David Kasik

Data science addresses the need to extract knowledge and information from data volumes, often from real-time sources in a wide variety of disciplines such as astronomy, bioinformatics, engineering, science, medicine, social science, business, and the humanities. The range and volume of data sources has increased enormously over time, particularly those generating real-time data. This has posed additional challenges for data management and data analysis of the data and effective representation and display. A wide range of application areas are able to benefit from the latest visual tools and facilities. Rapid analysis is needed in areas where immediate decisions need to be made. Such areas include weather forecasting, the stock exchange, and security threats. In areas where the volume of data being produced far exceeds the current capacity to analyze all of it, attention is being focussed how best to address these challenges. Optimum ways ofaddressing large data sets across a variety of disciplines have led to the formation of national and institutional Data Science Institutes and Centers. Being driven by national priority, they are able to attract support for research and development within their organizations and institutions to bring together interdisciplinary expertise to address a wide variety of problems. Visual computing is a set of tools and methodologies that utilize 2D and 3D images to extract information from data. Such methods include data analysis, simulation, and interactive exploration. These are analyzed and discussed.

Interactive display and visualization of large geometric and textured models is becoming a fundamental capability. There are numerous application areas, including games, movies, CAD, virtual prototyping, and scientific visualization. One of observations about geometric models used in interactive applications is that their model complexity continues to increase because of fundamental advances in 3D modeling, simulation, and data capture technologies. As computing power increases, users take advantage of the algorithmic advances and generate even more complex models and data sets. Therefore, there are many cases where we are required to visualize massive models that consist of hundreds of millions of triangles and, even, billions of triangles. However, interactive visualization and handling of such massive models still remains a challenge in computer graphics and visualization. In this monograph we discuss various techniques that enable interactive visualization of massive models. These techniques include visibility computation, simplification, levels-of-detail, and cache-coherent data management.We believe that the combinations of these techniques can make it possible to interactively visualize massive models in commodity hardware. Table of Contents: Introduction / Visibility / Simplification and Levels of Detail / Alternative Representations / Cache-Coherent Data Management / Conclusions / Bibliography