Quality, or QUal, is a domain ontology module of the Common Semantic Data Model (CSDM). It is a common ontology that enables to describe quality dimensions associated with data. It can be :
| ID | Question in natural language | Example |
|---|---|---|
| cq-1 | How elements of time series are grouped temporally ? | Each element of a time series corresponds to either a value observed at the timestamp |
| cq-2 | Is the value set complete regarding the temporal coverage period ? | The completeness of a value set can be evaluated using the percentage of missing values. |
| cq-3 | How was data produced ? | Data production process can be indicated to distinguish service-computed datasets from metered data for example. |
| cq-3 | WHat is the list of device properties described with the dataset ? | Data can represent grouped electrical consumption of a set of home appliances for example. The aggregation context should list these elements. |
qual:QualityMetric: The class of quantifiable measures used to assess a quality attribute.
qual:QualityAttribute: The class of attributes representing the degree to which data satisfy the requirements.
qual:Aggregation: The class of functions applied to aggregate data such as sum of values or average
qual:MeasuringProcedure: The class of procedures used to provide data
A qual:Aggregation enables the definition of what aggregation level to be used while interpreting values in a value set. The aggreation can occur on a TemporalContext indicating a time interval on which the aggregaton has been applied, a SpatialAggregationContext indicating a spatial location on which the aggreation has been applied or a LogicalAggregationContext to cover other grouping types.
If no aggregation indicated, the default interpretation can be to associate the exact value to the corresponding timestamp.
According to usage, individuals are created for each (function, Aggregation contexte). For example, qual:HourlySum indicates an Aggregation of type qual:Sum with a qual:TemporalAggregationContext of 1 hour.
The qual:MeasuringProcedure enables to indicate the procedure used to produce data. The use of hardware devices such as smart meters indicates a qual:Metering procedure, qual:Observation for sensors, while software simulators give qual:Simulation, scheduling services qual:Scheduling and qual:Prediction for other software predicting data.
A set of qual:QualityMetric can be associated with a ets:ValueSet to describe its quality. For each relevant quality metric, one individual should be created to indicate the qual:value. For example, to descrive the qual:Completeness of the electrical consumption time series ElectricityTimeSeries, a ETSPercentOfMissingValues individual of type qual:PercentOfMissingValues enables to provide a value of completeness.