Spatrem Compiler

This repository contains models and utilities developed for the Spatrem project. It comprises Python models of LRMoo classes and properties and code that uses those models to process tabular data files into RDF representations and serialize them into files that may be loaded into an RDF graph database like GraphDB or Jena and queried with SPARQL.

A single command-line script, compile_rdf.py, takes a pair of semicolon-separated tabular data files, one of translator records and one of translation records, and builds an internal graph of unique nodes representing authors, translators, names, magazines, languages, issues, genres, originals, and translations, which are then exported to files. The files are in the Turtle format, which is easier to read than other RDF serializations.

Notes

The compiler processes the translations file first. It proceeds record by record, cleaning the data where necessary and tracking the creation of new nodes to prevent duplication. Special attention is given to the "Anon." translator: a new node is created for each appearance of this token, even though the same name (Nomen) is attached to it. When all the translations have been processed, the translator file is processed, and the demographic data captured in it are added to the translator graphs.

The graph model (or schema, or ontology), is primarily LRMoo, with identifiers coming from the dcterms vocabulary. The following Spatrem-specific classes have been introduced:

• volume
• number
• language_area
• gender
• genre
• nationality
• pubDate
• year_birth
• year_death

Introducing the Spatrem-specific property pubDate makes it possible to eliminate the complex graphs of expressions, which are unneeded by Spatrem's application.

Identifiers

Nodes are generated in several namespaces. UUIDs are used to guarantee uniqueness, at the price of human legibility.

Building and Installing

Requirements:

Python 3.11 Poetry >= 1.6

To install, clone this repository, then run Poetry install:

git clone git@github.com:cwulfman/spatrem_parser.git spatrem_parser
cd spatrem_parser
poetry install

Running

There is a command-line script building graphs from translation and translator spreadsheets. Command-line help is available:

poetry run python compile_rdf.py --help

An Example: the German Dataset

For example, to process the German translations file into RDF, type the following at the command line, replacing the filepaths as appropriate:

poetry run python compile_rdf.py ~/spatrem/Datasets/translators/DE_Translators.csv \
~/spatrem/Datasets/translations/DE_Translations.csv \
~/spatrem/knowledge_graphs/rdf/DE