion rates.

Course forums offer an interactive channel for learners to express opinions and feedback, which contain valuable emotions and topic information towards courses. In this paper, we propose an emotion oriented topic probabilistic model that can be used to calculate distributions of emotion-topic over words to discover what students are most concerned about. An experiment on real-life data indicates that students had a positive attitude for knowledge applications, a negative experience for the learning system, and expressed confusion about the final exam. We also visualize the temporal trends of emotions of the whole group and the groups with different levels of achievement. The proposed model has a potential in discovering students' emotions in their feedback, thus improving the online learning experience, and identifying at-risk students timely.

Understanding students’ testing behaviors may help researchers design better computer-based assessment. For this reason, this study aims at characterizing students’ behavioral patterns in online reading test by k-means clustering. The clustering algorithm adopts eight indicators: reading time, answering time, the number of choosing articles, the number of choosing questions, the number of selecting options, the number of marking questions, the number of revisiting a test and the final testing scores. The result identifies five clusters of student testers: slow readers, fast readers, question markers, fast responders, and re-readers.