Evaluating the investment that healthcare organisations make in quality improvement requires knowledge of impact at multiple levels, including patient care, workforce and other organisational resources. The degree to which these resources help organisations to survive and thrive in the challenging contexts in which healthcare is designed and delivered is unknown. Investigating this question fro...

In many application domains, we encounter data which involves a graph encoding certain relationships and a set of items related to the graph. One example is in social websites where the users interact with each other, and share their interests on different items such as music or books. In this case, the direct interactions among the users can be represented as a graph, and the items like music ...

Contrastive learning has been widely applied to graph representation learning, where the view generators play a vital role in generating effective contrastive samples. Most of existing methods employ pre-defined generation methods, e.g., node drop or edge perturbation, which usually cannot adapt input data preserve original semantic structures well. To address this issue, we propose novel frame...

Multi-view learning has progressed rapidly in recent years. Although many previous studies assume that each instance appears all views, it is common real-world applications for instances to be missing from some resulting incomplete multi-view data. To tackle this problem, we propose a novel Latent Heterogeneous Graph Network (LHGN) learning, which aims use multiple views as fully possible flexi...

The thesis describes a concept for the abstraction of control flow graphs, view graphs, and some of its possible applications. Many computer science applications, especially in validation and verification, are directly or indirectly based on control flow analysis. Often, certain actions in the control flow are of special interest – e.g. input and output actions. These actions define a view of t...

In recent years, graph neural networks (GNNs)-based methods have been widely adopted for heterogeneous (HG) embedding, due to their power in effectively encoding rich information from a HG into the low-dimensional node embeddings. However, previous works usually easily fail fully leverage inherent heterogeneity and semantics contained complex local structures of HGs. On one hand, most existing ...

We give a unified proof of algorithmic weak and Szemerédi regularity lemmas for several well-studied classes sparse graphs, which only were previously known. These include core-dense low threshold rank (a version of) upper regular graphs. More precisely, we define cut pseudorandom prove our these then show that pseudorandomness captures all the above graph as special cases. The core approach is...