Digital Albini

The Giuseppe Albini Archive digitization project transforms hierarchical archival metadata into a semantic knowledge graph using the Records in Contexts (RiC-O) ontology. The conversion process is implemented in Python and produces RDF/Turtle output suitable for SPARQL querying and linked data integration.

Features

• Semantic Modeling: Uses RiC-O ontology for archival description
• External Enrichment: Integrates GeoNames data for geographic entities
• Authority Control: Links to VIAF for persons and corporate bodies
• Hierarchical Preservation: Maintains archive's original record structure

Table of Contents

1. Archive Structure
2. Technical Architecture
3. Dependencies and Configuration
4. Ontologies and Namespaces
5. Output Specifications

---

Archive Structure

The Albini collection follows a hierarchical organization mapped to RiC-O classes:

Serie (Series)
└── Sottoserie (Sub-series)
    └── Busta (Folder/Box)
        └── Fascicolo (File)
            └── Documento (Document)
                └── Immagine (Image)

Hierarchical Mapping

Level	Italian Term	RiC-O Class	Count
1	Serie	rico:RecordSet	4
2	Sottoserie	rico:RecordSet	7
3	Busta	rico:RecordSet	36
4	Fascicolo	rico:RecordSet	581
5	Documento	rico:Record	1,852
6	Immagine	rico:RecordResource	6,184

---

Technical Architecture

Input Files

The conversion process requires two Excel workbooks:

1. mapping1.xlsx - Defines column-to-predicate mappings for each archival level

   • Each sheet corresponds to an archival level (Serie, Sottoserie, Fascicolo, Documento)
   • Columns: Subject, Predicate, Object, Column Subject, Column Object

2. instances.xlsx - Contains the actual archival metadata

   • Sheet names must match those in mapping file (case-insensitive)
   • Each row represents an archival entity with its descriptive metadata

Output File

• letters_kg.ttl - RDF knowledge graph in Turtle format
• Base namespace: http://www.w3.org/ns/DigitalAlbini#
• Fully self-contained with all entity definitions and relationships

Script Location and Execution

# File structure
digital-albini/
├── letters_dataset/
│   ├── src/
│   │   └── script.py          # Main conversion script
│   ├── data/
│   │   ├── input/
│   │   │   ├── mapping1.xlsx
│   │   │   └── instances.xlsx
│   │   └── output/
│   │       └── letters_kg.ttl
│   └── docs/
│       ├── busta_uml.jpg
│       ├── serie_uml.jpg
│       └── sottoserie_uml.jpg

# Execute conversion
cd letters_dataset/src
python script.py

---

Dependencies and Configuration

Required Python Libraries

import pandas as pd        # Excel file processing
import re                  # Regular expressions for parsing
import os                  # File path handling
import requests            # GeoNames API calls
from rdflib import Graph, Namespace, URIRef, Literal
from rdflib.namespace import RDF, RDFS, XSD, OWL
from datetime import datetime

Optional but Recommended:

pip install geonamescache  # Local GeoNames database cache

Configuration Constants

# File paths (relative to project root)
BASE_DIR = os.path.dirname(os.path.abspath(__file__))
PROJECT_ROOT = os.path.dirname(BASE_DIR)
mapping_path = os.path.join(PROJECT_ROOT, "data", "input", "mapping1.xlsx")
instances_path = os.path.join(PROJECT_ROOT, "data", "input", "instances.xlsx")
output_path = os.path.join(PROJECT_ROOT, "data", "output", "letters_kg.ttl")

# Namespace configuration
BASE_NS = "http://www.w3.org/ns/DigitalAlbini#"

# External services
GEONAMES_USERNAME = "your_username_here"  # Required for API access

GeoNames Integration

The script supports two modes for geographic enrichment:

1. Local Cache (geonamescache) - Fast, offline lookups
2. API Fallback - Online queries when local data unavailable

Setup GeoNames Account:

1. Register at https://www.geonames.org/login
2. Enable free web services in account settings
3. Replace GEONAMES_USERNAME in script configuration

---

Ontologies and Namespaces

Primary Ontology

Records in Contexts Ontology (RiC-O) 0.2

• URL: https://www.ica.org/standards/RiC/ontology
• Developed by: International Council on Archives (ICA)
• Purpose: Archival description and relationships

Supporting Ontologies

Prefix	Namespace	Purpose
rico	https://www.ica.org/standards/RiC/ontology#	Core archival classes and properties
wgs84	http://www.w3.org/2003/01/geo/wgs84_pos#	Geographic coordinates
gn	http://www.geonames.org/ontology#	GeoNames feature classification
owl	http://www.w3.org/2002/07/owl#	Ontology alignment (sameAs)
xsd	http://www.w3.org/2001/XMLSchema#	Datatype definitions
rdfs	http://www.w3.org/2000/01/rdf-schema#	Labels and schema

Custom Namespaces (Digital Albini)

All local entities use the base namespace: http://www.w3.org/ns/DigitalAlbini#

Namespace	Usage	Example URI
record/	Individual documents (rico:Record)	DigitalAlbini#record/S1_SS1_B1_F1_D1
recordset/	Aggregations (rico:RecordSet)	DigitalAlbini#recordset/S1_SS1_B1
person/	Individual agents	DigitalAlbini#person/Giovanni_Gentile
corporateBody/	Institutional agents	DigitalAlbini#corporateBody/University_of_Bologna
place/	Geographic locations	DigitalAlbini#place/Roma
physloc/	Physical locations	DigitalAlbini#physloc/Bologna
date/	Temporal entities	DigitalAlbini#date/1920
title/	Document titles	DigitalAlbini#title/Corrispondenza_1920
internalIdentifier/	Archive reference codes	DigitalAlbini#internalIdentifier/S1_SS1_B1_F1
storageid/	Physical storage identifiers	DigitalAlbini#storageid/box12
inst/	Physical instantiations	DigitalAlbini#inst/S1_SS1_B1_F1_D1
type/	Identifier types	DigitalAlbini#type/InternalIdentifier
temp/	Internal processing (removed in final output)	temp:intermediateSender

---

Output Specifications

File Format

• Format: Turtle (RDF 1.1)
• Encoding: UTF-8
• File Extension: .ttl
• MIME Type: text/turtle

References

Standards and Ontologies

• RiC-O 0.2: https://www.ica.org/standards/RiC/ontology
• RDF 1.1: https://www.w3.org/TR/rdf11-primer/
• Turtle: https://www.w3.org/TR/turtle/
• SKOS: https://www.w3.org/TR/skos-reference/
• Dublin Core: https://www.dublincore.org/specifications/dublin-core/dcmi-terms/

External Services

• GeoNames: https://www.geonames.org/
• VIAF: https://viaf.org/
• geonamescache: https://pypi.org/project/geonamescache/

Related Documentation

• UML Data Model Diagrams: See docs/ folder
  • busta_uml.jpg - Folder-level model
  • serie_uml.jpg - Series-level model
  • sottoserie_uml.jpg - Sub-series level model

---

Last Updated: February 2025
Version: 1.0
Maintainer: Digital Humanities Advanced Research Center (/DH.arc) Institution: Department of Classical Philology and Italian Studies, University of Bologna