Seminar: Wednesday, April 14, 3 pm, SDSC Auditorium

Title: Complex Adaptive System Technology in a Distance
Collaboration Environment: VirtualCal CAST

Speakers:Van R. Whiting, Jr., Director, VirtualCal Distance Collaboration System (UCSD/USMEX) and Stephan Hollmann, Guest Scholar, UCSD VirtualCal CAST Contact: vwhiting@ucsd.edu

SDSC Host is: Amitava Majumdar majumdar@sdsc.edu

Abstract:
VirtualCal CAST offers the prospect of intelligent knowledge resources accessible during real time distance collaboration. The development of this application will utilize the infrastructure of
Internet2 connectivity, resource storage, and grid computing to bring intelligent agents to complex environments.

The system begins with a web presence and real time collaboration environment for scientists, engineers, and industry researchers. A variety of tools for meetings are included in an automated on-demand collaboration system. Participants are then invited to populate their domain of knowledge using Q methodology, a quantitative method for factor analysis of "the human factor," that is, human subjectivity regarding the knowledge domain.

The human factor is correlated for matrix analysis overlay to the corresponding technical domain. By rotating these factors, a theoretical "node probability system" or NPS will be created. Analyzing patterns semantically, temporally, and spatially allows the identification of signal/noise ratios in the system, to generate respective domain landscapes. SHRUTI based neural networks operate on selected domain landscapes containing candidate "stories" or areas of interest. Checked on plausibility, conformity, completeness, the neural system creates independent agent programs to validate assumptions.

VirtualCal CAST is an application building on twenty-first century cyber infrastructure, bringing intelligence, complex adaptive system techniques, and collaborative decision tools to science, industry, and government.