IVAPP 2011 Abstracts


Full Papers
Paper Nr: 8
Title:

VISUAL ANALYSIS OF DOCUMENT TRIAGE DATA

Authors:

Zhao Geng, Robert S. Laramee, Fernando Loizides and George Buchanan

Abstract: As part of the information seeking process, a large amount of effort is invested in order to study and understand how information seekers search through documents such that they can assess their relevance. This search and assessment of document relevance, known as document triage, is an important information seeking process, but is not yet well understood. Human-computer interaction (HCI) and digital library scientists have undertaken a series of user studies involving information seeking, collected a large amount of data describing information seekers’ behavior during document search. Next to this, we have witnessed a rapid increase in the number of off-the-shelf visualization tools which can benefit document triage study. Here we set out to utilize existing information visualization techniques and tools in order to gain a better understanding of the large amount of user-study data collected by HCI and digital library researchers. We describe the range of available tools and visualizations we use in order to increase our knowledge of document triage. Treemap, parallel coordinates, stack graph, matrix chart, as well as other visualization methods, prove to be insightful in exploring, analyzing and presenting user behavior during document triage. Our findings and visualizations are evaluated by HCI and digital library researchers studying this problem.
Download

Paper Nr: 12
Title:

MULTIMODAL SEARCH FOR GRAPHIC DESIGNERS

Authors:

Sandra Skaff, David Rouquet, Emmanuel Dellandrea, Achille Falaise, Valérie Bellynck, Hervé Blanchon, Christian Boitet, Didier Schwab, Liming Chen, Alexandre Saidi, Gabriela Csurka and Luca Marchesotti

Abstract: This paper describes OMNIA, a system and interface for searching in multimodal image collections. OMNIA includes a set of tools which allow the user to retrieve assets using different features. The tools are based on extracting different types of asset features, which are content, aesthetic, and emotion. Visual-based features are used to retrieve assets using each of these tools. In addition, text-based features can be used to retrieve image assets based on content. Different datasets are used in OMNIA and retrieved assets are displayed in such a way which facilitates user navigation. It is shown how OMNIA can be used for simple, efficient, and intuitive asset search in the context of graphic design applications.
Download

Paper Nr: 13
Title:

VISUALIZING DYNAMIC QUANTITATIVE DATA IN HIERARCHIES - TimeEdgeTrees: Attaching Dynamic Weights to Tree Edges

Authors:

Michael Burch and Daniel Weiskopf

Abstract: In this paper we introduce a technique for visualizing the dynamics of quantitative data in static hierarchical structures. We exploit the straight links of orthogonal tree diagrams as a timeline on which we visually encode dynamic quantitative information. We use color coding and varying thicknesses to represent the time-varying data. Bimodal data can also be displayed by exploiting both sides of the time axes simultaneously. Our TimeEdgeTrees tool allows us to explore dynamic quantitative data in tree diagrams by interactive data filtering and zooming. The spatial proximity of neighboring hierarchically structured elements allows us to easily explore trends, countertrends, periodicity, temporal shifts, or anomalies during the evolution synchronously. Interactive features such as expanding or collapsing of subhierarchies additionally help to detect the aforementioned phenomena on different levels of granularity. The usefulness of our visualization technique is illustrated by water level data acquired at more than 450 measurement stations along German rivers for 768 points in time.
Download

Paper Nr: 14
Title:

DEVELOPMENT PROCESS AND EVALUATION OF THE ONTOLOGY VISUALIZATION TOOL KNOOCKS - A Case Study

Authors:

Simone Kriglstein and Günter Wallner

Abstract: Ontologies were becoming popular in various communities and disciplines over the past few years. The reason for their popularity is that they allow to specify concepts and their interconnections of a domain which are understandable by computers and by people. However without visualizations, ontologies - especially large ones - would be hard to comprehend by users. Visualizations are essential because they make ontologies with their structure more manageable. Furthermore, visualization tools are helpful to analyze the interconnections between concepts. For an effective usage of such visualization tools, it is necessary that the design is easy to understand, provides a good overview about the structure and allows fast access to detail information. Furthermore, it is necessary to consider the domain for which the ontology is developed along with its users and their needs. This paper presents the development process of Knoocks – a visualization tool for OWL Lite ontologies – where users were already involved during the process to improve the usability of the tool. Moreover, the presented case study underlines the importance of repeated usability evaluations during the development process to identify weak points of the design and missing features which are relevant for the intended users.
Download

Paper Nr: 26
Title:

MULTI-LAYER DISTORTED 1D NAVIGATION

Authors:

Michael Wörner and Thomas Ertl

Abstract: We propose a novel interactive user interface for the efficient, one-dimensional navigation in massive, ordered datasets. Applying a multi-layered, distorted visualization we provide both context and detail information about the investigated data and allow for both large scale and precise navigation. Adjacent layers display the data at gradually varying scales. This creates a visual relation between data points at different levels of detail and enables the user to browse through the data at varying speeds. Thereby, a sense of global context is retained, while at the same time detail information about the current focus position is kept in view. We believe that this approach is suitable for the investigation of massive, ordered datasets such as time-based data series, lists, or tables and discuss its possible applications in an example scenario.
Download

Short Papers
Paper Nr: 5
Title:

TEMPORAL SOCIAL NETWORK ANALYSIS FOR HISTORIANS - A Case Study

Authors:

John Haggerty and Sheryllynne Haggerty

Abstract: Social network analysis has received much attention across disciplines. Recently, historians have begun to complicate their understanding of networks and are increasingly using visualisation to elucidate tangible and intangible information in their data sets. Current approaches make use of static social network analysis and whilst they provide a number of tools to explore a social network, they are unable to provide temporal analysis. This paper presents Matrixify, a practical visual application for the exploratory temporal analysis of social networks for historians and others not familiar with visual representations of data. This approach aims to deconstruct the complexity of social networks to provide a temporal analysis to answer a key problem in historical studies, that of ‘analysis of change over time’. In this way, historians are able to identify real relationships in their data sets; actors in contact at a particular point in time and shown over time. The case study presented in this paper demonstrates the applicability of the approach in inter-disciplinary studies.
Download

Paper Nr: 10
Title:

CREATING AUDIENCE SPECIFIC GALACTIC SIMULATIONS USING EYE-TRACKING TECHNOLOGY

Authors:

Janelle Arita, Jenniffer Feliz, Dennis Rodriguez, Hans-Peter Bischof, Manjeet Rege and Reynold Bailey

Abstract: This study uses eye-tracking to investigate the differences in the way professional astrophysicists and novices observe simulations of galactic events. The results of this study provide insight into which aspects of the data are important and allow us to tailor the visualizations for a specific group. We hypothesized that the gaze patterns of professionals and novices would vary considerably. A user study was performed on two groups: trained astrophysicists and novices. Each group was presented with a randomized sequence of images and a video while their gaze patterns were recorded with an eye-tracker. We discovered that although both groups observed each image for the same duration, experts limited their fixations to a smaller area. Novices, on the other hand, had fixations which were spread across the images. For the video, the astrophysicists were more focused on simulations in which most of the data was visible and the camera angles had minimal axis change.
Download

Paper Nr: 11
Title:

TOWARDS AFFORDABLE GESTURE BASED INTERFACES - An Exploration with Wii Remotes

Authors:

Amy Ciavolino, Camille Marvin, Jason Creighton, James Coddington, Hans-Peter Bischof and Reynold Bailey

Abstract: The traditional keyboard and mouse computer interface is well suited for 2D applications such as document editing, but as 3D environments become more prevalent there arises a need for new methods of user input. An immersive 3D interface is preferable, but is often cost prohibitive. This paper presents the design and implementation of a Natural User Interface (NUI), named NuWii. NuWii was designed to be used with the Spiegel visualization framework for astrophysical data. The interface makes use of Nintendo Wii Remotes as infrared tracking cameras to detect 3D gestures made by the user. These gestures are interpreted and used to control the viewing camera’s parameters throughout the course of a visualization. Our project provides a new, more intuitive way of manipulating the camera in the Spiegel visualization system. Furthermore, NuWii provides an expandable base that could be used for gesture control in many other applications.
Download

Paper Nr: 16
Title:

KNOVA: INTRODUCING A REFERENCE MODEL FOR KNOWLEDGE-BASED VISUAL ANALYTICS

Authors:

Stefan Flöring and H.-Jürgen Appelrath

Abstract: When creating interactive applications for data exploration three major challenges can be identified: The integration of heterogeneous data sources at runtime, the integration of suitable visualization methods and the availability of interaction methods which enable domain experts to (implicitly) apply their expert knowledge in the knowledge driven exploration process. To address these challenges we introduce the KnoVA (Knowledge-Based Visual Analytics) reference model, which allows for generating a description of visualization methods, interaction methods and data sources. We then outline how this model can be useful to create knowledge based visual analytics systems in a model driven software development process.
Download

Paper Nr: 25
Title:

TOP-DOWN DATA ANALYSIS WITH TREEMAPS

Authors:

Martijn Tennekes and Edwin de Jonge

Abstract: Statistics Netherlands produces statistics about the economic activity in The Netherlands. These statistics are based on survey data and administrative sources, such as value added tax (VAT). A recent trend in the production of official statistics is to use a top-down analysis, which means that analysts first analyze high level aggregated data, and then zoom in on a more detailed level when necessary. In this paper, we discuss how treemap visualizations can be used for this top-down approach. We use comparison treemaps and density treemaps. Finally, we introduce a method to visualize confidence intervals in treemaps.
Download

Paper Nr: 31
Title:

REQUIREMENTS FOR INTERACTIVE ONTOLOGY VISUALIZATION - Using Hypertree +2.5D Visualization for Exploring Relationships between Concepts

Authors:

Isabel Cristina Siqueira da Silva and Carla Maria Dal Sasso Freitas

Abstract: Ontologies are used for sharing among people or software agents the common understanding of the information structure in a certain domain. Usually, ontologies are represented as static 2D graphs where the relationships are displayed as edges, which often overlap and cause cognitive overload. Three-dimensional representations can also lead to confusion due to occlusion. Moreover, as the ontology grows, incorporating new concepts (and their relationships) increases the visualization complexity either in 2D or in 3D. In this paper, we present a study about the requirements of visualization and interaction with ontologies. In order to do that, we interviewed with four experts on ontology creation and use. From the results, we propose the design of a 2.5D visualization tool for exploring relationships between ontology concepts.
Download

Paper Nr: 1
Title:

A VISUALIZATION BASED SIMULATOR FOR SYN FLOOD ATTACKS

Authors:

Tom Terry Jr., Huiming Yu, Ken Williams, Xiaohong Yuan and Bill Chu

Abstract: In this paper we present a visualization based simulator to help students gain knowledge of SYN flood attacks. The design principles of the simulator are user friendly, visualization based, and easy to install and use. The animation simulator demonstrates normal network traffic, exhibits how a SYN flood attack occurs, and shows how using a firewall might prevent SYN flood attacks. A set of short quizzes has been designed to help students to understand the concepts. The visualization based simulator for SYN flood attacks has been designed and implemented in the Department of Computer Science at North Carolina A&T State University to enhance information assurance education. This simulator was used in COMP 620 Information, Privacy and Security in the fall 2010 and received excellent results.
Download

Paper Nr: 22
Title:

VISUAL ANALYTICS OF MULTIMODAL BIOLOGICAL DATA

Authors:

Hendrik Rohn, Christian Klukas and Falk Schreiber

Abstract: Biological data is measured in increasing quantity and quality, resulting in data describing biological systems from different perspectives. Based on data integration methods, visual data mining and visual analytics can be used to promote the understanding of combined biological data and facilitate the exploration process. In this paper a number of view types are presented and integrated into a comprehensive software tool, in order to support researchers in visualizing flexible combinations of multimodal biological data and to create integrated views on comprehensive datasets spanning multiple “omics” areas. A number of interaction techniques accompany these views, enabling the efficient exploration of the data.
Download

Paper Nr: 23
Title:

TOWARD A POST-SERIALITY MAP OF TV SERIES - Visualizing the New TV Seriality System with Protovis

Authors:

Pier Giuseppe Mariconda and Stefano Perna

Abstract: The aim of this work is to visualize some information resulted from an analysis on 60 series from the top rated ones in the 2001-2010 decade. Arc-Diagram is presented as an overview of the TV series co-actor network. Airing networks and ratings are highlighted in the circular graph. Major insights are discussed.
Download

Paper Nr: 24
Title:

INTERACTIVE SEARCH AND RESULT VISUALIZATION FOR CONTENT BASED RETRIEVAL

Authors:

Levente Kovács

Abstract: This paper presents visual query, search and result visualization application which is interactive, robust, and flexible to be usable for different image and video retrieval applications. The main novelty of our approach is that, at the same time, it provides a text and model based search interface, a visual browsing interface, a distribution visualization interface based on a number of content based features, an annotation editing interface and a content classification interface, all combined together in an easy to use prototype.
Download

Paper Nr: 32
Title:

USING BLENDER TO DESIGN AND IMPLEMENT DATA VISUALIZATION COMPONENTS FOR X3D AND X3DOM

Authors:

Mario A. Bochicchio, Antonella Longo and Lucia Vaira

Abstract: We discuss how to use Blender to design and implement data-bound components for Web-based data visualization. The goal is twofold: to use Blender, with its declarative approach based on the logic brick programming model, for prototyping data visualization components of Web3D applications, and to test the advanced visualization capabilities of the new X3DOM framework based on HTML5 and WebGL. In order to validate the approach we have extended Blender with a software module able to map both Blender geometries and behaviours (logic bricks) code into X3D and X3DOM code.
Download