sions related to medical protection and treatment as well as those that inform intelligence, operations planning, medical intelligence, and attribution. Just as S&T contributes considerably to medical decisions, it also contributes to military and policy decisions related to possibly nefarious activities involving biological and chemical agents. The committee has identified four S&T capabilities that are required to provide technologies needed for chemical, biological, radiological, and nuclear (CBRN) intelligence, surveillance, and reconnaissance (CBRN ISR): information acquisition and analysis, health monitoring, environmental monitoring, and unknown agent identification and characterization.
Information Acquisition and Analysis
The core concept for this capability is that multiple dense sources of information pertinent to CBRN ISR exist and need to be properly mined, managed, integrated, distilled, and efficiently queried. A few examples of information sources in the biological domain include ProMed Mail (outbreak reports), ARGUS (a system that searches World Wide Web news media for signs of social unrest that could be due to pathogen outbreaks), Global Public Health Information Network, social media such as Facebook or Twitter,5 and global weather forecasts and reports. To be fully effective, an information acquisition and analysis system needs to be capable of extracting pertinent facts from a large number of languages and information source types. This involves numerous difficult and largely unsolved problems in natural language processing and information fusion.
DoD needs to be well informed of potential and actual biological and chemical outbreaks and incidents as they relate to both existing and potentially needed detection capabilities. There are multiple government and non-government sources of relevant information for enabling CBRN ISR. Additionally, there are multiple government-funded groups that already attempt various levels of information integration in the chemical and biological domain. It is likely that no existing systems individually meet all current DoD needs. However, DoD should strive to leverage as much as possible from collaborations with existing systems, rather than launching any de novo efforts at developing such capabilities. The committee noted that the current program plans to search social media for outbreak-relevant information did not appear to have a plan for integrating with other available information sources. CBDP should be better aligned with other relevant government information acquisition and analysis efforts in the biological and chemical domains.
5 We note that legal issues involving the Privacy Act are faced by any information systems that mine information about US citizens, even if such information is in the public domain.