Relational information between different types of entities is often modelled by a multilayer network (MLN) - a network with subnetworks represented by layers. The layers of an MLN can be arranged in different ways in a visual representation, however, the impact of the arrangement on the readability of the network is an open question. Therefore, we studied this impact for several commonly occurring tasks related to MLN analysis. Additionally, layer arrangements with a dimensionality beyond 2D, which are common in this scenario, motivate the use of stereoscopic displays. We ran a human subject study utilising a Virtual Reality headset to evaluate 2D, 2.5D, and 3D layer arrangements. The study employs six analysis tasks that cover the spectrum of an MLN task taxonomy, from path finding and pattern identification to comparisons between and across layers. We found no clear overall winner. However, we explore the task-to-arrangement space and derive empirical-based recommendations on the effective use of 2D, 2.5D, and 3D layer arrangements for MLNs.

The emergence of multilayer networks as a concept from the field of complex systems provides many new opportunities for the visualization of network complexity, and has also raised many new exciting challenges. The multilayer network model recognizes that the complexity of relationships between entities in real-world systems is better embraced as several interdependent subsystems (or layers) rather than a simple graph approach. Despite only recently being formalized and defined, this model can be applied to problems in the domains of life sciences, sociology, digital humanities, and more. Within the domain of network visualization there already are many existing systems, which visualize data sets having many characteristics of multilayer networks, and many techniques, which are applicable to their visualization. In this Synthesis Lecture, we provide an overview and structured analysis of contemporary multilayer network visualization. This is not only for researchers in visualization, but also for those who aim to visualize multilayer networks in the domain of complex systems, as well as those solving problems within application domains. We have explored the visualization literature to survey visualization techniques suitable for multilayer network visualization, as well as tools, tasks, and analytic techniques from within application domains. We also identify the research opportunities and examine outstanding challenges for multilayer network visualization along with potential solutions and future research directions for addressing them.

The visual exploration of large and complex network structures remains a challenge for many application fields, such as systems biology or social sciences. Often, various domain experts would like to work together to improve the analysis time or the quality of the analysis results. Collaborative visualization tools can facilitate the analysis process in such situations. Moreover, a growing number of real world networks are multivariate and often interconnected with each other. Entities in a network may have relationships with elements of other related data sets, which do not necessarily have to be networks themselves, and these relationships may be defined by attributes that can vary greatly. A challenge is to correctly assign the attributes and relations between different data sets and graphs in order to be able to analyze them visually afterwards. The navigation between the resulting visualizations is also difficult. How can users be guided to other interesting data points relevant to their current view and how can this information be additionally displayed in a graph without losing the overview of the data?

In this dissertation, we propose our new web-based visualization environment OnGraX, which supports distributed, synchronous and asynchronous collaboration of networks and related multivariate data sets. In addition to standard collaboration features like event tracking or synchronizing, our client/server-based system provides a rich set of visualization and interaction techniques for better navigation and overview of the input network. Changes made by specific analysts or even just visited network elements can be highlighted by heat maps, which enable us to visualize user behavior data without affecting the original graph visualization. We evaluate the usability of the heat map approach against two alternatives in a user experiment.

Additional features of OnGraX include a comprehensive visual analytics approach that supports researchers to specify and subsequently explore attribute-based relationships across networks, text documents, and derived secondary data. Our approach provides an individual search functionality based on keywords and semantically similar terms over an entire text corpus to find related network nodes. For examining these nodes in the interconnected network views, we introduce a new interaction technique, called Hub2Go, which facilitates the navigation by guiding the user to the information of interest. To showcase these features, we use a large text corpus collected from papers listed in the IEEE VIS publications data set (1990--2015) that consists of 2,752 documents. We analyze relationships between various heterogeneous networks, a Bag-of-Words index, and a word similarity matrix, all derived from the initial corpus and metadata. We also propose a design for the interactive specification of degree-of-interest functions, which can be used to provide and evaluate configurations for guidance based on network attributes and logged user data in heterogeneous networks.

Analysts of specific application domains, such as experts in systems biology or social scientists, are often interested to visually analyze a number of different network structures in conjunction, for example by using various visual structures of so-called multilayer networks. From the perspective of the human analyst, a sufficient perception and, consequently, a good understanding of those visual representations of multilayer networks is a non-trivial and often challenging task. Despite this practical importance and the clearly interesting visualization challenges, only few evaluation studies exist that investigate usability and cognitive issues of complex networks or, more specifically, multilayer networks. In this position paper, we address two main goals. On the one hand, we discuss existing studies from the fields of human-computer interaction and cognitive psychology that could inform the designers of multilayer network visualization in the future. On the other hand, we formulate first tentative recommendations for the design of multilayer networks, identify open issues in this context, and clarify possible future directions of research.

The visual exploration of large and complex network structures remains a challenge for many application fields. Moreover, a growing number of real world networks are multivariate and often interconnected with each other. Entities in a network may have relationships with elements of other related data sets, which do not necessarily have to be networks themselves, and these relationships may be defined by attributes that can vary greatly. In this work, we propose a comprehensive visual analytics approach that supports researchers to specify and subsequently explore attribute-based relationships across networks, text documents, and derived secondary data. Our approach provides an individual search functionality based on keywords and semantically similar terms over the entire text corpus to find related network nodes. For examining these nodes in the interconnected network views, we introduce a new interaction technique, called Hub2Go, which facilitates the navigation by guiding the user to the information of interest. To showcase our system, we use a large text corpus collected from research papers listed in the IEEE VIS publications dataset that consists of 2752 documents over a period of 25 years. Here, we analyze relationships between various heterogeneous networks, a Bag-of-Words index, and a word similarity matrix, all derived from the initial corpus and metadata. 

Heterogeneous networks and multi-level networks occur in several application fields where their integration, combination, comparison, analysis, and visualization poses major challenges. In this chapter, we analyze the general characteristics of this type of data and identify examples in three application domains: biology, social sciences, and software engineering. Conceptually, we focus on sets of multivariate networks at two or more levels. Each level may describe a specific scale, and within each level several related heterogeneous networks are represented. We allow n:m mappings within the same level, but only 1:n mappings across levels that must be consecutive. This leads to a structured data set that is the basis for further visual analysis. Our chapter ends with ideas to visualize those networks together with the relationships between them and highlights research challenges.