(diff) ←Older revision | Current revision (diff) | Newer revision→ (diff)
Review Summary: The pattern brings a nontrivial modeling problem to the domain of process and workflow representation.
If the pattern is revised to reuse existing patterns, provides less in-house namespacing and integrates more entity annotations, it will be a fine addition and a prime candidate for the catalogue. To that end, it would be good to issue a revised version with its own version IRI.
Some rather feasible presentation issues in the pattern page should be addressed as well.
Reviewer Confidence: High (ontologies, OWL, ontology design patterns), Medium (process modeling)
Problems: - From the competency questions, it is not clear if input, output, conditions and triggers (events) need to be instantiated as actual values, or simply *categories* of parameters, conditions and events.
- Wouldn't it be possible to relax existential restrictions on input/output parameters for the Process class?
- Typo: Property "hasEnvironemntalCondition" should be "hasEnvironmentalCondition"
Community Relevance: High, due to its attempt at tackling the cause/effect representation problem and bringing it to the domain of process models.
Relation to Best Practices: It could set a standard solution for process modeling in various domains, provided that its relation with patterns like Parameter and Reaction is well-defined.
Reusability: Reusable, with the possible drawback of namespacing. Every entity is defined in-house with namespace "
http://purl.org/biomass/ReactorPattern#", which means that if I reuse this pattern, it could remain isolated from other imported definitions of Process, Event, etc. unless I manually align them. The pattern itself should reuse these base concepts.
Relations to Other Patterns: It would probably be better off specializing some other content pattern, e.g. Reaction.
The "Parameter" pattern could also be specialized.
Overall Understandability: The intent and implementation are very clear, sans the issues described in other fields.
Clear Problem Description: Clear, but it seems to tackle two problems, i.e. parametric processes and cause/effect relations. Content Patterns should be atomic solutions whenever possible.
Clear Relevance and Consequences: Clear. Thers's one extra newline in the "Consequences" field.
Clear Figures and Illustrations: Clear and understandable, though it would be better to develop the figure vertically a little more.
Missing Information: - Labels and comments for most entities
- There is a scenario annotation in the ontology - It would be good to add a more elaborate version of it to the "Scenarios" section of the pattern page (not just the field in the general description).
- The OntoMDL implementation should be mentioned (with a link to the ontology) in the "Known Uses" field.
- I personally would suggest that Content Patterns start to use OWL 2 version IRI, given that they are subject to refinement -> versioning. Please note that version IRIs should resolve to the specific version of the ontology, while the ontology IRI should resolve to the latest stable release.
