III: EAGER: Mapping Three-Dimensional Virtual Worlds

University Of Arkansas, Fayetteville AR

Investigators

Abstract

Navigating in online virtual worlds is similar to searching the Web prior to the development of the first search engine. Users can find regions of interest to them through serendipity, by browsing or word of mouth, but they can only search the specific world they are on. This exploratory project investigates an enabling mobile crawler technology that can collect information across multiple worlds, and allows virtual worlds to grow in a distributed, multi-server, multi-platform manner. Of particular interest is content in 3D environments that is interconnected with the traditional flat Web. In this research explores strategies for gathering content from 3D environments and, using that content, gain a better understanding of how users are utilizing 3D environments and how they might benefit from more accessible, cross-platform search services. Crawling in 3D environments has many similarities with flat Web crawling but has additional constraints that may constrain how much content can reasonably be collected. The gathered data will further be explored, e.g., how frequently content changes in 3D environments and how users utilize linking in virtual worlds as compared to the linkage structures on the Web. In addition, this project investigates how traffic patterns in 3D environments are related to links, recommendations, and groups. The research challenges in this project are: (1) Gathering content from distributed 3D environments. This includes approaches for navigating around the 3D environment for information collection are explored, mapping the 3D environment space, analysis of the impact of additional information that can only be collected via interaction, and characterization of the additional information that is available through internal and external links. (2) Gaining a better understanding of user behavior. Research focuses on investigating ways in which user linkage patterns are similar to, and different from, user linkages on the flat Web and whether or not avatar traffic can be used to identify sites of broader user interest that can be exploited by the search process. (3) Supporting search across multiple platforms. The aim is to develop a 3D 3D object ontologies and a schema, similar to RSS, that will enable virtual world servers to more easily support sharing of their data with an external search service. As 3D environments become both more prevalent and more diverse, the need for a cross-platform search service will continue to grow. By increasing the visibility of content across platforms and servers, this project will allow content creators to focus on creating higher quality 3D environment by providing effective search mechanism for finding the desired content in a large, distributed, virtual world. The expected results are expected to provide foundations for exploring questions of how to discover and rank a wide range of content in virtual reality, mirror world, and augmented reality environments. By creating an open platform, this project will provide opportunities for more researchers to work in 3D environments. The database of regions, traffic, and linkages will be made available for downloading and analysis by other researchers. Research results, gathered data and the crawler will be made available as freeware via the project Web site (http://virtualworldsearch.uark.edu).