The SWRL Temporal Ontology is a core component of the SWRLAPI's temporal subsystem. It defines a valid-time temporal model that can be used to model complex interval-based temporal information in OWL ontologies. It also defines a library of SWRL built-ins to perform temporal operations on information described using this ontology. The use of this library with this temporal ontology is described here.

This ontology has the default prefix temporal. A copy of it can be found here.

An introduction to this ontology and an explanation of how can be used to perform temporal reasoning is covered in the following paper:

• "A Method for Representing and Querying Temporal Information in OWL" M. J. O'Connor and A. K. Das. Biomedical Engineering Systems and Technologies (Selected Papers), Springer Verlag, CCIS 127, 97-110, 2011.

Table of Contents SWRL Temporal Ontology Representing Granularity Representing Durations Representing Valid Times Representing Facts or Propositions Temporal Built-Ins

The SWRL Temporal Ontology defines OWL entities that can be used to represent an interval-based temporal information in OWL. This ontology defines a temporal model that provides the formal underpinning for this representation. This model is based on the valid-time temporal model, a commonly-used model to represent temporal information in many systems.

The valid-time temporal model provides an approach for consistently representing temporal information. In this model, a piece of information - which is often referred to as a fact or a proposition - can be associated with instants or intervals denoting the time or times that the fact is held to be true. Such facts have a value and one or more valid times. In other words, every temporal fact holds information denoting the facts's valid-time. Conceptually, this representation means that every temporal fact is held to be true or valid during the time or times associated with this fact. No conclusions can be made about the fact for time periods outside of its valid-time. More information about the valid-time temporal model can be found here.

The SWRL Temporal Ontology provides OWL entities for representing all the entities defined by this model, including propositions, valid-times (both instants and intervals), granularity, and duration.

A granularity is specified using a class called temporal:Granularity, which has instances Years, Months, Days, Hours, Minutes, Seconds, and Milliseconds.

Durations are represented by the class temporal:Duration, which has a functional integer property hasCount and a functional granularity property with a range of temporal:Granulairty.

Valid-times are represented using the class ValidTime. Each valid-time class has a temporal:hasGranularity property that specifies the accuracy of the temporal information associated with it. The range of this property is temporal:Granularity.

Two subclasses of this class are defined in the temporal ontology to represent instants and intervals. These classes are called temporal:ValidInstant and temporal:ValidPeriod, respectively.

The temporal:ValidInstant class is used to represent instants. It has a temporal:hasTime data valued property containing an XML Schema dateTime.

The temoral:ValidPeriod class is used to model intervals. It has temporal:hasStart and temporal:hasFinish XML Schema dateTime properties.

One possible use of ValidInstant and ValidPeriod classes is to take an existing OWL class and adding a property with a range of one of these two classes. The choice of class depends on whether one wishes to model an activity that occurs at a single instant in time or one that takes place over an interval of time. Also, if the activity occurs only once the property will be functional, whereas an activity that may repeat will use a non-functional property.

Say, for example, we would like to work with blood pressure measurements and have an existing class called BloodPressureMeasurement that contains properties for both the systolic and diastolic components of a measurement. Blood pressures are typically recorded as instantaneous measurements so the ValidInstant class would be the appropriate property range choice here. A useful name for this property could be hasValidTimes. By using this class all instances of the BloodPressureMeasurement class can now use the hasTime and hasGranularity properties associated with the valid time property, which will allow them to consistently record the temporal information associated with the measurement. Similarly, if we would like to work with prescriptions using an existing class called Prescriptions, we might chose to use the ValidPeriod class as the range of a property called hasValidPeriod.

A more elegant us of valid-times is to associate them with a class representing a fact or proposition and to use that class directly to model entities with temporal extent. In the SWRL Temporal Ontology this class is called temporal:ExtendedProposition. It can be used to represent entities that can extend over time.

This class has a property called temporal:hasValidTimes with a range of temporal:ValidTime that holds the propositions valid-times. Instances of the extended proposition class will have either instants or intervals as values of this property.

Say, for example, we would like to take our earlier blood pressure measurements class and model it as an extended proposition. We can take the BloodPressureMeasurement class and make it a subclass of the temporal:ExtendedProposition} class. Instances of this class will now be able to use the temporal:hasValidTimes property to store their valid instants. Similarly, if we would like to work with these prescriptions class, we can again subclass the temporal:ExtendedProposition class and store

One of the advantages of subclassing the extended proposition class is that the existing ontology does not have to be modified and can be used as is.

Representing temporal entities using the extended proposition class also clarifies the distinction between temporal and non temporal entities in an ontology. It also provides a consistent mechanism for representing all temporal information.

A significant additional advantage is that one all temporal information is represented consistently, it can then be manipulated consistently. To this end we have provided a SWRL built-in library called the [XXX?SWRLTemporalBuiltInsAdvanced] that supports an array of temporal operations on entities defined using the classes in the SWRL Temporal Ontology.

The ontology also contains a set of SWRL built-ins than can be used to reason with temporal information defined using its temporal model. These built-ins are described here.