Datasets:

Simple7up
/

IOM

triad_index int64 1 13	source stringclasses 78 values	target stringclasses 78 values	relation stringclasses 3 values
1	T01_adv_P0	T01_adv_P1	none
1	T01_adv_P0	T01_adv_P2	none
1	T01_adv_P0	T01_ret_P0	none
1	T01_adv_P0	T01_ret_P1	creates
1	T01_adv_P0	T01_ret_P2	mirrorOf
1	T01_adv_P1	T01_adv_P0	none
1	T01_adv_P1	T01_adv_P2	none
1	T01_adv_P1	T01_ret_P0	creates
1	T01_adv_P1	T01_ret_P1	mirrorOf
1	T01_adv_P1	T01_ret_P2	creates
1	T01_adv_P2	T01_adv_P0	none
1	T01_adv_P2	T01_adv_P1	none
1	T01_adv_P2	T01_ret_P0	mirrorOf
1	T01_adv_P2	T01_ret_P1	creates
1	T01_adv_P2	T01_ret_P2	none
1	T01_ret_P0	T01_adv_P0	none
1	T01_ret_P0	T01_adv_P1	creates
1	T01_ret_P0	T01_adv_P2	mirrorOf
1	T01_ret_P0	T01_ret_P1	none
1	T01_ret_P0	T01_ret_P2	none
1	T01_ret_P1	T01_adv_P0	creates
1	T01_ret_P1	T01_adv_P1	mirrorOf
1	T01_ret_P1	T01_adv_P2	creates
1	T01_ret_P1	T01_ret_P0	none
1	T01_ret_P1	T01_ret_P2	none
1	T01_ret_P2	T01_adv_P0	mirrorOf
1	T01_ret_P2	T01_adv_P1	creates
1	T01_ret_P2	T01_adv_P2	none
1	T01_ret_P2	T01_ret_P0	none
1	T01_ret_P2	T01_ret_P1	none
2	T02_adv_P0	T02_adv_P1	none
2	T02_adv_P0	T02_adv_P2	none
2	T02_adv_P0	T02_ret_P0	none
2	T02_adv_P0	T02_ret_P1	creates
2	T02_adv_P0	T02_ret_P2	mirrorOf
2	T02_adv_P1	T02_adv_P0	none
2	T02_adv_P1	T02_adv_P2	none
2	T02_adv_P1	T02_ret_P0	creates
2	T02_adv_P1	T02_ret_P1	mirrorOf
2	T02_adv_P1	T02_ret_P2	creates
2	T02_adv_P2	T02_adv_P0	none
2	T02_adv_P2	T02_adv_P1	none
2	T02_adv_P2	T02_ret_P0	mirrorOf
2	T02_adv_P2	T02_ret_P1	creates
2	T02_adv_P2	T02_ret_P2	none
2	T02_ret_P0	T02_adv_P0	none
2	T02_ret_P0	T02_adv_P1	creates
2	T02_ret_P0	T02_adv_P2	mirrorOf
2	T02_ret_P0	T02_ret_P1	none
2	T02_ret_P0	T02_ret_P2	none
2	T02_ret_P1	T02_adv_P0	creates
2	T02_ret_P1	T02_adv_P1	mirrorOf
2	T02_ret_P1	T02_adv_P2	creates
2	T02_ret_P1	T02_ret_P0	none
2	T02_ret_P1	T02_ret_P2	none
2	T02_ret_P2	T02_adv_P0	mirrorOf
2	T02_ret_P2	T02_adv_P1	creates
2	T02_ret_P2	T02_adv_P2	none
2	T02_ret_P2	T02_ret_P0	none
2	T02_ret_P2	T02_ret_P1	none
3	T03_adv_P0	T03_adv_P1	none
3	T03_adv_P0	T03_adv_P2	none
3	T03_adv_P0	T03_ret_P0	none
3	T03_adv_P0	T03_ret_P1	creates
3	T03_adv_P0	T03_ret_P2	mirrorOf
3	T03_adv_P1	T03_adv_P0	none
3	T03_adv_P1	T03_adv_P2	none
3	T03_adv_P1	T03_ret_P0	creates
3	T03_adv_P1	T03_ret_P1	mirrorOf
3	T03_adv_P1	T03_ret_P2	creates
3	T03_adv_P2	T03_adv_P0	none
3	T03_adv_P2	T03_adv_P1	none
3	T03_adv_P2	T03_ret_P0	mirrorOf
3	T03_adv_P2	T03_ret_P1	creates
3	T03_adv_P2	T03_ret_P2	none
3	T03_ret_P0	T03_adv_P0	none
3	T03_ret_P0	T03_adv_P1	creates
3	T03_ret_P0	T03_adv_P2	mirrorOf
3	T03_ret_P0	T03_ret_P1	none
3	T03_ret_P0	T03_ret_P2	none
3	T03_ret_P1	T03_adv_P0	creates
3	T03_ret_P1	T03_adv_P1	mirrorOf
3	T03_ret_P1	T03_adv_P2	creates
3	T03_ret_P1	T03_ret_P0	none
3	T03_ret_P1	T03_ret_P2	none
3	T03_ret_P2	T03_adv_P0	mirrorOf
3	T03_ret_P2	T03_adv_P1	creates
3	T03_ret_P2	T03_adv_P2	none
3	T03_ret_P2	T03_ret_P0	none
3	T03_ret_P2	T03_ret_P1	none
4	T04_adv_P0	T04_adv_P1	none
4	T04_adv_P0	T04_adv_P2	none
4	T04_adv_P0	T04_ret_P0	none
4	T04_adv_P0	T04_ret_P1	creates
4	T04_adv_P0	T04_ret_P2	mirrorOf
4	T04_adv_P1	T04_adv_P0	none
4	T04_adv_P1	T04_adv_P2	none
4	T04_adv_P1	T04_ret_P0	creates
4	T04_adv_P1	T04_ret_P1	mirrorOf
4	T04_adv_P1	T04_ret_P2	creates

End of preview. Expand in Data Studio

IOM (Realidad I.O)

IOM es un marco formal para construir y verificar unidades informativas autocontenidas mediante cinco perspectivas, trece tríadas y relaciones de transformación (mirrorOf / creates).

La unidad informativa es la mínima estructura autocontenida capaz de conservar una distinción, relacionarla con su opuesto, integrarla en un sistema y transformarse sin perder la trazabilidad de su identidad.

La especificación canónica (data/iom_spec.json) se verifica con Lean 4, SHACL y regeneración idempotente de los artefactos derivados. Los dominios, anotadores y métricas de recuperación de reglas se tratan como encuentro, no como fundamento.

Datos tabulares

Archivo	Contenido
`datasets/triad_nodes.csv`	78 nodos de las 13 tríadas
`datasets/relation_candidates.csv`	390 relaciones candidatas entre nodos
`datasets/vector_steps.csv`	20 pasos de vectores

Contenido del repositorio

IOM/
├── IOM/                  # módulos Lean 4 (Core, FormalRules, SelfGenesis, ...)
├── data/                 # iom_spec.json y su JSON Schema
├── datasets/             # CSVs, evaluación independiente, plantillas de ingestión
├── ontology/             # io_shapes.ttl (formas SHACL)
├── scripts/              # generación, validación y verificación
├── tests/                # pruebas
├── experiments/          # mapeo de dominios (encuentro)
├── docs/                 # especificaciones, auditorías y manuales
├── .github/workflows/    # CI
├── paper.md
└── LICENSE

Reproducibilidad

python3 -m venv .venv && .venv/bin/python -m pip install -r requirements-lock.txt
.venv/bin/python scripts/self_bootstrap.py --passes 2
lake build
.venv/bin/python scripts/validate_shacl.py
.venv/bin/python scripts/measure_invariance.py

La verificación del núcleo se ejecuta también en GitHub Actions en cada push y pull request. Consulta REPRODUCCION.md para más detalle.

Alcance

El proyecto implementa una construcción formal de unidad informativa. No afirma, por sí solo, una unidad informativa absoluta independiente del entorno de ejecución (Python, Lean, sistema operativo).

Uso

from datasets import load_dataset

nodos = load_dataset("Simple7up/IOM", "triad_nodes")
relaciones = load_dataset("Simple7up/IOM", "relation_candidates")
pasos = load_dataset("Simple7up/IOM", "vector_steps")

Enlaces

Código fuente: https://github.com/luisreygarciafigueroa-creator/IOM

Licencia

MIT

Cita

@misc{iom,
  title  = {IOM (Realidad I.O)},
  author = {Garc{\'i}a Figueroa, Luis Rey},
  year   = {2026},
  url    = {https://huggingface.co/datasets/Simple7up/IOM}
}

Downloads last month: 175