Model Sensemaking Strategies
This page describes what are sensemaking strategies, and how are they classified.
A model sensemaking strategy is a purposeful model visualization based on a visual metaphor. They are defined as reusable patterns that yield tailored visualizations when applied to meta-models. They are aimed at accomplishing specific model understanding tasks, offering visualization metaphors for it – like charts, plots, maps, graphs, or matrices. SMSs are reusable since each strategy exposes a small meta-model pattern that when bound to a target meta-model results in a tailored visualization for the target meta-model instances. SMSs are also flexible as they can be used to understand (meta-)models at any level of abstraction, as well as entire modelling ecosystems.
SMS are instantiated with a binding connecting the context meta-model of the SMS (i.e., its pattern) to a target meta-model.
Structure of an SMS
SMSs are described according to the following structure:
Intent.
It describes the goal of the strategy, i.e., the sensemaking task(s) it aims to answer.
Presentation metaphor.
Similar SMSs are grouped by presentation metaphor, including: data, grouping, metric, time, and structural.
Context metamodel.
It defines the application pattern of an SMS. It must be bound to a target meta-model so that the SMS can be used to visualise meta-model instances. Each element of the context meta-model (i.e., classes, attributes, and references) can be deemed as conceptual “holes” to be populated by elements of the target meta-model. Context meta-models usually contain few elements, enabling their application to a wide range of target meta-models.
Visualisation variants.
The different ways in which an SMS can be visualised. They can be switched at run-time.
Properties.
Properties allow customisating the resulting visualisations. Each SMS defines its own set of properties, which can be mandatory or optional, and are typed. Properties are independent of the target meta-model and are populated by value.
Catalogue of supported SMSs
Dandelion offers a catalogue of SMSs, summarised in the following table. Each queryable SMS is detailed in its own page.
| PM | SMS | Visualisation variants | Description |
|---|---|---|---|
| D | Numerical | line graph, area chart, scatter plot | For (x, y) coordinates. |
| D | Numerical + Frequency | bubble chart | For (x, y) coordinates with an associated frequency. |
| G | Categorical | vertical/horizontal bar chart, (semi-circle) donut chart, pie chart, proportional area chart | To partition data into categories. |
| M | Metric Distribution | boxplot | To depict the most important percentiles of a metric. |
| M | Free Metric | highlighted number, icon and number | To visualise an unconstrained value. |
| M | Bounded Metric | angular gauge, gauge chart | To visualise a value bounded in a min..max range. |
| M | Literal Metric | word cloud | To represent frequency in a textual field. |
| T | Time-Based | Gantt chart | For timed tasks with start and finish dates. |
| S | Connectivity | adjacency matrix/heat map, chord diagram | For cross-referenced objects. |
| S | Weighted Hierarchy | treemap | For nested objects. |
Where PM = presentation metaphor: D = data, G = grouping, M = metric, T = time, S = structural.