The proposed project will develop a new interdisciplinary methodology for both the analysis of user experience in virtual worlds and the design of such worlds. The methodology combines ideas from game design, computer science, information visualization, New Media Art and media theory. The project will combine two research efforts: The Scalable City developed in the Experimental Game Lab directed by Professor Sheldon Brown, and Cultural Analytics developed by the Software Studies Initiative directed by Dr. Lev Manovich. The Scalable City is an interactive artwork that uses custom computer graphics software to create a dynamically changing 3D urban environment. First, satellite imagery of real cities is manipulated through image processing routines to generate new, highly patterned terrains. After spatial analysis, areas are "zoned" for development using a variety of eccentric "real estate development" schemes - space filling curves, dense grids, mazes, and combinations of these forms. As a critical artwork, the Scalable City has been designed to challenge the existing conventions of 3D narrative environments. It engages the users in a narrative of discovery, not that of a story line, but of interacting with the world in self-reflective ways. Cultural Analytics is a computational method to quantitatively analyze patterns and rhythms in visual media such as photography, design, film, animation, choreography, and music composition. It has been developed specifically for humanities scholars of visual culture and media studies to harness the powers of computational methods to identify patterns of creativity.
Using the Cultural Analytics framework, Brown and Manovich propose to develop a new methodology for the analysis of users? activities and visual experiences in a multi-user version of the Scalable City. The Cultural Analytics Framework will be embedded into the Scalable City generative software. This new analytical software layer will enable the transformation of the virtual world in real-time in response to patterns in users? behavior and visual experience. At the same time, by incorporating the new analytics techniques in virtual world generation software, the project aims to advance the current research on how to create interfaces and simulations which analyze user performance and dynamically adapt based on the results of the analysis If successful, this research is likely to have transformative effects in a number of fields. Game designers, HCI researchers, and games and media scholars will be able to analyze, visualize and interpret the dimensions of user experiences with interactive time-based cultural artifacts such as video games, animated interfaces, and interactive artworks which are not captured with current analytics techniques such as network analysis (connectivity, load, and latency), econometrics on virtual economies, and profiling player game play.
The project developed a novel methodology for analyzing the experiences of participants in multi-user virtual worlds. The methodology uses a number of new analytics techniques developed by PIs since 2008, which have already been successfully employed to study non-interactive media such as film and animation. In the new project, these technique sare used for the first time to study important dimensions of user experiences in a virtual world that are not captured by analytics methods currently used in game industry. The new analytics technology is also integrated with world generation software. As a result, the virtual world can respond to patterns in users’ actions and visual experiences in real time by modifying its architecture and behavior. The project was carried out by creative practitioners who blend artistic work, computer science research andcultural theoretical analysis. It combines the work of two labs at the Center for Research in Computing and the Arts (CRCA) within the California Institute of Information Technologies and Telecommunications (Calit2) at UCSD: 1) the Cultural Analytics framework, developed by the Software Studies Initiative directed by Lev Manovich; 2) the Scalable City virtual world, developed by the Experimental Game Lab directed by Sheldon Brown. This analytics methodology can have a transformative effect on both user testing and design of 3D interactivevirtual worlds such as simulations, computer games and virtual reality communities. By making possible a quantitative analysis of users’ visual experience in the virtual world and integration of real-time analytics into world generation, our approach enhances the aesthetic richness and complexity of the virtual world, while providing new opportunities for users to make unexpected discoveries. The integration of the new analytics technologies with world generation software goes beyond the use of existing analytics methods in the industry – to make user interaction more efficient (as in a typical HCI application), or to increase the "playability" of the world (as in a typical commercial computer game). The project also contributes new methods for the design of procedural virtual worlds. The Scalable City creates a simulated urban environment using a custom 3D world generation engine. In contrast to existing virtual worlds, this environment already responds to patterns in user navigation by growing and transforming in real time. As a result of this grant, the new versions of the Scalable City run with a built-in analytics engine, and are able to guide users to discover new patterns of interaction and offer more rich and varied visual experience. Scalable City engages a broad audience, including students of all levels, through its exhibitions at science and art museums around the world. Supplementing this are ongoing lectures and presentation of the work at conferences and festivals in media and digital art in addition to technical conferences in game and virtual world design. Cultural Analytics techniques from this project have also created visualizations of user activity and visual experiences that are be shown in gallery and museums in the U.S. and in other countries. Project PI’s have also incorporated presentation of novel techniques for virtual world generation and the analysis and visualization of user experiences developed in the proposed project in their regular classes which reach several hundred undergraduate students at UCSD who major in computer science, digital art and media and PhD students in computer science at CUNY. The project has contributed new knowledge to a number of areas of computer science, engineering, and design: multicore distributed computing, HCI, interface design, and the design of virtual worlds.
