[linked-art] Automatic JSON-LD to Zooming Data model visualisation

36 views
Skip to first unread message

Joseph Padfield

unread,
Aug 16, 2023, 10:24:20 AM8/16/23
to Art Linked
Dear All,

As many of you will know most of the Linked.Art examples provided as part of the documentation are displayed as graphical models. These models are generated using the Mermaid (https://github.com/mermaid-js/mermaid) diagram library.

I have been working for sometime on a system that simplifies the creation of these models from simple tab separated text see: https://national-gallery.github.io/dynamic-modelling/

It has also been possible to use the tool to automatically generate similar models directly from JSON-LD files, but until recently the results have been spotty.

Over the last few days I have managed to improve this functionality so that the Linked.Art model visualisations can be automatically generated. For example the diagram shown on https://linked.art/model/object/physical/#dimension-statement can be replicated from the supplied JSON-LD with: 

One can then zoom in and out and explore the model as well as potentially editing it as needed.

A simplified version of the the presentation which could be included in something like an iframe also now works, so that live visualisations can be embedded:
* https://national-gallery.github.io/dynamic-modelling/?url=https://linked.art/example/object/44&simple

The system does also work with larger Linked.Art examples such as: 

I have gone through most of the main examples on the Linked.Art website and all of the models and formatting seem to be replicated quite closely, but if anyone spots anything with those or larger examples please do let me know.

I should also say that the github url https://national-gallery.github.io/dynamic-modelling/ does redirect to a National Gallery server, but as the name of the local server is going to be changed in the near future it is better to use the github url in any recorded or bookmarked links.

Any thoughts or comments please let me know.

Thanks

Joe
Reply all
Reply to author
Forward
0 new messages