The Linked Open Data Site Generator (LOD-SG) is a tool to create a static website from and for an RDF dataset. It uses the Jekyll static site generator with the Jekyll RDF plugin to create one HTML page per RDF resource in the input graph. The project contains a number of basic page templates to produce a generic website. By adding additional templates for specific kinds of RDF resources, LOD-SG can be the starting point for developing tailor-made sites for individual datasets.

… for generating and publishing a LOD dataset with Jekyll

• folder structure
• basic configuration
• basic dataset description using the VoID vocabulary
• a Makefile for orchestrating common tasks such as merging and converting RDF input files, generating the static website

… for generating HTML pages out of RDF resources (using the Liquid template language).

• a default page template
• a default page header
• a default page footer
• a dataset template (for instances of void:Dataset)
• some more templates for common resource types such as organizations, people, geographic features etc.
• page includes such as:
  • statement tables (resource as subject and resource as object)
  • a Mermaid graph showing the same statements
  • a Leaflet map display for geographic features

• The Jekyll static site generator with the jekyll-rdf plugin. Jekyll is based on the Ruby language, so you first need to make sure that you have Ruby available on your command line. Once that is ensured, you can install Jekyll like this (this command installs the dependencies that are defined in the Gemfile):

$ bundle install

• Python 3 with the rdflib library. If you have that, create a virtual environment, activate it and install rdflib via the requirements.txt file:

$ python -m venv venv
$ . venv/bin/activate
(venv) $ pip install -r requirements.txt

To generate a new static site for your dataset, you need to perform the following steps:

• Create a new repository that will be the home for your LOD-site.
  • Create this repository using the LOD-Browser repository as the template.
• Put a .ttl (Turtle) file with your RDF data in /data.
  • This is the bare minimum. Of course, you could also add code or other assets that will be used to generate the Turtle-file in /data. I usually put some scripts in /bin that convert the source (non-RDF) data into RDF, maybe some static boilerplate RDF in /data/static, and some plumbing in a /Makefile to orchestrate and automate the process of building the output data.
• Adjust the void.ttl file that describes your dataset.
• Adjust _config.yml to configure the creation of the static site.