REFORMERS — Ontology Extensions & Knowledge Graphs

Knowledge-graph workspaces for the REFORMERS project. Each subfolder is one partner's workspace. Each partner fills in a shared Excel template with their data, then converts it to a TTL file. That TTL is the workspace's knowledge graph.

🚀 First time here? Read STEP_BY_STEP.md. It walks you through everything in about 15 minutes.

Project phases

Phase 1 is the priority. Get the data in first.

Phase	What happens	Status
1. Knowledge graphs & ontology extensions	Each partner fills the Excel template and converts it to TTL. New classes go in the workspace's ontology/extensions/.	Current priority
2. Services & scenarios	Register service requirements. Build what-if scenarios.	After Phase 1
3. RDP integration	Integrate workspaces and platform into the REFORMERS RDP architecture.	After Phase 2

Workspaces

Folder	Lead
WindForecasting_AIT	AIT
InvestaModel_AIT	AIT
InvestaScenarios_AIT	AIT
HeilooControl_TUD	TU Delft
CongestionPolicyTool_TUD	TU Delft
SolarForecasting_CIRCE	CIRCE
EVForecasting_VUB	VUB
PriceForecasting_CERTH	CERTH
BuildingFlexibility_EMPA	Empa
EnergySystemTechnologyDatabase	Worked example

The Excel template

Every workspace ships with a blank copy of the template at:

<WorkspaceFolder>/ingestion/input/data_ingestion_template.xlsx

Rename it to match your workspace (e.g. WindForecasting_AIT.xlsx). The converter uses the filename for the project URI.

The latest blank template is always at the platform repo: data_ingestion_template.xlsx.

How to fill it in: template README. It covers the header rows, the 15 attribute types, and how to cite your sources.

How to populate your workspace

Full walkthrough: STEP_BY_STEP.md. Short version:

1. Open your workspace's ingestion/input/data_ingestion_template.xlsx. Rename it.

2. Fill in the rows.

3. Run this:

   docker exec digicities-streamlit \
       python -m backend.replica_builder.cli \
       /app/data/usecases/<WorkspaceFolder>/ingestion/input/<your_workbook>.xlsx

4. The TTL lands at <WorkspaceFolder>/ingestion/output/<your_workbook>.ttl. Commit it.

That's it for Phase 1. Scenarios, services, and custom queries come later.

CLI options

python -m backend.replica_builder.cli <xlsx-path> [--output PATH] [--project-uri URI] [--uri-mode MODE]

• <xlsx-path>: required. Container path to the workbook.
• --output / -o: override output path. Default: <workspace>/ingestion/output/<stem>.ttl.
• --project-uri: override the URI. Default: https://digicities.info/proj/<stem>.
• --uri-mode: default (most workbooks), full-uri-in-cell, or complete-project-uri. See the template README.

Look at the example first

EnergySystemTechnologyDatabase/ has a real filled-in workbook (MotelDB.xlsx) and its converted TTL (MotelDB.ttl). Open it in the UI to see what a populated workspace looks like before doing your own.

Plug this repo into Digicities

The platform discovers workspaces by scanning a parent folder. Mount this repo at that folder:

# digicities-platform/docker-compose.override.yml
services:
  streamlit:
    volumes:
      - ../REFORMERS_Ontology-Extensions-and-Knowledge-Graphs:/app/data/usecases
    environment:
      USECASES_DIR: /app/data/usecases

Restart Streamlit, refresh http://localhost:8501. All workspaces in this repo show up.

Customise your workspace

Two files control how your workspace appears on the landing page.

workspace_meta/metadata.json sets the name, description, and tags.

{
  "id": "WindForecasting_AIT",
  "name": "Wind Forecasting (AIT)",
  "description": "Short-term wind power forecasting models and supporting data.",
  "created": "2026-05-28",
  "tags": ["forecasting", "wind", "AIT"]
}

Edit name, description, and tags freely. Don't change id. The platform uses it internally.

workspace_meta/image.png is the thumbnail. Replace it with your own (.png, .jpg, or .jpeg). About 400x300 px works well.

Changes appear on the next workspace load. The platform caches for 5 minutes, so hard-refresh (Ctrl+F5) or restart the streamlit container if you want them sooner.

Conventions

• One topic, one folder. Don't mix data from two partners.
• Folder names: <Topic>_<Partner>. PascalCase topic, partner abbreviation.
• Excel workbooks in ingestion/input/ are gitignored by default. The example workbook in EnergySystemTechnologyDatabase is tracked as an exception. To track your own, add it explicitly.

Citing the platform and ontology

This repo is just workspace data. It doesn't carry its own citation. If you publish work using the underlying tools, cite these two:

• Digicities platform (Apache 2.0): see its README for the BibTeX.
• Digicities ontology (CC BY 4.0): see its README for the BibTeX.

Need help?

Open an issue on this repository.

Related

• Digicities platform: uesl-empa/digicities-platform
• Core ontology: uesl-empa/digicities-ontology
• Excel template: digicities-platform/data/ingestion_template/

Funding & acknowledgements

Digicities was funded through the SFOE P+D program under the ERA-Net Smart Energy Systems joint initiative Digital Transformation for the Energy Transition, grant agreement No 88397.