Due to their great performance in many challenges, Deep Learning (DL) techniques keep gaining popularity in many fields. They have been adapted to process graph data structures to solve various complicated tasks such as graph classification and edge prediction. Eventually, they reached the Graph Drawing (GD) task. This article is an extended version of the previously published(DNN)2and presents a framework to leverage DL techniques for graph drawing (DL4GD). We demonstrate how it is possible to train a Deep Learning model to extract features from a graph and project them into a graph layout. The method proposes to leverage efficient Convolutional Neural Networks, adapting them to graphs using Graph Convolutions. The graph layout projection is learned by optimizing a cost function that does not require any ground truth layout, as opposed to prior work. This paper also proposes an implementation and benchmark of the framework to study its sensitivity to certain deep learning-related conditions. As the field is novel, and many questions remain to be answered, we do not focus on finding the most optimal implementation of the method, but rather contribute toward a better understanding of the approach potential. More precisely, we study different learning strategies relative to the models training datasets. Finally, we discuss the main advantages and limitations of DL4GD.

In 2D visualizations, visibility of every datum's representation is crucial to ease the completion of visual tasks. Such a guarantee is barely respected in complex visualizations, mainly because of overdraws between datum representations that hide parts of the information (e.g., outliers). The literature proposes various Layout Adjustment algorithms to improve the readability of visualizations that suffer from this issue. Manipulating the data in high-dimensional, geometric or visual space; they rely on different strategies with their own strengths and weaknesses. Moreover, most of these algorithms are computationally expensive as they search for an exact solution in the geometric space and do not scale well to large datasets. This article proposes GIST, a layout adjustment algorithm that aims at optimizing three criteria: (i) node visibility guarantee (at least 1 pixel), (ii) node size maximization, and (iii) the original layout preservation. This is achieved by combining a search for the maximum node size that enables to draw all the data points without overlaps, with a limited budget of movements (i.e., limiting the distortions of the original layout). The method's basis relies on the idea that it is not necessary for two data representations to be strictly not overlapping in order to guarantee their visibility in visual space. Our algorithm therefore uses a tolerance in the geometric space to determine the overlaps between pairs of data. The tolerance is optimized such that the approximation computed in the geometric space can lead to visualization without noticeable overdraw after the data rendering rasterization. In addition, such an approximation helps to ease the algorithm's convergence as it reduces the number of constraints to resolve, enabling it to handle large datasets. We demonstrate the effectiveness of our approach by comparing its results to those of state-of-the-art methods on several large datasets.

In many 2D visualizations, data points are projected without considering their surface area, although they are often represented as shapes in visualization tools. These shapes support the display of information such as labels or encode data with size or color. However, inappropriate shape and size selections can lead to overlaps that obscure information and hinder the visualization's exploration. Overlap Removal (OR) algorithms have been developed as a layout post-processing solution to ensure that the visible graphical elements accurately represent the underlying data. As the original data layout contains vital information about its topology, it is essential for OR algorithms to preserve it as much as possible. This article presents an extension of the previously published FORBID algorithm by introducing a new approach that models OR as a joint stress and scaling optimization problem, utilizing efficient stochastic gradient descent. The goal is to produce an overlap-free layout that proposes a compromise between compactness (to ensure the encoded data is still readable) and preservation of the original layout (to preserve the structures that convey information about the data). Additionally, this article proposes SORDID, a shape-aware adaptation of FORBID that can handle the OR task on data points having any polygonal shape. Our approaches are compared against state-of-the-art algorithms, and several quality metrics demonstrate their effectiveness in removing overlaps while retaining the compactness and structures of the input layouts.

The rise of Deep Learning (DL) has led to a breakthrough in the research field of content-based multimedia indexing. Newly developed systems based on complex models outperform classic machine learning algorithms in object detection, image segmentation or classification tasks. However, despite their high performance, these systems still make mistakes. To be used in industrial conditions, these systems must be able to provide trustworthy decisions with guarantees or justifications. Therefore, it is crucial to provide means to analyze and comprehend the decision process that leads a model to its decision. Image classification implies tracking and understanding which input features the model relies on to make its prediction. This paper focuses on features attribution techniques and proposes Heatmaps by Hierarchical Occlusion (H2O), a novel method for detecting pattern-relevant features in an image. We also propose two new pairs of metrics that overcome some evaluation issues: (a) Insertion and Deletion Spearman correlation coefficients which both estimate a correlation between the computed scores in a saliency map and the importance for the model of the associated pixels in the image. (b) Insertion Positive and Deletion Negative Gradient Sums both estimate the coherence of the scores in the saliency maps. Both visual inspection and evaluation on 7 metrics show that H2O is competitive against state-of-the-art methods.

Abstract The use and creation of machine‐learning‐based solutions to solve problems or reduce their computational costs are becoming increasingly widespread in many domains. Deep Learning plays a large part in this growth. However, it has drawbacks such as a lack of explainability and behaving as a black‐box model. During the last few years, Visual Analytics has provided several proposals to cope with these drawbacks, supporting the emerging eXplainable Deep Learning field. This survey aims to (i) systematically report the contributions of Visual Analytics for eXplainable Deep Learning; (ii) spot gaps and challenges; (iii) serve as an anthology of visual analytical solutions ready to be exploited and put into operation by the Deep Learning community (architects, trainers and end users) and (iv) prove the degree of maturity, ease of integration and results for specific domains. The survey concludes by identifying future research challenges and bridging activities that are helpful to strengthen the role of Visual Analytics as effective support for eXplainable Deep Learning and to foster the adoption of Visual Analytics solutions in the eXplainable Deep Learning community. An interactive explorable version of this survey is available online at https://aware‐diag‐sapienza.github.io/VA4XDL .

Current deep learning approaches are cutting-edge methods for solving classification tasks. Arising transfer learning techniques allows applying large generic model to simple tasks whereas simpler models could be used. Large models raise the major problem of their memory consumption and processor usage and lead to a prohibitive ecological footprint. In that paper, we present a novel visual analytics approach to interactively prune those networks and thus limit that issue. Our technique leverages a novel sparkline matrix visualization technique as well as a novel local metric which evaluates the discriminatory power of a filter to guide the pruning process and make it interpretable. We assess the well- founded of our approach through two realistic case studies and a user study. For both of them, the interactive refinement of the model led to a significantly smaller model having similar prediction accuracy than the original one.

Projecting a set of$n$points on a grid of size$\sqrt{n}\times\sqrt{n}$provides the best possible information density in two dimensions without overlap. We leverage the Voronoi Relaxation method to devise a novel and versatile post-processing algorithm called VRGrid: it enables the arrangement of any 2D data on a grid while preserving its initial positions. We apply VRGrid to generate compact and overlap-free visualization of popular and overlap-prone projection methods (e.g., t-SNE). We prove that our method complexity is$O(\sqrt{n}.i.n.log(n))$, with i a determined maximum number of iterations and$n$the input dataset size. It is thus usable for visualization of several thousands of points. We evaluate VRGrid's efficiency with several metrics: distance preservation (DP), neighborhood preservation (NP), pairwise relative positioning preservation (RPP) and global positioning preservation (GPP). We benchmark VRGrid against two state-of-the-art methods: Self-Sorting Maps (SSM) and Distance-preserving Grid (DGrid). VRGrid outperforms these two methods, given enough iterations, on DP, RPP and GPP which we identify to be the key metrics to preserve the positions of the original set of points.

Current machine learning and deep learning approaches are cutting-edge methods for solving classification tasks. Comparing the performances of classification models has become a prominent task since the outbreak of these techniques. The performance of such classification models is measured by the ratio between the correctly predicted samples and the others. The most widely used visualization to represent this information is the Confusion matrix. Yet, if this technique is suited to apprehend one model performances, very few works use this representation to compare models. In that paper, we present the Relative Confusion Matrix (RCM), a new matrix visualization that leverages Confusion matrices and a color encoding to expose the class-wise differences of performances between two models. We conduct a user evaluation to compare RCM with two confusion matrix variants. Our results show that RCM encoding leads to a more efficient comparison of two models than existing approaches.

The design of efficient representations is well established as a fruitful way to explore and analyze complex or large data. In these representations, data are encoded with various visual attributes depending on the needs of the representation itself. To make coherent design choices about visual attributes, the visual search field proposes guidelines based on the human brain’s perception of features. However, information visualization representations frequently need to depict more data than the amount these guidelines have been validated on. Since, the information visualization community has extended these guidelines to a wider parameter space. This paper contributes to this theme by extending visual search theories to an information visualization context. We consider a visual search task where subjects are asked to find an unknown outlier in a grid of randomly laid out distractors. Stimuli are defined by color and shape features for the purpose of visually encoding categorical data. The experimental protocol is made of a parameters space reduction step (i.e., sub-sampling) based on a machine learning model, and a user evaluation to validate hypotheses and measure capacity limits. The results show that the major difficulty factor is the number of visual attributes that are used to encode the outlier. When redundantly encoded, the display heterogeneity has no effect on the task. When encoded with one attribute, the difficulty depends on that attribute heterogeneity until its capacity limit (7 for color, 5 for shape) is reached. Finally, when encoded with two attributes simultaneously, performances drop drastically even with minor heterogeneity.

In information visualization, it has become mandatory to assess visualization techniques efficiency either to write a survey, optimize a technique or even design a new one. To do so, the common way is to conduct user evaluations through which human subjects are asked to solve a task on different visualization techniques while their performances are measured to assess which technique is the most efficient. These evaluations can be complex to design and setup in order not to be biased and, in the end, their results can become contestable when the evaluation methods standards evolve. To overcome these flaws, new evaluation methods are emerging, mostly making use of modern and efficient computer vision techniques such as deep learning. These new methods rely on a strong assumption that has not been studied deeply enough yet: humans and deep learning models performances can be correlated. This paper explores the performances of both a state-of-the-art deep neural network and human subjects on an outlier detection task taken from a previous experiment of the literature. The objective is to study whether the machine and humans behaviors were different or if some correlations can be observed. Our study shows that their results are significantly correlated and a machine learning model efficiently learned to predict human performances using deep neural network metrics as input. Hence, this work presents a use case where using a deep neural network to assess human subjects performances is efficient.

Many end-user evaluations of data visualization techniques have been run during the last decades. Their results are cornerstones to build efficient visualization systems. However, designing such an evaluation is always complex and time-consuming and may end in a lack of statistical evidence and reproducibility. We believe that modern and efficient computer vision techniques, such as deep convolutional neural networks (CNNs), may help visualization researchers to build and/or adjust their evaluation hypothesis. The basis of our idea is to train machine learning models on several visualization techniques to solve a specific task. Our assumption is that it is possible to compare the efficiency of visualization techniques based on the performance of their corresponding model. As current machine learning models are not able to strictly reflect human capabilities, including their imperfections, such results should be interpreted with caution. However, we think that using machine learning-based pre-evaluation, as a pre-process of standard user evaluations, should help researchers to perform a more exhaustive study of their design space. Thus, it should improve their final user evaluation by providing it better test cases. In this paper, we present the results of two experiments we have conducted to assess how correlated the performance of users and computer vision techniques can be. That study compares two mainstream graph visualization techniques: node-link (NL) and adjacency-matrix (AM) diagrams. Using two well-known deep convolutional neural networks, we partially reproduced user evaluations from Ghoniem et al. and from Okoe et al.. These experiments showed that some user evaluation results can be reproduced automatically.

In the context of multiple views, coordination is essential to navigate and grasp the relationships lying behind the different juxtaposed views. Linked highlighting is a typical example of coordination where a subset of the data points is emphasized simultaneously on all views. The strength of this approach is that the selected data can be studied within its context. Other approaches have been used to implement coordination such as using varying levels of transparency or visual links. We propose to use spatial distortion to contribute a similar effect in multiple views. It is particularly suited to the context of multiple views since it alleviates the lack of screen space by reallocating it based on a certain definition of user interest. The proposed method targets coordination between views that represent the same entities and readily adapts to various visualization forms. It is based on a user degree-of-interest function, defined on these entities, that acts as a common ground for the distortion of all views. Views are distorted such that empty areas and areas holding entities of lesser interest are compressed to the benefit of areas holding entities of higher interest. To demonstrate its feasibility and versatility, we describe how to technically apply our approach to several common visualization techniques.

Nowadays, one of the main challenges in biology is to make use of several sources of data to improve our understanding of life. When analyzing experimental data, researchers aim at clustering genes that show a similar behavior through specific external conditions. Thus, the functional interpretation of genes is crucial and involves making use of the whole subset of terms that annotate these genes and which can be relatively large and redundant. The manual expertise to clearly decipher the main functions that may be related to the gene set is timeconsuming and becomes impracticable when the number of gene sets increases, like in the case of vaccine/drug trials. To overcome this drawback, it may be necessary to reduce the dataset with the aim to apply visualization approaches. In this paper, we propose a new pipeline combining enrichment and annotation terms simplification to produce a synthetic visualization of several gene sets simultaneously. We illustrate the efficiency of our method on a case study aiming at analyzing the immune response in diseases.

Heatmap visualization is a well-known type of visualization to alleviate the overplot problem of point visualization. As such, it is well suited to visualize Big Data. In order to tackle the velocity problem of Big Data, one has to leverage streaming computations. Recently, canopy clustering was shown to be well suited for Big Data heatmap visualization. In this article, we present how to design a streaming algorithm to compute canopy clustering using Apache Spark. This result is directly applicable to be included into a lambda architecture.

BACKGROUND: Bacterial sRNA-mediated regulatory networks has been introduced as a powerful way to analyze the fast rewiring capabilities of a bacteria in response to changing environmental conditions. The identification of mRNA targets of bacterial sRNAs is essential to investigate their functional activities. However, this step remains challenging with the lack of knowledge of the topological and biological constraints behind the formation of sRNA-mRNA duplexes. Even with the most sophisticated bioinformatics target prediction tools, the large proportion of false predictions may be prohibitive for further analyses. To deal with this issue, sRNA target analyses can be carried out from the resulting gene lists given by RNA-SEQ experiments when available. However, the number of resulting target candidates may be still huge and cannot be easily interpreted by domain experts who need to confront various biological features to prioritize the target candidates. Therefore, novel strategies have to be carried out to improve the specificity of computational prediction results, before proposing new candidates for an expensive experimental validation stage. RESULT: To address this issue, we propose a new visualization tool rNAV 2.0, for detecting and filtering bacterial sRNA targets for regulatory networks. rNAV is designed to cope with a variety of biological constraints, including the gene annotations, the conserved regions of interaction or specific patterns of regulation. Depending on the application, these constraints can be variously combined to analyze the target candidates, prioritized for instance by a known conserved interaction region, or because of a common function. CONCLUSION: The standalone application implements a set of known algorithms and interaction techniques, and applies them to the new problem of identifying reasonable sRNA target candidates.

Many real world information can be represented by a graph with a set of nodes interconnected with each other by multiple type of relations called edge layers (e.g., social network, biological data). Edge bundling techniques have been proposed to solve cluttering issue for standard graphs while few efforts were done to deal with the similar issue for multilayer graphs. In multilayer graphs scenario, not only the clutter induced by large amount of edges is a problem but also the fact that different type of edges can overlap each other making useless the final visualization. In this paper we introduce a new multilayer graph edge bundling technique that firstly produces a preliminary edge bundling independently of the different edge layers and then deals with the specificity of multilayer graphs where more than one type of edges can be routed on the same bundle. The proposed visualization is tested on a real world case study and the outcomes point out the ability of our proposal to discover patterns present in the data.

Biometric authentication systems suffer from several performance limitations. Many performance metrics exist to assess the overall performance of such systems. However, these metrics provide a quantitative assessment in terms of errors without explaining the reasons behind the set of users who significantly contributed for these errors. Towards contributing to solve this problem, we present a novel method (named Zoo Graph) to visualize the performance of a biometric system as a graph thanks to a database of recognition scores. Our approach is an improvement of the Zoo Plot and emphasizes on the relations between the individuals of the database and allows interactive manipulations to track these relations and understand why the biometric authentication method reacts this way. This graph provides researchers with an additional visual assessment tool that would identify problematic users. Such information would allow researchers to update their developed authentication algorithms to reduce those errors.

The central dogma in molecular biology postulated that 'DNA makes RNA makes protein', however this dogma has been recently extended to integrate new biological activities involving small non-coding RNAs, called sRNAs. In particular, it has been shown that sRNAs regulate the production of proteins by interacting on mRNAs to regulate positively or negatively their translations. That regulation of the mRNA translation is done by forming a base-pairing between the RNAs sequences of bases. In silico methods have been proposed by the bioinformatics community toprovide a list of putative interactions to be experimentally validated. However, such approaches suffers from a poor specificity and therefore produce a large number of false predictions. In this paper, we present a new visualization technique for sRNA-mRNA interactions emphasizing theinvolved regions on the sRNA secondary structure drawing. Our approach also supports interactive exploration as the user can select and highlight interactions. We demonstrate the usefulness of our approach by a case study on E. coli bacteria performed by domain experts.

The context of this paper is the labelling of a document image database in an industrial process. Our work focuses on the quality assessment of a given labelled database. In most practical cases, a database is manually labelled by an operator who has to browse sequentially the images (presented as thumbnails) until the whole database is labelled. This task is very repetitive; moreover the filing plan defining the names and number of classes is often incomplete, which leads to many labelling errors. The question is then to certify if the quality of a labelled batch is good enough to globally accept it. Our objective is to ease and speed up that evaluation that needs up to 1.5 more times than the labelling work itself. We propose an interactive tool for visualizing the data as a graph. That graph enhances similarities between documents as well as the labelling quality. We define criteria on the graph that characterize the three types of errors an operator can do: an image is mislabelled, one class should be split in more pertinent subclasses, several classes should be merged in another. This allows us to focus the operator attention on potential errors. He can then count the errors encountered while auditing the database and assess (or not) the global labelling quality.

Graph is a powerful tool to model relationships between elements and has been widely used in different research areas. Size and complexity of newly acquired graphs prohibit manual representations and urge a need for automatic visualization methods. We are interested with the node-links diagram which represents each node as a glyph and edge as a line between the corresponding nodes. % We present a novel layout algorithm that emphasizes the cores of very large networks (up to several hundred thousand of nodes and million of edges) in few seconds or minutes. Our method uses a hierarchical coreness decomposition of the graph and a combination of existing layout algorithms according to the clusters topologies. Area-aware drawing algorithms which produce node overlap-free drawings are used to reduce the visual clutter. Edges are bundled along the hierarchy of clusters to highlight the network communities and reduce edge visual clutter. % We validated our approach by comparing our method against one of the fastest method of the state of the art on a benchmark of 23 large graphs extracted from various sources. We have statistically proved that our method performs faster while providing meaningful results.

The most common depictions of graphs are node-link diagrams (NLDs) and matrix-based diagrams (MBDs). Making valid comparisons between these two visualisation techniques is difficult because they are each subject to a variety of representation parameters with respect to graph layout (NLD) and node ordering (MBD), meaning that any given choice of layout and order (even if they fulfil some aesthetic criteria) may influence experimental results. To overcome this problem, we propose a MBD-based technique which hybridises the entity visual encoding of a MBD with the edge visual encoding of a NLD. Using a typical MBD, we propose three edge visual encoding evolutions to ultimately render edges like in a NLD while preserving nodes depiction and order. Such encoding evolutions allow us to perform an experimental evaluation of user performances for a path finding task without the above limitations. We show that for a path finding task, our edge visual encoding evolutions tend to improve the user experience when analysing and interacting with a MBD.

Visualization of hierarchical weighted and directed graphs are usually done with node-link or adjacency matrix diagrams. However, these representations suffer from various drawbacks: low readability in a context of Big Data, high number of edge crossings, difficulty to efficiently represent the weighting. With the stated goal of reducing these drawbacks, we designed Adjasankey, a hybrid visual representation of weighted and directed graphs using hierarchical abstractions. This technique combines adjacency matrices readability of large graphs and flow diagrams visual design efficiency for weighting depiction. Associated to Big Data computing and light-weight web rendering, our tool allows to depict and interact in real time on huge dataset and supports user multi-scale exploration and analysis. To show the efficiency of Adjasankey, we present a case study on the analysis of a Customer to Customer website.

The revolution in high-throughput sequencing technologies has enabled the acquisition of gigabytes of RNA sequences in many different conditions and has highlighted an unexpected number of small RNAs (sRNAs) in bacteria. Ongoing exploitation of these data enables numerous applications for investigating bacterial transacting sRNA-mediated regulation networks. Focusing on sRNAs that regulate mRNA translation in trans, recent works have noted several sRNA-based regulatory pathways that are essential for key cellular processes. Although the number of known bacterial sRNAs is increasing, the experimental validation of their interactions with mRNA targets remains challenging and involves expensive and time-consuming experimental strategies. Hence, bioinformatics is crucial for selecting and prioritizing candidates before designing any experimental work. However, current software for target prediction produces a prohibitive number of candidates because of the lack of biological knowledge regarding the rules governing sRNA-mRNA interactions. Therefore, there is a real need to develop new approaches to help biologists focus on the most promising predicted sRNA-mRNA interactions. In this perspective, this review aims at presenting the advantages of mixing bioinformatics and visualization approaches for analyzing predicted sRNA-mediated regulatory bacterial networks.

Technological advances in biology lead to a profusion of quantitative data, raising analytical challenges. Visual analytics is particularly well suited to address these difficulties. It helps to interactively move through the different levels of analysis and to simultaneously investigate data with different point of views. It is especially the case when dealing with biological networks that can contain hundreds of elements. In these studies biologists generally follow the same analytic process which consists in first getting an overview of the data before focussing on a few relevant subnetworks. In this article we present, Systrip, a visual environment for the analysis of time-series data in the context of biological networks. In particular we focus on the study of metabolism. Systrip gathers bioinformatics and graph theoretical algorithms that can be assembled in different ways to help biologists in their visual mining process. This framework had been used to analyse various real biological data. In this article we describe how it helped in understanding drug effects on the metabolism of the parasite of the tsetse fly causing sleeping sickness.

Pattern discovery plays an important part in the graph analysis process. Good examples are the detection of communities in social networks or the clustering into pathways of metabolic networks. However, elements may be shared by several clusters, making the patterns entangled. When mining such data, experts are usually interested in both each individual cluster and their overlaps. Dedicated visualization methods are therefore necessary to efficiently support their exploration process. In this article, we propose a new method that emphasizes patterns in a node-link diagram representation and allows to easily identify overlaps between these patterns as well. Our technique combines graph topology and embedding to compute concave hulls with holes surrounding the patterns of interest.

Abstract Improvements in biological data acquisition and genomes sequencing now allow to reconstruct entire metabolic networks of many living organisms. The size and complexity of these networks prohibit manual drawing and thereby urge the need of dedicated visualization techniques. An efficient representation of such a network should preserve the topological information of metabolic pathways while respecting biological drawing conventions. These constraints complicate the automatic generation of such visualization as it raises graph drawing issues. In this paper we propose a method to lay out the entire metabolic network while preserving the pathway information as much as possible. That method is flexible as it enables the user to define whether or not node duplication should be performed, to preserve or not the network topology. Our technique combines partitioning, node placement and edge bundling to provide a pseudo‐orthogonal visualization of the metabolic network. To ease pathway information retrieval, we also provide complementary interaction tools that emphasize relevant pathways in the entire metabolic context.

Abstract PrEd [ Ber00 ] is a force‐directed algorithm that improves the existing layout of a graph while preserving its edge crossing properties. The algorithm has a number of applications including: improving the layouts of planar graph drawing algorithms, interacting with a graph layout, and drawing Euler‐like diagrams. The algorithm ensures that nodes do not cross edges during its execution. However, PrEd can be computationally expensive and overly‐restrictive in terms of node movement. In this paper, we introduce ImPrEd: an improved version of PrEd that overcomes some of its limitations and widens its range of applicability. ImPrEd also adds features such as flexible or crossable edges, allowing for greater control over the output. Flexible edges, in particular, can improve the distribution of graph elements and the angular resolution of the input graph. They can also be used to generate Euler diagrams with smooth boundaries. As flexible edges increase data set size, we experience an execution/drawing quality trade off. However, when flexible edges are not used, ImPrEdproves to be consistently faster than PrEd.

Visualization of graphs containing many nodes and edges efficiently is quite challenging since representations generally suffer from visual clutter induced by the large amount of edge crossings and node-edge overlaps. That problem becomes even more important when nodes positions are fixed, such as in geography were nodes positions are set according to geographical coordinates. Edge bundling techniques can help to solve this issue by visually merging edges along common routes but it can also help to reveal high-level edge patterns in the network and therefore to understand its overall organization. In this paper, we present a generalization of [18] to reduce the clutter in a 3D representation by routing edges into bundles as well as a GPU-based rendering method to emphasize bundles densities while preserving edge color. To visualize geographical networks in the context of the globe, we also provide a new technique allowing to bundle edges around and not across it.

Abstract Visualizing graphs containing many nodes and edges efficiently is quite challenging. Drawings of such graphs generally suffer from visual clutter induced by the large amount of edges and their crossings. Consequently it is difficult to read the relationships between nodes and the high‐level edge patterns that may exist in standard node‐link diagram representations. Edge bundling techniques have been proposed to help solve this issue, which rely on high quality edge rerouting. In this paper, we introduce an intuitive edge bundling technique which efficiently reduces edge clutter in graphs drawings. Our method is based on the use of a grid built using the original graph to compute the edge rerouting. In comparison with previously proposed edge bundling methods, our technique improves both the level of clutter reduction and the computation performance. The second contribution of this paper is a GPU‐based rendering method which helps users perceive bundles densities while preserving edge color.

Community detection in social networks varying with time is a common yet challenging problem whereby efficient visualization of evolving relationships and implicit hierarchical structure are important task. The main contribution of this paper is towards establishing a framework to analyze such social networks. The proposed framework is based on dynamic graph discretization and graph clustering.The framework allows detection of major structural changes over time, identifies events analyzing temporal dimension and reveals command hierarchies in social networks.We use the Catalano/Vidro dataset for empirical evaluation and observe that our framework provides a satisfactory assessment of the social and hierarchical structure present in the dataset.

In this paper, we present an algorithm to lay out a particular class of graphs coming from real case studies: the quasi-tree graph class. Protein and internet mappings projects have shown the interest of devicing dedicated tools for visualizing such graphs. Our method addresses a challenging problem which consists in computing a layout of large graphs (up to hundred of thousands of nodes) that emphasizes their tree-like property in an efficient time. In order to validate our approach, we compare our results on real data to those obtained by well known algorithms.

We present an application for integrated visualization of gene expression data from time series experiments in gene regulation networks and metabolic networks. Such integration is necessary, since it provides the link between the measurements at the transcriptional level and the observable characteristics of an organism at the functional level. Our application can (i) visualize the data from time series experiments in the context of a regulatory network and a metabolic network; (ii) identify and visualize active regulatory subnetworks from the gene expression data; (iii) perform a statistical test to identify and subsequently visualize affected metabolic subnetworks. Initial results show that our integrated approach speeds up data analysis, and that it can reproduce results of a traditional approach that involves many manual and time-consuming steps.

This article is addressing a recurrent problem in biology: mining newly built large scale networks. Our approach consists in comparing these new networks to well known ones. The visual backbone of this comparative analysis is provided by a network classification hierarchy. This method makes sense when dealing with metabolic networks since comparison could be done using pathways (clusters). Moreover each network models an organism and it exists organism classification such as taxonomies. Video demonstration: http://www.labri.fr/perso/bourqui/video.wmv.

Visualization of clustered graphs has been a research area since many years. In this paper, we describe a new approach that can be used in real application where graph does not contain only topological information but also extrinsic parameters (i.e. user attributes on edges and nodes). In the case of force-directed algorithm, management of attributes corresponds to take into account edge weights. We propose an extension of the GRIP algorithm in order to manage edge weights. Furthermore, by using Voronoi diagram we constrained that algorithm to draw each cluster in a non overlapping convex region. Using these two extensions we obtained an algorithm that draw clustered weighted graphs. Experimentation has been done on data coming from biology where the network is the genes- proteins interaction graph and where the attributes are gene expression values from microarray experiments.

A metabolic network is a set of interconnected metabolic pathways (subnetworks). Until recently, metabolic studies were dedicated to a single pathway, but current researches now consider the entire network. As matter stands, existing visualization tools cannot be used to undertake these global studies since they have been designed to probe metabolic pathways. For the purpose of making it feasible, this paper presents a graph drawing algorithm for the whole metabolic network. Our collaboration with biologists led us to introduce drawing constraints which take into account the decomposition of the network into metabolic pathways as well as biochemical textbook drawing conventions. These constraints raise numerous graph drawing problems which are solved by first recursively decomposing the network then applying suitable graph drawing algorithms. Finally, we present an application that illustrates the advantage of this representation when visualizing groups of reactions which span several metabolic pathways.