UCO should clarify role of informal type string fields vs. implemented classes

[ajnelson-nist]

Background

UCO has several owl:DatatypePropertys that are used to house plain strings (sometimes that are part of semi-open vocabularies) which can be used to describe a type. For example, observable:accountType describes a type that is some specialization of observable:Account, and the vocabulary suggests some values (ldap, nis, openid, etc.).

These properties are provided for UCO users to strike a balance between specificity and agility. For instance, an account type might be sufficiently niche in an investigation that there would be little to no benefit to attempting to standardize it. But, less-niche options would suffer from not standardizing. For instance, the "Crossover WMD" CASE example uses the string "Phone" as a value of observable:AccountType - but, the class observable:PhoneAccount is in UCO.

This proposal introduces a property core:informalType to clarify the usage of these plain string fields that proxy formal types, by inlining documentation into one place within the ontology.

Requirements

Requirement 1

String fields that are used as a substitute of OWL-encoded or SHACL-reviewed classes must include a link to documentation on their role in data creation, management, and interchange.

Requirement 2

Any grouping of these informal-typing string fields must not impede usage of semi-open vocabularies.

Risk / Benefit analysis

Benefits

This proposal introduces structural-purpose documentation, and a structural link to that embedded documentation, to many properties throughout UCO.

Risks

Subproperties have the potential to introduce comprehension complexity from multiple rdfs:domains and rdfs:ranges going up the superproperty hierarchy. Requirement 2 is intended to avoid this situation with the new property, and UCO happens to not use rdfs:domain on any of the properties impacted by this proposal. So, no new risks are currently believed to be associated with this change.

Competencies demonstrated

Competency 1

Competency Question 1.1

What are the informal type properties currently in UCO?

Result 1.1

The solution is drawn from the implementing PR (link coming momentarily after posting), and this SPARQL query which assumes introduction of core:informalType:

PREFIX core: <https://ontology.unifiedcyberontology.org/uco/core/>
SELECT ?nProperty
WHERE {
  ?nProperty
    rdfs:subPropertyOf+ core:informalType ;
    .
}
ORDER BY ?nProperty

The result comes from running that query against test/uco_monolithic.ttl after make check was run.

?nProperty
configuration:dependencyType
configuration:itemType
core:eventType
location:addressType
marking:definitionType
observable:MSISDNType
observable:accountLogonType
observable:accountType
observable:actionType
observable:archiveType
observable:audioType
observable:blockType
observable:callType
observable:cellSiteType
observable:contentType
observable:dataType
observable:deviceType
observable:diskPartitionType
observable:diskType
observable:driveType
observable:eventType
observable:extFileType
observable:fileSystemType
observable:hiveType
observable:icmpType
observable:imageType
observable:libraryType
observable:messageType
observable:mimeType
observable:passwordType
observable:peType
observable:pictureType
observable:rangeOffsetType
observable:serviceType
observable:startType
observable:triggerSessionChangeType
observable:triggerType
observable:urlTransitionType
observable:whoisContactType
tool:toolType

Solution suggestion

core:informalType is introduced as a property. Its lengthy documentation string is given here with line-breaks:

Informal Type serves as a parent property for string-valued properties meant to describe a type without implementing a class design. This property hierarchy supports a balancing point between semantic specificity and operational agility.

The known benefits of describing types rather than implementing them include swift extensibility of some existing, or possibly non-existing, subclass hierarchy in UCO without requiring training in ontological development, taxonomic specification, or OWL, SHACL, or RDF maintenance logistics.

The known detractions of using string-literals for type descriptions include that used vocabularies may require careful maintenance among data-sharing parties;
that vocabularies require independent logistics (external to UCO) for providing definitions (i.e., dictionary-style semantics) to string-literals chosen;
and that string-literals cannot by themselves encode hierarchical structure or entailments, such as the informal device type string 'ExamplePhone 8 P4321' entailing 'ExamplePhone 8', 'ExamplePhone', or 'ExamplePhone models discontinued in 2020'.

Usage of Informal Type to house strings should be weighed against usage of classes when classes are available, and should periodically be reviewed for potential additions to UCO's class hierarchy or downstream extensions thereof.

All properties P that currently house an informal type are set as sub-properties of core:informalType: P rdfs:subPropertyOf core:informalType. Note this includes one property in CASE, investigation:investigationForm.

In satisfaction of Requirement 2, core:informalType does not have an rdfs:range. Though all of the child properties have the option of xsd:string in their range, some use a union of xsd:string with a semi-open vocabulary datatype. Even though Issue 629 will adjust typing of vocabulary members to be xsd:string, the sub-property linkage may interact poorly with owl:unionOf, since multiple rdfs:range statements are interpreted in OWL as an intersection.

Coordination

• Tracking in Jira ticket OCUCO-149
• Administrative review completed, proposal announced to Ontology Committees (OCs) on 2024-10-28
• Requirements to be discussed in OC meeting, 2024-11-21
• Requirements Review vote occurred, passing, on 2024-11-21
• Requirements development phase completed.
• Solution announced to OCs on 2024-11-22
• Solutions Approval to be discussed in OC meeting, 2024-12-10
• Solutions Approval vote occurred, passing, on 2024-12-10
• Solutions development phase completed.
• Backwards-compatible implementation merged into develop for the next UCO release
• Backwards-compatible implementation merged into develop for the next CASE release
• develop state with backwards-compatible implementation merged into UCO develop-2.0.0
• develop state with backwards-compatible implementation merged into CASE develop-2.0.0
• Backwards-incompatible implementation merged into develop-2.0.0 (N/A)
• Milestone linked
• Documentation logged in pending UCO release page
• Documentation logged in pending CASE release page