Online health communities (OHCs) provide an essential platform for patients with similar health conditions to share experiences and offer moral support. However, many time-sensitive questions from patients often remain unanswered due to the multitude of threads and the random nature of patient visits in OHCs. Traditional recommendation systems solely based on similarity for recommendations cannot be directly applied in OHCs. They tend to overlook the influence of patients' dynamically changing features (e.g., health stages), affecting their ability to provide meaningful responses to questions. To address this, we propose a novel recommender system scenario designed for OHCs, which differs from traditional recommender systems in several ways. Firstly, it's challenging to model the social support factors that form helper-seeker links in OHCs. Secondly, the impact of patients' historical activities is complex to quantify. Lastly, ensuring recommended helpers have the requisite expertise is crucial. To overcome these challenges, we develop a Monotonically regularIzed diseNTangled Variational Autoencoders (MINT) model. This model formulates interactions between seekers and helpers as a dynamic graph, using encoded historical activities as node features. We also introduce a graph-based disentangle VAE to capture patient features and a monotonic regularizer to ensure the logical pairing of seekers and helpers. Our extensive experiments show the effectiveness of our approach.

Learning user preferences by modeling historical purchase behaviors has significantly succeeded in existing recommender systems. Most use trained models to make predictions for users, and they assume that the training data samples and test data sample sets come from the same distribution. However, in practical applications, the distribution of users’ true preferences may be more complicated, and data drift can easily make the trained model invalid on the test dataset. In this case, to accurately model user preferences based on their historical behavior, two difficulties need to be addressed. First, it is difficult to model various purchase behavior shift situations due to their complexity. Second, inferring users’ true preferences from the complicated shifting cases is challenging. To solve the above problems, we build a robust recommender system to predict possible user purchase shifts and make recommendations for users. First, we propose a simulating strategy to cover possible scenarios when user purchase behavior shifts. Second, we build a novel voting framework to ensure the robustness of predicting results based on learned preferences. Extensive experiments were conducted, and the results demonstrate the outstanding performance of the proposed method on MovieLens-1M and LastFM datasets, providing at most 37.31% and 30.95% relative performance gains, respectively.

Even though conventional methods have contributed a lot to understanding information diffusion to a certain extent, they are limited by the neglect of considering survivors (i.e., non-participants). As the counterpart to participants, survivors exert analogously a vital role in cascade analysis, which represent the inaccessible scope of the message. To characterize participants and survivors simultaneously and assemble them into the macro-level cascade representation, we propose an end-to-end model, named complement coupling network, which utilizes multi-gating mechanism to coalesce inhomogeneous input. We first design a novel strategy for sampling survivors, which extracts a sequence of emblematic survivors corresponding to the sequence of observed participants. Afterwards, the complement gate is designed to weigh the contributions of the participant and survivor to the cascade at each timestamp. The reset and output gates are reformed to update the cell state and output the cascade snapshot, respectively. Furthermore, an attention mechanism keyed by the source node is introduced to assemble all snapshots within the observation window for predicting multiple subsequent activated users. Extensive experiments on two real-world datasets demonstrate that the proposed model significantly outperforms state-of-the-art approaches.

Aspect-based sentiment analysis (ABSA) aims to identify the sentiment of one or more aspects in the text. Existing methods pay attention to the syntactic structure, and significant progress has been achieved by using graph convolutional network (GCN). However, they ignore internal connections between different types of syntactic structure from a fine-grained perspective, which may lead to underexploring critical syntactic information of sentences. Additionally, the aspect-oriented syntactic dependency is omitted that generally provides penetrating insights of the corresponding sentiment. To tackle these problems, we propose a multi-grained (both coarse-grained and fine-grained) syntactic dependency-aware graph convolutional network model (named MSD-GCN). Particularly, in the initial representation layer, we redesign the aspect-enhanced coarse-grained dependency graph and construct five fine-grained dependency graphs by taking into account the types of syntactic structure. Moreover, we explore a multi-grained syntactic enhancement layer, which employs GCN and attention mechanism over multi-grained dependency graphs to capture more abundant syntactic information. Experimental results on five datasets illustrate that our proposed MSD-GCN model outperforms other representative ones in terms of Accuracy and Macro-Averaged F1.