-
Notifications
You must be signed in to change notification settings - Fork 2
Home
Eugeniu Costetchi edited this page Apr 25, 2019
·
7 revisions
Need to migrate the report away from the Latex towards single page HTML5. Ideally the approach would use a combination of a template language (e.g. Mustache, Jinja etc.) and SPARQL queries.
Alternatively search for projects that implement some forms of document generation based on RDF structures.
One such technology is [SWP](SPARQL Web Pages - User Guide) highlighted here. By the way I just discovered that TopBraid just made SWP available in all versions of TopbraidComposer (presumably the free one as well). Is there any independent SWP implementation? Can we mimic something like that with Mustache or implement the spec?
Next is a list of possibly relevant projects:
- Another project implement this goal of LODE and WiDoco projects focusing on visualizing ontologies.
- OntoSpy
- Frensel
- RDForm
- OSLO spec gen
- RDF2HTML converter1
- RDF2HTML converter2
- BIB LOD to HTML
- how about setting up a simple project which takes a structured set of Sparql queries, generated JSON structure with the sparql query results, which is then fed into the document templates using mustache templating language
- once there is an xsl-fo document then it can be transformed into HTML, PDL, word, Markup, etc
- revise the concepts
- search for lexicalisation and annotation properties ** rdfs:label, rdfs:comment ** skos:prefLabel, skos:altLabel,skos:definition,skos:scopeNote (expecting them to be data properties) ** skos:prefLabel, skos:altLabel,skos:definition,skos:scopeNote (expecting them to be reified data properties like VocBench3 does) ** skosxl:prefLabel, skosxl:altLabel (expecting them to be object properties)
- add possibility to configure other, new annotation properties to be included
- create the lexicalisation profile per resource type i.e. what lexicalisation properties are used and which languages are covered for each of them
- note: un-typed resources with annotations are are of undefined type
- create AP for the lexicalisation properties only (i.e. domain type, property, range-language, cardinality, confidence/frequency)
- take the SPARQL queries from the VOID spec
- add a report on dependecy between namespaces (see already existing source code in another project)
- given a source and target vocabulary and an aliment (in EDOAL or SKOS formats) provide statistics answering some questions as the following.
- what is the number of distinct subjects from source involved in alignment?
- what is the number of distinct objects from target involved in alignment?
- what are the relative portions covered by the alignment?
- for each concept that has children (i.e. is a parent) get the number of direct children, distance to the top concept, number of parents
- aggregated number of children at depth level
- simplified diagram of the graph shape
- tree symmetry metrics
- search and implement for graph/tree metrics
- search and implement forest metrics (e.g compare trees to each other)
- read few papers and add ontology metrics that apply to SKOS and other instance data