This research creates an innovative method for visualizing and analyzing space-time dimensions of the spread of ideas through semantic webs represented in texts posted on the Internet and the Web pages. The project integrates geographic information systems (GIS), geographic information science (GIScience), computational linguistics (CL), and semantic web (computer-based ontology) technologies to track and analyze public-accessible websites for progressively more defined clusters of words and phrases that characterize actual and potentially developing networks of social processes. Human and natural crises (e.g., epidemics), and hostile social movements (e.g., militia and hate groups) are examined as illustrative exemplars of the utility of such an integrated system. Website pages and web contents with identified clusters of words or phrases will be mapped (by geo-referencing their web addresses, URL, place names, gazetteers, blogs, etc.) over a world map (using GIS tools) with time stamps. The resulting map provides a visual ?information landscape? consisting of hundreds of website locations (using real world coordinate systems) containing related keywords or similar ideas. When integrated with time-series analyses, this map allows the examination of the paths and speed of information dissemination, as well as the evolving varieties of various ideas and their relationships. By creating a Semantic Web Automatic Reasoning and Mapping System (SWARMS) prototype, researchers can visualize the spread of concepts, ideas and news over time and space. Clusters of keywords and phrases are identified and classified, and references to key ideas, provocative events, and important text sources are collected. This innovative methodology can be applied in multiple languages and other applications, such as cellular phone text messages and social network messages (such as twitter messages).
Intellectual Merit: This project forms a new multidisciplinary research framework in connecting social science and computer science for analyzing the dynamic information landscape on the Internet in which ideas spread quickly. By combining GIScience and geo-locating skills in a Web search mechanism, this project illustrates a new approach to information query, retrieval, and analysis methods. The creation of semantic knowledge bases upgrades traditional ?data mining? methods to advanced ?information mining? approaches. This research facilitates the development of space-time analysis methodology, a rapidly growing field in the context of the traditional separation between time series analysis and spatial analysis. The innovative semantic and spatiotemporal analysis framework provides a new direction for social science research to track the spread of ideas, to analyze where they go and how fast, and to analyze the ideological, social, and religious conditions that promote that spread. There are numerous current for- and non-profit organizations as well as government organizations that routinely monitor the web for various purposes. The unique aspect of this project, however, is that it brings together disparate academic disciplines not only to monitor the web, but also introduces a coherent methodological framework to analyze the spread of an idea and its impact through cyberspace.
Broader Impacts: This project helps to foster the integration of research and education in multiple disciplines, including geography, linguistics, political science, and communication. Four graduate students at SDSU are directly funded and trained through this project. Students across a variety of disciplines and curricula also directly benefit from this research project. This research provides an excellent opportunity to involve students from diverse backgrounds. San Diego State University is ranked number 8 in the nation for bachelor?s degrees awarded to Hispanics and number 11 for bachelor?s degrees awarded to all ethnic minorities, and was ranked number 22 in the nation for ethnic diversity by U. S. News and World Report in 2006. Given the makeup of the student body, this project features underrepresented groups in both graduate and undergraduate programs. The project website publicizes research findings to the general public and creates a discussion forum to involve multidiscipline researchers. Four Cyber-Discovery and Information Landscape workshops are planned to facilitate future multidisciplinary collaborations. This research improves understanding of the technical, psychological, and political mechanisms and conditions that facilitate the spread of certain ideas and provide a new direction for social science research.
