Broadly, sensemaking is the process of extracting and synthesizing information from data. This task spans many domains, including finance, cyber or intelligence analysis, and emergency management. The data itself is not limited to text, but can also be numerical data, multimedia, or a combination. Many analytical strategies exist to do sensemaking tasks, but we will not go into that much depth here.
Several sensemaking models exist that are based on actions analysts perform or the cognitive processes. The Sensemaking Loop model is composed of a series of iteratively accessible steps that are organized into two sub-loops: a foraging loop where the analyst gathers evidence, and a sensemaking loop where the analyst synthesizes information using schemas and hypotheses, resulting in the presentation of results [Pirolli and Card]. In emergency management, this presentation phase is replaced by relaying instructions and next steps to the staff and first-responders.
The recognition primed model also applies here, with it's different aspects of sensemaking [Klein et al]:
- The initial account people generate to explain events
- The elaboration of that account
- The questioning of that account in response to inconsistent data
- Fixation on the initial account
- Discovering inadequacies in the initial account
- Comparison of alternative accounts
- Reframing the initial account and replacing it with another
- The deliberate construction of an account when none is automatically recognized
- Pirolli and Card designed the Sensemaking Loop through cognitive task analysis of intelligence analysts.
- Klein et al.’s recognition-primed model from "A Data-Frame Theory of Sensemaking" describes the different aspects of sensemaking.