Measuring Situational Awareness through Analysis of Communications: A Preliminary Exercise

Network centric warfare promises to increase information sharing and allow distribution of decision making. This will improve military effectiveness, but only if the situational awareness (SA) of warfighters is correctly aligned. Modern natural language processing techniques, such as Network Text Analysis (Carley, 1993), are designed to infer the cognitive states of individuals and groups engaged in cognitive collaboration and measure group SA by exploiting data on the information that team members access and generate. An integrated software application, IMAGES, utilizes AutoMap (Diesner & Carley, 2004) as the primary analysis engine to take advantage of the large amounts of communication and report text that naturally occur in collaborative environments. The text generated in the normal course of work is collected and changed into forms that can be compared and analyzed. A comparison of networks based on text from several individuals or groups yields information about the similarity of their respective mental models. Differences among maps may reflect misalignments of SA, which can be remedied by information sharing and targeted communication. An exercise was conducted to assess the potential of NTA as implemented in AutoMap and IMAGES. The results indicate that NTA will allow analysts to effectively assess SA through passive means.
The original document contains color images.