Re: Visualizing RDF graphs in Jupyter notebooks

Hi Agnieszka,

If you're not opposed to using a triplestore like blazegraph you might
checkout the graph-notebook nbextenstion from AWS:
https://github.com/aws/graph-notebook

Cheers,

Nolan

On Wed, Feb 24, 2021 at 1:07 PM Agnieszka Lawrynowicz <
alawrynowicz@cs.put.poznan.pl> wrote:

> Dear Martin and Nicholas,
>
> Thank you a lot for all your feedback!
> It really helps.
>
> Best Regards and cheers,
> Agnieszka
>
> > Hi,
> >
> > On Wed, Feb 24, 2021 at 04:00:06PM +0100, Martin Hepp wrote:
> >
> >> I would likely start with Graphviz
> >
> > There is a rdf2dot tool in the rdflib library that might prove useful:
> >
> >
> https://rdflib.readthedocs.io/en/stable/apidocs/rdflib.tools.html?highlight=rdf2dot#module-rdflib.tools.rdf2dot
> >
> >> > For instance, imagine a notebook that uses rdflib (and possibly
> >> owlready2)
> >> > to create triples (a knowledge graph in general), which later on one
> >> would
> >> > like to visualize from within the same notebook.
> >
> > When it comes to manipulating and visualizing graphs in Python,
> > networkx is an option you need to consider:
> >
> >    https://networkx.org/
> >
> > On the Jupyter blog, you can also read about Cytoscape:
> >
> >    https://cytoscape.org/
> >
> > Looking at their "app store", I find
> > http://apps.cytoscape.org/search?q=semantic and
> > http://apps.cytoscape.org/apps/semscape
> >
> > As you can see, there are many options.
> >
> > I can not recommend one in particular, for I agree with Martin Hepp
> > that generic visualisation of RDF graphs is not that useful. Most of
> > the time you need to do a lot of filtering and pre-processing before
> > you can display something that provides readable information.
> >
> > Hope this helps !
> >
> > --
> > Nicolas Chauvat
> >
> > logilab.fr - services en informatique scientifique et gestion de
> > connaissances
> >
> >
>
>
>
>