Second International Conference on Coordinated & Multiple Views in Exploratory Visualization (CMV2004)

Preface This is the second 'International Conference on Coordinated & Multiple Views in Exploratory Visualization', the first of which was held in 2003 in London. The aim of this conference is to bring together top researchers and practitioners who are interested in the area of coordinated and multiple views in exploratory visualization (CMV). Multiple views can be useful for several reasons. For example, they allow data to be presented in different forms that in turn allow the user to perceive the information through different perspectives. They can provide a history of experimentation, which can enable direct comparison of various parameterizations; as the views are often tightly coupled one view can be used to control other views, such as filtering or enhancing the information that is displayed in other related views (providing 'rapid coordinated investigation and exploration'). Moreover, there are many coordinated operations that could be applied such as selection (brushed highlighting), navigation (e.g. zooming, scrolling), filtering and coupled windowing operations (e.g. window placement, deletion). The conference provides an ideal venue to discuss state-of-the-art concepts and future opportunities in this area. Indeed, there are still many exciting research challenges and problems to solve. One of the challenges, especially in today's hands-on society, is the need to design useful coordinated exploration tools that are highly interactive and allow the user to satisfy their inquisitive nature. Users expect to manipulate their data, try-out scenarios, search for pertinent information and compare results. In particular, developers need to tread a fine line between functionality and usability: Functionally rich to satisfy the hungry demands and inquisitive nature of the user, yet being ?easy to use? and intuitive to learn. Thus, it is important to work out what is effective, useful and usable. During the 2003 conference there was a congenial discussion about the quantity of views, whether dual, three or n views are best. Various opinions were voiced: 'We have two eyes, so two views must be best'. 'Three views are often used in CAD presentations hence three views are best'. However, 'N views allow the user freedom''. The discussion remains open. Accordingly, it is imperative to generate empirical studies of the novel designs. Moreover, even in a climate of increasing complexity and ever growing dataset sizes, users expect to operate these exploratory systems and visualize their data at interactive speeds, to extend the tool (incorporating their own algorithms) or use the software for new datasets with different data-configurations. Thus, developers need to address software engineering and computer science challenges to develop efficient and effective algorithms, incorporated within an extensible and maintainable system that perhaps interoperates with other software. The papers in this volume present current research and novel ideas of coordinated and multiple view exploratory visualization. They demonstrate the breadth of this area, detail solutions towards some of the challenges, and allow the reader to further study this exciting research area. Jonathan C. Roberts CMV2004 Conference Chair