Qualitative Analysis Toolkit
The Causemos HMI supports a qualitative Causal Analysis Graph (CAG) assembly workflow that allows analysts to leverage automatically processed literature to identify the main drivers and relationships for target concepts.
CAGs can be turned into semi-quantitative computational models used to analyze different scenarios and interventions. In this scenario analysis workflow, available data (see Quantitative Analysis for more information) is automatically mapped to concepts. While quantification is not required for scenario analysis, when included it allows World Modelers engines to infer the nature and strength of relationships in the CAG. To validate the mapped defaults and resolve any data quality issues, the system solicits analysts to provide their own human judgment.
A parameterization of the semi-quantification computational models is sent to the three inference engines (DySE, Delphi, Sensei) to create models and project the baseline for the selected timescale (months or years). Projections visualized in Causemos are meant to provide a sense of directionality of trends and whether changes are small or large (especially compared to historical dynamics), and account for all the influences in the complex system. Analysts can apply “clamps” to future values for one or more drivers, to simulate interventions or specific worst case/best case scenarios, see cascading impacts across the system, and compare to baseline trends. An explicit representation of the system makes levers available for interventions more explicit and supports creative thinking. This helps mitigate framework issues and encourages analysts to consider possible hindrances to interventions and their unanticipated effects.