Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Suggested Representation Improvements #3

Open
1 of 10 tasks
VladimirAlexiev opened this issue Dec 1, 2023 · 0 comments
Open
1 of 10 tasks

Suggested Representation Improvements #3

VladimirAlexiev opened this issue Dec 1, 2023 · 0 comments

Comments

@VladimirAlexiev
Copy link

VladimirAlexiev commented Dec 1, 2023

  • Many strings have datatype ^^xsd:string.
    This is the default literal datatype and according to specs, it is redundant: eg modality "Texts" and "Texts"^^xsd:string are precisely the same literal.
    However, Virtuoso doesn't treat them the same:

Eg foaf:name values are printed with a long ugly datatype, so eg I had to do this to have them printed as a simple string:

  ?task a lpwcc:task; foaf:name ?name1
  bind(str(?name1) as ?name)

Eg if you're looking for Text datasets, you have to write like this. If you omit the datatype, nothing is found

?dataset a lpwcc:dataset; lpwc:modality "Texts"^^xsd:string

So please: change the mapping to NOT emit ^^xsd:string if possible

  • Uses two ontology prefixes. Would be better to use just one:
PREFIX lpwc: <https://linkedpaperswithcode.com/property/>
PREFIX lpwcc: <https://linkedpaperswithcode.com/class/>
  • The diagram shows class names Capitalized but in RDF they are in lowercase.
    Would be better to use Capitalized; else fix the diagram
  • Semantic resolution without Accept header returns RDF/XML. Would be better to return Turtle
  • For every resource, include a link to the original site, eg
<https://linkedpaperswithcode.com/task/analogy-questions-bats>
  rdfs:seeAlso <https://paperswithcode.com/task/analogy-questions-bats>

Actually it's very easy to go from the semantic URL to the web URL (just remove the word "linked"), but it would be nicer to have a link as well.

  • Deploy YASGUI: it's a better SPARQL editor than Virtuoso's: prefix addition, code highlighting, ontology term autocompletion, viewing the query and results together.

https://yasgui.triply.cc/ can be used with the https://linkedpaperswithcode.com/sparql endpoint without any problem, see query link below, and more in #1 and #2

  • Add to the repo all prefixes from the diagram so they are used in the response.

Eg here is a saved query that defines all relevant prefixes, so the "property" column in the response is short

I've added https://prefix.cc/lpwc (should appear shortly), but can add only 1 per day: add the other one

The idea to reuse ontologies is a good one, but you should reuse only appropriate ontologies.

  • Don't use dbp: properties. Unlike dbo:, these are the "raw" DBpedia props coming from English Wikipedia.
    Eg dbp:area is completely inappropriate since it expresses the acreage of some land, not an area of research.

  • The correct namespace for dbp: uses http, as you can see at https://prefix.cc/dbp . However, you define it to use https
    As a result, if you convert linkedpaperswithcode-ontology.owl to ttl, you see long URLs:

@prefix dbp:     <http://dbpedia.org/property/> . # NOT used below
<https://dbpedia.org/property/acronym>
        rdf:type     owl:DatatypeProperty ;
        rdfs:domain  <https://linkedpaperswithcode.com/class/conference> ;
        rdfs:range   xsd:string .
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant