IEEE VIS Publication Dataset

Many real world graphs have small world characteristics, that is, they have a small diameter compared to the number of nodes and exhibit a local cluster structure. Examples are social networks, software structures, bibliographic references and biological neural nets. Their high connectivity makes both finding a pleasing layout and a suitable clustering hard. In this paper we present a method to create scalable, interactive visualizations of small world graphs, allowing the user to inspect local clusters while maintaining a global overview of the entire structure. The visualization method uses a combination of both semantical and geometrical distortions, while the layout is generated by a spring embedder algorithm using recently developed force model. We use a cross referenced database of 500 artists as a running example

The presented work aims to identify major research topics, co-authorships, and trends in the IV Contest 2004 dataset. Co-author, paper-citation, and burst analysis were used to analyze the dataset. The results are visually presented as graphs, static Pajek [1] visualizations and interactive network layouts using Pajek’s SVG output feature. A complementary web page with all the raw data, details of the analyses, and high resolution images of all figures is available online at http://iv.slis.indiana.edu/ref/iv04contest/.

In Web data, telecommunications traffic and in epidemiological studies, dense subgraphs correspond to subsets of subjects (i.e. users, patients) that share a collection of attributes values (i.e. accessed Web pages, email-calling patterns or disease diagnostic profiles). Visual and computational identification of these "clusters" becomes useful when domain experts desire to determine those factors of major influence in the formation of access and communication clusters or in the detection and contention of disease spread. With the current increases in graphic hardware capabilities and RAM sizes, it is more useful to relate graph sizes to the available screen real estate S and the amount of available RAM M, instead of the number of edges or nodes in the graph. We offer a visual interface that is parameterized by M and S and is particularly suited for navigation tasks that require the identification of subgraphs whose edge density is above certain threshold. This is achieved by providing a zoomable matrix view of the underlying data. This view is strongly coupled to a hierarchical view of the essential information elements present in the data domain. We illustrate the applicability of this work to the visual navigation of cancer incidence data and to an aggregated sample of phone call traffic

We describe an exploratory technique based on the direct interaction with a 2D modified scatterplot computed from two different metrics calculated over the elements of a network. The scatterplot is transformed into an image by applying standard image processing techniques resulting into blurring effects. Segmentation of the image allow to easily select patches on the image as a way to extract subnetworks. We were inspired by the work of Wattenberg and Fisher [M. Wattenberg et al. (2003)] showing that the blurring process builds into a multiscale perceptual scheme, making this type of interaction intuitive to the user. We explain how the exploration of the network can be guided by the visual analysis of the blurred scatterplot and by its possible interpretations

In this paper, we describe the process and result of creating a visualization to capture the past 10 years of history in the field of Information Visualization, as part of the annual InfoVis Conference Contest. We began with an XML file containing data provided by the contest organizers, scrubbed and augmented the data, and created a database to hold the information. We designed a visualization and implemented it using Flash MX 2004 Professional with ActionScript 2.0, PHP, and PostgreSQL. The resulting visualization provides an overview of the field of Information Visualization, and allows users to see the connections between areas of the field, particular researchers, and documents.

We present a method by which force-directed algorithms for graph layouts can be generalized to calculate the layout of a graph in an arbitrary Riemannian geometry. The method relies on extending the Euclidean notions of distance, angle, and force-interactions to smooth nonEuclidean geometries via projections to and from appropriately chosen tangent spaces. In particular, we formally describe the calculations needed to extend such algorithms to hyperbolic and spherical geometries

Color is often used to convey information, and color compositing is often required while visualizing multiattribute information. This paper proposes an alternative method for color compositing. In order to present understandable color blending to the general public, several techniques are proposed. First, a paint-inspired RYB color space is used. In addition, noise patterns are employed to produce subregions of pure color within an overlapped region. We show examples to demonstrate the effectiveness of our technique for visualization

We explore the use of nature’s phyllotactic patterns to inform the layout of hierarchical data. These naturally occurring patterns provide a non-overlapping, optimal packing when the total number of nodes is not known a priori. We present a family of expandable tree layouts based on these patterns.

This paper addresses the problem of how to enable users to visually explore and compare large sets of documents that have been retrieved by different search engines or queries. The Rank-Spiral enables users to rapidly scan large numbers of documents and their titles in a single screen. It uses a spiral mapping that maximizes information density and minimizes occlusions. It solves the labeling problem by exploiting the structure of the special spiral mapping used. Focus+Context interactions enable users to examine document clusters or groupings in more detail.

In many application domains, data is collected and referenced by its geospatial location. Nowadays, different kinds of maps are used to emphasize the spatial distribution of one or more geospatial attributes. The nature of geospatial statistical data is the highly nonuniform distribution in the real world data sets. This has several impacts on the resulting map visualizations. Classical area maps tend to highlight patterns in large areas, which may, however, be of low importance. Cartographers and geographers used cartograms or value-by-area maps to address this problem long before computers were available. Although many automatic techniques have been developed, most of the value-by-area cartograms are generated manually via human interaction. In this paper, we propose a novel visualization technique for geospatial data sets called RecMap. Our technique approximates a rectangular partition of the (rectangular) display area into a number of map regions preserving important geospatial constraints. It is a fully automatic technique with explicit user control over all exploration constraints within the exploration process. Experiments show that our technique produces visualizations of geospatial data sets, which enhance the discovery of global and local correlations, and demonstrate its performance in a variety of applications

This interactive poster proposes a novel, explorative way to browse a database containing links to resource systems-related information online. Our approach is an illustrative one, and draws on our combined backgrounds in computer science, graphic and interaction design, sustainability, community organization, and urban design. The data visualized in our prototype was collected by students in the course Sustainable Habits, which Lauren Dietrich taught at Stanford University during Winter 2004.

We present the novel high-level visualization taxonomy. Our taxonomy classifies visualization algorithms rather than data. Algorithms are categorized based on the assumptions they make about the data being visualized; we call this set of assumptions the design model. Because our taxonomy is based on design models, it is more flexible than existing taxonomies and considers the user's conceptual model, emphasizing the human aspect of visualization. Design models are classified according to whether they are discrete or continuous and by how much the algorithm designer chooses display attributes such as spatialization, timing, colour, and transparency. This novel approach provides an alternative view of the visualization field that helps explain how traditional divisions (e.g., information and scientific visualization) relates and overlap, and that may inspire research ideas in hybrid visualization areas

Current implementations of multidimensional scaling (MDS), an approach that attempts to best represent data point similarity in a low-dimensional representation, are not suited for many of today's large-scale datasets. We propose an extension to the spring model approach that allows the user to interactively explore datasets that are far beyond the scale of previous implementations of MDS. We present MDSteer, a steerable MDS computation engine and visualization tool that progressively computes an MDS layout and handles datasets of over one million points. Our technique employs hierarchical data structures and progressive layouts to allow the user to steer the computation of the algorithm to the interesting areas of the dataset. The algorithm iteratively alternates between a layout stage in which a subselection of points are added to the set of active points affected by the MDS iteration, and a binning stage which increases the depth of the bin hierarchy and organizes the currently unplaced points into separate spatial regions. This binning strategy allows the user to select onscreen regions of the layout to focus the MDS computation into the areas of the dataset that are assigned to the selected bins. We show both real and common synthetic benchmark datasets with dimensionalities ranging from 3 to 300 and cardinalities of over one million points

In today's fast-paced world it is becoming increasingly difficult to stay abreast of the public discourse. With the advent of hundreds of closed-captioned cable channels and internet-based channels such as news feeds, blogs, or email, knowing the "buzz" is a particular challenge. TextPool addresses this problem by quickly summarizing recent content in live text streams. The summarization is a dynamically changing textual collage that clusters related terms. We tested TextPool with the content of several RSS newswire feeds, which are updated roughly every five minutes. TextPool was able to handle this bandwidth well, and produced useful summarizations of feed content.

This article presents the InfoVis toolkit, designed to support the creation, extension and integration of advanced 2D information visualization components into interactive Java swing applications. The InfoVis toolkit provides specific data structures to achieve a fast action/feedback loop required by dynamic queries. It comes with a large set of components such as range sliders and tailored control panels required to control and configure the visualizations. These components are integrated into a coherent framework that simplifies the management of rich data structures and the design and extension of visualizations. Supported data structures currently include tables, trees and graphs. Supported visualizations include scatter plots, time series, parallel coordinates, treemaps, icicle trees, node-link diagrams for trees and graphs and adjacency matrices for graphs. All visualizations can use fisheye lenses and dynamic labeling. The InfoVis toolkit supports hardware acceleration when available through Agile2D, an implementation of the Java graphics API based on OpenGL, achieving speedups of 10 to 200 times. The article also shows how new visualizations can be added and extended to become components, enriching visualizations as well as general applications

This research demonstrates how principles of self-organization and behavior simulation can be used to represent dynamic data evolutions by extending the concept of information flocking, originally introduced by Proctor & Winter (1998), to time-varying datasets. A rule-based behavior system continuously controls and updates the dynamic actions of individual, three-dimensional elements that represent the changing data values of reoccurring data objects. As a result, different distinguishable motion types emerge that are driven by local interactions between the spatial elements as well as the evolution of time-varying data values. Notably, this representation technique focuses on the representation of dynamic data alteration characteristics, or how reoccurring data objects change over time, instead of depicting the exact data values themselves. In addition, it demonstrates the potential of motion as a useful information visualization cue. The original information flocking approach is extended to incorporate time-varying datasets, live database querying, continuous data streaming, real-time data similarity evaluation, automatic shape generation and more stable flocking algorithms. Different experiments prove that information flocking is capable of representing short-term events as well as long-term temporal data evolutions of both individual and groups of time-dependent data objects. An historical stock market quote price dataset is used to demonstrate the algorithms and principles of time-varying information flocking

Graph drawing is a basic visualization tool. For graphs of up to hundreds of nodes and edges, there are many effective techniques available. At greater scale, data density and occlusion problems often negate its effectiveness. Conventional pan-and-zoom, and multiscale and geometric fisheye views are not fully satisfactory solutions to this problem. As an alternative, we describe a topological zooming method. It is based on the precomputation of a hierarchy of coarsened graphs, which are combined on the fly into renderings with the level of detail dependent on the distance from one or more foci. We also discuss a related distortion method that allows our technique to achieve constant information density displays

We present a model and prototype system for tracking user interactions within a visualization. The history of the interactions are exposed to the user in a way that supports non-linear navigation of the visualization space. The interactions can be augmented with annotations, which, together with the interactions, can be shared with other users and applied to other data in a seamless way. The techniques constitute a novel approach for documenting information provenance.