Intuitiveness

Intuitiveness as the next stage of open data
Dataset design and complexity

---

What is Intuitiveness?

Intuitiveness is a methodology and Python package for transforming raw, complex datasets into purpose-built data that directly answers your questions.

The method works through a descent-ascent cycle:

• Descent (L4 → L0): Strip away complexity to find the core truth
• Ascent (L0 → L3): Rebuild with YOUR intent, adding only relevant dimensions

The 5 Levels of Abstraction

Level	Name	Description
L4	Raw Dataset	Original tabular data
L3	Entity Graph	Knowledge graph of relationships
L2	Domain Categories	Grouped by semantic domains
L1	Feature Vector	Unified numeric representation
L0	Core Datum	Single atomic value (the truth)

Features

• Natural Language Search: Query French open data (data.gouv.fr) in plain French using SmolLM3-3B
• Interactive Streamlit App: Visual descent-ascent workflow
• Knowledge Graph: Neo4j-powered entity relationships
• Semantic Matching: AI-assisted domain categorization

Quick Start

# Clone the repository
git clone https://github.com/ArthurSrz/intuitiveness.git
cd intuitiveness

# Install dependencies
pip install -r requirements.txt

# Run the app
streamlit run app.py

Scientific Article

This work is documented in a peer-reviewed research paper:

• PDF: scientific_article/Intuitiveness.pdf
• Revised (v2): scientific_article/v2_intuitive_datasets_revised.md

Prerequisites

Neo4j Database

docker run -d --name neo4j -p 7474:7474 -p 7687:7687 \
  -e NEO4J_AUTH=neo4j/password -e NEO4J_PLUGINS='["apoc"]' neo4j:latest

HuggingFace Token (for NL queries)

Set HF_TOKEN environment variable or add to .streamlit/secrets.toml:

HF_TOKEN = "your_token_here"

Project Structure

intuitiveness/
├── intuitiveness/          # Core package
│   ├── data_sources/       # MCP client & NL query engine
│   ├── services/           # data.gouv.fr search service
│   └── ui/                 # Streamlit components
├── scientific_article/     # Research paper
├── app.py                  # Main Streamlit app
└── requirements.txt

Acknowledgments

Part of the Dataflow research project, funded by:

• Datactivist
• UNESCO Chair in AI and Data Science for Society

---

_{Designed by Arthur Sarazin & Mathis Mourey}