semantic-schemas 0.7.0

OO-LD schema library and tooling for materials science data

Semantic Schemas

A library of templates for recording materials science data in a way that machines can read, search, and connect to other datasets.

Each template covers one concept (e.g. chemical composition, specimen). You fill in a plain JSON file with your values; the tooling converts it into RDF, a standard format for linked data, using vocabulary from materials science ontologies such as DCAT and PMDCo.

New to this? Each schema folder has a Jupyter notebook in docs/. It walks you through the whole process step by step.

---

Repository structure

schemas/                       # Schema library
  <domain>/
    <ontology>/
      README.md                # What this schema is for and how to use it
      specs/
        schema.oold.yaml             # Full schema definition (expert reference)
        shape.ttl                    # SHACL validation rules (checks the output RDF)
        schema.simplified.json       # Input field reference
        transform.simplified.jsonata # Converts your input to the structured format
      docs/
        example.input.json     # Ready-to-edit example (start here)
        *.ipynb                # Step-by-step workflow notebook
templates/
  schema.oold.yaml             # Blank template for writing a new schema
docs/
  1_concepts.md                  # How the semantic schemas are designed and why
  2_oold-primer.md               # How the OO-LD schema format works
  3_schema-format.md             # Field-by-field reference for schema authors
  4_schema-patterns.md           # Inheritance and composition between schemas
.github/
  ISSUE_TEMPLATE/              # Propose or correct a schema
  PULL_REQUEST_TEMPLATE.md

---

Installation

Using pip (recommended)

pip install semantic-schemas

Development installation

For development or contributions, clone the repository and install in editable mode:

git clone https://github.com/Semantic-Dataspace/semantic-schemas
cd semantic-schemas
python3 -m venv .venv
source .venv/bin/activate           # Windows: .venv\Scripts\activate
pip install -e ".[dev]"

Note: The semantic_schemas package is used by all workflow notebooks. Without it, the cells that call from semantic_schemas import Schema will fail.

---

Quick start

I want to record data for an existing schema

1. Find the schema in CATALOG.md and open its folder.
2. Copy docs/example.input.json and fill in your values (no ontology knowledge required).
3. Open the Jupyter notebook in docs/ and run all cells (Kernel → Restart & Run All). It converts your file to RDF and validates the result automatically.

I want to understand how the semantic schemas work

Read docs/1_concepts.md for the big picture: why there are two schemas per concept (one simple for users, one expert for machines), what the transform does, and how the pipeline fits together. Then docs/2_oold-primer.md if you want to go deeper into the OO-LD (Object-Oriented Linked Data) format itself.

I want to contribute a new schema

1. Read CONTRIBUTING.md for the workflow and conventions.
2. Copy templates/schema.oold.yaml as your starting point.
3. Open a New Schema issue to discuss the pattern before submitting a PR.

---

Documentation

Document	Content
docs/1_concepts.md	How the semantic schemas are designed: two layers, the transform, the full pipeline
docs/2_oold-primer.md	How the OO-LD (Object-Oriented Linked Data) schema format works
docs/3_schema-format.md	Field reference for writing and reviewing schemas
docs/4_schema-patterns.md	Inheritance and composition between schemas
docs/5_llm-schema-guide.md	Step-by-step guide for drafting a new schema with an LLM

---

Related projects

Project	Role
OO-LD	The schema format specification and Python tooling
PMDCo	Platform MaterialDigital Core Ontology
BWMD Ontology	Fraunhofer IWM materials ontology

---

The k-item field type

Some fields in these schemas link to data containers known as Knowledge Items (short: k-items) in a live knowledge management system rather than to a static list of values from an ontology. These containers encapsulate data, metadata, attachments, executables (apps), and more. In JSON, the value is a URI pointing to the k-item’s web page.

See docs/3_schema-format.md for the field syntax. Background and motivation:

Nahshon, Y.; Morand, L.; Büschelberger, M.; Helm, D.; Kumaraswamy, K.; Zierep, P.; Weber, M.; de Andrés, P. (2025). Semantic Orchestration and Exploitation of Material Data: A Dataspace Solution Demonstrated on Steel and Copper Applications. Advanced Engineering Materials, 27(8), 2401448. https://doi.org/10.1002/adem.202401448

---

License

Schemas are published under CC0 1.0 Universal, with no rights reserved. You may use, adapt, and redistribute them freely, including for commercial purposes, without attribution.