Knowledge based public health situation awareness

There have been numerous efforts to create comprehensive databases from multiple sources to monitor the dynamics of public health² and most specifically to detect the potential threats of bioterrorism before widespread dissemination. But there are not many evidences for the assertion that these systems are timely and dependable, or can reliably identify man made from natural incident^1,3. One must evaluate the value of so called 'syndromic surveillance systems' along with the costs involved in design, development, implementation and maintenance of such systems and the costs involved in investigation of the inevitable false alarms¹. In this article we will introduce a new perspective to the problem domain with a shift in paradigm from 'surveillance' toward 'awareness'. As we conceptualize a rather different approach^4,5 to tackle the problem, we will introduce a different methodology in application of information science, computer science, cognitive science and human-computer interaction concepts in design and development of so called 'public health situation awareness systems'. We will share some of our design and implementation concepts for the prototype system that is under development in the Center for Biosecurity and Public Health Informatics Research, in the University of Texas Health Science Center at Houston. The system is based on a knowledgebase containing ontologies with different layers of abstraction, from multiple domains, that provide the context for information integration, knowledge discovery, interactive data mining, information visualization, information sharing and communications. The modular design of the knowledgebase and its knowledge representation formalism enables incremental evolution of the system from a partial system to a comprehensive knowledgebase of 'public health situation awareness' as it acquires new knowledge through interactions with domain experts or automatic discovery of new knowledge.