Talk:Semantic Web Ontology Portal Evaluation Approach
From Mike Huhns
We might do a cost/benefit analysis for ESIP maintaining an ontology portal. The cost of maintaining a portal is the cost of a server and its connection to the Internet, plus the cost of someone to occasionally patch the software and update the portal software (currently based on Stanford's Bioportal). The benefit is harder to characterize, but might be estimated as follows. Suppose a researcher in earth science gathers data that is intended to be preserved and shared. The data is "more valuable" if it uses standard terms and concepts, which would be found in an existing ontology, which would be found at an ESIP portal. The data is also more valuable if it is organized logically, that is, if its terms and concepts are organized according to an ontology. The researcher could create his/her own ontology, which might take several person-months. The cost of the person-months is $X. Suppose there are N such researchers in the world. The ontology portal is worthwhile if Annual Maintenance Cost < N * $X Researchers would have to be convinced to pay a fee to ESIP for using the portal, thereby offsetting its maintenance cost. The fee would have to be included in the budget for their grant.
The definition given seems inaccurate or in the least too broad:
'ontology' refers to any semantic artifact, whether expressed in OWL/RDF, another OWL representation, or a simpler form like Excel or CSV that can be converted into a semantic artifact.
- Any semantic artifact? I think we need to be more specific here.
- Ontology is the conceptualization of entities and interactions in some particular domain of knowledge or practice.
- Then there is a representation of the ontology, such as OWL/XML, RDF/XML, ttl, n3, xsl, csv, etc...
- Are we using portal and repository interchangeably?
- Are we using this evaluation for the development of an ESIP repository? Or is this just a general evaluation of ontology repositories?
- Would an ontology portal simply store information about an ontology and not the ontology itself?
- What about discovery of ontologies? I.E. does the repository provide markup such as schema.org or RDFa for easier discoverability of ontologies?
- Visualizations of an ontology, there could be conceptual visualization, such as a CMAP, vs. a human readable visualization, such as LODE, vs. a machine readable representation such as RDF/XML.
- Is the repository domain specific?
- Ontology feedback? Can users provide comments or review an ontology?
- Citation capabilities? Perhaps a prov pingback capability or a means to cite an ontology, display citations for an ontology ...
Re: Ontology Definition and some thoughts -- Brandon 05:21, 15 May 2016 (MDT)
- +1 to everything Patrick has mentioned.
- Perhaps explicitly stating there will be a focus on, or restriction to, W3C standards (proposed or accepted) would help for the time being.