On Sun, 2020-03-08 at 12:06 +0530, Diptendu Dutta wrote:
> I have used Jena to generate RDF/XML of the model:
> 
> <rdf:RDF
>     xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#";
>     xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#";>
>   <rdf:Description>
>     <rdf:subject rdf:parseType="Resource">
>       <rdfs:label>TheConsignor</rdfs:label>
>     </rdf:subject>
>     <rdf:predicate rdf:resource="http://www.lke.com/lke.owl#shall%20deliver
> "/>
>     <rdf:object rdf:parseType="Resource">
>       <rdfs:label>theConsignedUnits</rdfs:label>
>     </rdf:object>
>   </rdf:Description>
>   <rdf:Description>
>     <rdf:subject rdf:parseType="Resource">
>       <rdfs:label>rulesofinterpretation</rdfs:label>
>     </rdf:subject>
>     <rdf:predicate rdf:resource="http://www.lke.com/lke.owl#apply%20in"/>
>     <rdf:object rdf:parseType="Resource">
>       <rdfs:label>thisConsignmentAgreement</rdfs:label>
>     </rdf:object>
>   </rdf:Description>
> .
> .
> </rdf:RDF>
> 
> Do I need to use some other format apart from RDF/XML so that
> the output my be be visualised in a browser?
> 
> I looked at "cytoscape.js", Owl2Vowl, and some others. They all require
> the data to be in some specific format.
> 
> Are there libraries available for  transforming the RDF/XML to a format
> suitable for display in browser?
> 
> Which approach would you suggest?

I have not worked in this area since 2015, but it remains the biggest
problem (and opportunity) in semantic technology.

Must you work with RDF/XML format? Jena provides many other options,
including csv, json-ld, and sparql results format (srx). These are all
easier to transform for presentation than RDF/XML. Will you be
transforming the RDF on the server or in the browser?

Next, do you want to create a static information display, or dynamic
display that responds to user input? Do you want a primarily text-based
layout (paragraphs, lists, tables), or graphical (boxes or bubbles and
lines)?

The approach I found most promising was to use d3.js [1] in the browser.
This allows you to create either text or graphical layouts, static or
dynamic. You can make ajax sparql queries, either CONSTRUCT or SELECT,
and transform the results to HTML5 (including SVG and canvas). What is
lacking, however, is a standard way to select and transform significant
graph patterns, analogous to XSLT templates. Perhaps the graph shape
languages, SHACL and ShEx, could be of some help in this area, but I
have not kept up in the last 5 years.

One of the big disappointments of semantic technologies is that we
haven't gotten far past the "connected bubbles" visualization of RDF
graphs. That is easy to do and most unhelpful. It is much harder to
create useful schema-specific information presentations from semantic
data that meet the real needs of the information consumers.

Regards,
--Paul

[1] https://d3js.org/

> 
> Regards,
> 
> Diptendu Dutta
> 
> Regards,
> 
> Diptendu Dutta


Reply via email to