Research Platform

The BECACO
Tool Suite

A set of purpose-built digital tools for cultural heritage provenance research — from knowledge graph integration to AI-powered narratives, conversational queries, and multi-dimensional analytics.

Request Access →
4
Tools
CIDOC
CRM Standard
LLM
AI Powered
Graph DB
SPARQL Backend

One Integrated Platform

01
🔗
BecaGraph
Transforms tabular museum exports into a CIDOC CRM knowledge graph with AI entity resolution across institutions.
02
💬
BecaChat [in progress]
Conversational interface to query the knowledge graph in plain language — no Cypher expertise required.
03
◈
Beca Narratives [in progress]
Generates historically informed narratives from graph data — with text-to-speech audio playback for museum audiences.
04
📊
Analytics Dashboard
Six analytical modules plus five specialised tools — provenance, actors, collections, and data quality.
◆ ◆ ◆

BecaGraph

CIDOC CRM Entity Resolution Graph DB SPARQL
From Spreadsheet
to Knowledge Graph
Data Integration Engine · The Foundation

BecaGraph is the integration engine that powers the entire platform. It transforms raw Excel or CSV exports from museum collection management systems into a CIDOC CRM-compliant RDF knowledge graph queryable via SPARQL — with LLM-powered entity resolution that identifies the same actor, place, or object across multiple institutional datasets.

📂
Upload objects, actors, events, and institutions from any museum data export (XLSX · CSV)
🧠
Semantic matching resolves duplicate entities across institutions at configurable thresholds
🗺️
CIDOC CRM classes mapped — E22 Objects, E39 Actors, E8 Acquisitions, E53 Places…
Request Access →
becagraph · pipeline
// Connecting to Graph DB (SPARQL)...
status: connected ✓
 
// Loading objects (1,247 rows)
museum_id: "MQB"
 
// Entity resolution @ 87%
clusters: 312 detected ✓
nodes: 3,891  ·  edges: 7,234
 
▸
📂
Ingest
🗺️
Map CRM
🧠
Resolve
🕸️
Graph
◆

BecaChat [in progress]

Conversational AI Live SPARQL
Ask the Graph,
Get Answers
Natural Language Knowledge Graph Query

BecaChat lets researchers and museum professionals query the BECACO knowledge graph in plain language — no SPARQL expertise needed. Type a question, and the LLM translates it into a SPARQL query, executes it against the Graph DB endpoint, and returns a structured, readable answer.

🗣️
Accepts questions in plain English — "Who acquired this object from whom and when?"
⚡
Automatically generates and executes SPARQL queries against the live Graph DB endpoint
🔍
Returns structured, explainable answers with provenance chain details
Who supplied objects from Mexico to European museums between 1950 and 1980?
BecaChat · Graph Query
SELECT ?actor (COUNT(?obj) AS ?count)
WHERE { ?obj :culture "Mexico" ;
        :acquisitionYear ?y .
  FILTER(?y >= 1950)
  ?actor :supplied ?obj }
ORDER BY DESC(?count)
Found 180 actors — top result: G. Schmidt Pizarro (5120 objects across 7 museums).
Show me the full provenance chain for object MQB-0447.
BecaChat · SPARQL Trace
Tracing 4-step chain: Oaxaca (1962) → Schmidt Pizarro → Sotheby's NY (1967) → Leiden Volkenkunde (1971)
Ask about actors, objects, events…
→
◆ ◆ ◆

BecaNarratives [in progress]

Story Generation Text-to-Speech Multi-Modal
Objects with
Stories to Tell
Object Story Weaver · LLM Narrative Engine

Beca Narratives bridges the gap between raw provenance data and human-readable stories. Enter any object identifier from the knowledge graph, and the tool retrieves its structured metadata, feeds it to an LLM, and generates a compelling, historically informed narrative. The text is then converted to audio for accessible, multi-modal playback — validated by both domain experts and museum visitors.

🔍
Autocomplete search across all object identifiers in the graph
📜
Structured metadata panel + LLM-generated narrative displayed side by side
🔊
Text-to-speech playback with play/stop controls — accessible to non-specialist audiences
ID › MQB-0447
Generate
📜 Structured Information
Object: Mixtec turquoise mosaic
Culture: Mixtec · Mexico
Date: ca. 1400–1521 CE
Acquired: Leiden, 1971 · G. Schmidt Pizarro
◈ Generated Narrative
This turquoise mosaic, its surface still catching light as it did six centuries ago in Oaxaca, carries the memory of hands that shaped it long before it crossed the Atlantic. Acquired by Leiden in 1971 through the dealer Guillermo Schmidt Pizarro, its journey from a ritual deposit to a European showcase reflects the complex networks at the heart of the BECACO project…
▶ Play Narrative
■ Stop
◆

Analytics Dashboard

6 + 5 Modules Interactive Charts 3D Network
Six Lenses on
Collection Data
Multi-Dimensional Analytics Suite · + 5 Specialised Tools

The Analytics Dashboard provides six integrated modules for exploring the knowledge graph from temporal, spatial, relational, and qualitative angles. It also groups five additional specialised tools — provenance tracing, actor intelligence, collection browsing, and data quality monitoring.

⏱️
Time & PeriodsTimeline · cultural periods · production vs acquisition comparison
🌍
GeographySpatial distribution of production places · country breakdown
🌐
Production3D force-graph network linking places, techniques, actors and production events
📈
AcquisitionOwnership transfer timeline · top acquiring institutions · transaction type breakdown
🔍
Advanced SearchMulti-criteria queries across object, spatial-temporal, and actor dimensions with 7 search modes
Also included
🔎
Provenance Explorer
Full ownership chains from production to current location
👤
Actors & Provenance
Actor network browser with cross-museum timelines
🧑‍🏛️
Actor Profile
Deep-dive: objects created, collected, acquired per actor
📦
Collections Overview
Cross-museum object browser with filters & export
🏆
Data Quality
Completeness scores per field and institution — material, date, actor, place coverage with missing-data heatmaps
Time
Geography
Production
+3
⏱️
Time
🌍
Geography
🌐
Production
📈
Acquisition
🔵
Clustering
🔍
Advanced
Objects by century
Earliest: 3000 BC  ·  Latest: 2024
Fuzzy dates: 142 (11%)  ·  Avg gap: 287y

Want to Use
These Tools?

The BECACO tool suite is designed for research collaboration with ethnographic and archaeological museums. If you are interested in integrating your collection data or exploring any of these tools for your own provenance research, get in touch.

↑