User Requirements
The users for this product are referred to as analysts. They have technical training and experience in social sciences. Their main goals are to review information for evidence to use in reporting. They don't mind reading a lot of information.
- Review large amounts of data.
- Ability to attach evidence to a model quickly.
- Review results of evidence against the model.
- Adjust evidence evaluations.
- Assign evidence into a custom, user generated taxonomy.
Technical Requirements
- Collect user inputs for evidence relevancy model.
- Evidence must be evaluated before it is assigned to the intent model.
- Evaluation includes assigning the evidence to a model node, selecting an indicator, and assigning a numeric weight to the indicator.
This is the base workflow for the software. Supporting workflows include, training an evidence recommendation algorithm, editing evidence, editing evidence evaluations, adding text as evidence that is not found the documents, creating a user driven taxonomy structure. These workflows have been developed at the request of analysts and through user studies results.
Evidence Curation Workflow
Process sketches that explore layout and interactions
Early iterations of the software. These were evaluated by users both qualitative and quantitatively.
Results of User Studies
- Manage more evidence, how to manage evidence against more complex models. Explore different document list visualizations (i.e. tree map, graph, list hierarchy)
- Work with multiple projects within the same data set. Easily move evidence between projects.
- Start page to visualization available projects or start a new analysis.
