Recently, modular ontology modeling has become a more popular ontology engineering paradigm. With it, the need for additional metadata associated with ontology design patterns has grown. The Ontology Pattern Language (OPLa) was developed to facilitate annotating ontologies with useful metadata, as well as supporting tooling infrastructure. In this paper, we detail three extensions to OPLa into a reorganized namespace: OPLa-core, containing the original annotations; OPLa-SD, for use in detailing schema diagrams; and OPLa-CP, an adaptation of the content-pattern annotation schema.
As the pattern-based, modular ontology engineering methodology has matured, it has become apparent that additional metadata describing that patterns and modules that compose the modular ontology is necessary. Indeed, we have seen that the documentation of an ODP has three facets which are espeically useful during development: structural and provenance, content and usage, and visualization.
Structural and provenance documentation refers to those annotations that
describe how patterns and modules may interact with each other, as well as
related patterns, how they connect, or their most general form. Recently, [
Content and usage has been historically covered by the cp-annotation-schema; these metadata annotations are critical for the discovery, reuse, and interoperability with existing patterns. When a pattern is submitted to the ontology design patterns web portal3, the form requires that the pattern be annotated with the cp-annotation-schema.4 This schema includes a number of annotation properties that allow a developer to express certain usage characteristics of their content ontology design pattern. However, it is not always clear how to use these annotations and may pose as an obstacle for advanced querying. ? Copyright c 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). 3 http://ontologydesignpatterns.org 4 url2cp-annotation.owl
Pre x Namespace opla-core: http://ontologydesignpatterns.org/opla-core# opla-cp: http://ontologydesignpatterns.org/opla-cp# opla-sd: http://ontologydesignpatterns.org/opla-sd#
Furthermore, we have seen that schema diagrams are an important tool for
understanding the purpose, content, and structure of an ontology design
pattern or module [
To support these needs, this paper describes the extension and slight reorganization of OPLa into three distinct sub-namespaces. Table 1 shows these new namespaces. Thus, we enable a simpli ed documentation process, as well as provide multiple dimensions for pattern discovery. The contributions are as follows. 1. OPLa-Core: the new namespace of the original OPLa ontology and the annotations therein remain unchanged. 2. OPLa-CP: contains the annotation properties describing a number of nontechnical facets on the usage of a particular ontology design pattern or module. These annotation properties are adapted from the cp-annotation-schema. 3. OPLa-SD: the foundational positional axioms used for expressing visual location of an entity on a canvas. 2
As mentioned in the previous section, there are three distinct representations of a pattern that we wish OPLa to cover. Furthermore, we wish to have each of the annotations to support discoverability, as such, some disambiguation of the uses of the annotations was necessary. In the next sections we brie y discuss the usage characteristics of the new OPLa-cp annotations and two annotations from OPLa-core. The updated OWL le for these OPLa extensions (as well as the original speci cation) is available online.5 2.1
In this section, we merely disambiguate some common questions pertaining to two original OPLa annotations and a new core annotation, as adapted from cpannotation-schema. While some of these annotations were in the original speci cations for OPLa, their use-cases were not explicit. opla-core:isNativeTo should be used to express the provenance of some ontological entity. This annotation is not functional, e.g. some entity may be native to more than one pattern, e.g. the Agent may belong to both AgentRole and Provenance patterns. opla-core:ofExternalType should be used to indicate that another pattern may hook into this entity. For example, a hook might indicate that the pattern developer acknowledges a certain concept is out of scope of the particular pattern. opla-core:extractedFrom is adapted from cp-annotation-schema This annotation should be used to indicate that a pattern has been created from where one originally did not exist. Otherwise, opla-core:derivedFrom should be used. 2.2
The adaptation of cp-annotation-schema is the primary contribution in this paper. The purpose of the cp-annotation-schema is to describe many aspects pertaining to the usage of an ontology design pattern, thus we have adapted many of the annotations for use in OPLa. However, not all of the annotations in cpannotation-schema were necessary to be included, as their purposes were covered by existing OPLa annotations. Additionally, few of the original annotations implied that their payload was plural, e.g. hasConsequences. However, in order to aid in discovery and minimze complicated text-processing, we felt that it was more natural to have singular natural language payloads, i.e. hasConsequence. opla-cp:hasConsequence Describes a potential gain and drawback when using the annotated module or pattern. For example, it may be used to express the impact of an ontological commitment. opla-cp:coversRequirement Should point to a blank node, wherein a competency question and SPARQL query should be paired. This pair should pretain to a question answerable by the original pattern itself without being instantiated for a speci c usecase. opla-cp:hasCompetencyQuestion Points to an example competency question that can be evaluated against the annotated ontology, as expressed in natural language. 5 https://github.com/cogan-shimizu-wsu/Extended-OPLa/ :Agent rdf:type owl:Class ; opla-sd:entityPosition [ opla-sd:entityPositionX "19.608958656638492"^^xsd:double ; opla-sd:entityPositionY "89.84401282241834"^^xsd:double ; rdfs:comment "This is an entity positioning annotation generated by CoModIDE (https://comodide.com/). Removing this annotation will break rendering the CoModIDE schema diagram view."@en ] . opla-cp:hasUnitTest Points to an example SPARQL query that can be evaluated against the content of the annotated ontology. It should be paired with a natural language description, e.g. Competency Question. opla-cp:addressesScenario Describes a potential or existing usecase or instantiation of the ontology design pattern. It should describe the commitments necesarry to make the instantiation possible. 2.3
OPLa-SD includes a set of annotations speci cally for tooling software to query
and utilize. Currently, there are only three annotations belong to OPLa-SD,
as detailed below. Figure 1 shows an excerpted opla-sd:EntityPosition from the
annotated version of the AgentRole pattern in [
This paper describes the reorganization of the OPLa namespace into three new sub-namespaces, OPLa-core, OPLa-CP, and OPLa-SD. This reorganization, and the population of the new sub-namespaces, supports the documentation of ODPs, in three critical facets, as well providing descriptions to annotations previously without documentation. Each namespace has a set of annotations for specialized roles and functionality.
These new OPLa annotations can be seen in two ongoing projects: MODL: a
modular ontology design library [
Acknowledgement. Quinn Hirt acknowledges funding from the Air Force O ce of Scienti c Research under award number FA9550-18-1-0386. Cogan Shimizu acknowledges support by the Dayton Area Graduate Studies Institute (DAGSI).