KNOWLEDGE REPRESENTATION

Starting from CSV files that describe our ontology items in natural language listing the triples defined in the theoretical model, we mapped predicates and entities to known RDF namespaces and saved these mappings in JSON files. All entities have been linked to their corresponding Wikidata URIs to take advantage of Wikidata as an authority hub that connects to multiple external authority files (e.g. Viaf, Geonames).

Using Python, we then processed each CSV file, converting rows into RDF triples based on the mappings. Each resulting RDF graph was serialized into a Turtle (.ttl) file, preserving a one-to-one correspondence between input CSVs and output RDF files. Moreover, we produced a comprehensive Turtle file for the entire ontology, including both the individual files for each item and our custom-defined entities, organized into the following sections: person, type, format and organization.

From CSV...

subjectpredicateobject
Nana (anime)has titleNana
Nana (anime)has typeanime
Nana (anime)is adaptation ofNana (manga)
Nana (anime)has date of publication2007
Nana (anime)has publisherTerminal Video
Nana (anime)has audio languageItalian
Nana (anime)has audio languageJapanese
Nana (anime)has subtitle languageItalian
Nana (anime)has subtitle languageJapanese
Nana (anime)has directorMorio Asaka
Nana (anime)is held byBiblioteca Sala Borsa
Nana (anime)has formatslipcase

... to RDF


                        @base <https://github.com/metamuses/lod-save-the-queen/> .
                        @prefix dc: <http://purl.org/dc/elements/1.1/> .
                        @prefix dct: <http://purl.org/dc/terms/> .
                        @prefix schema: <https://schema.org/> .
                        @prefix wd: <https://www.wikidata.org/entity/> .

                        <item/anime> a wd:Q1107 ;
                            dc:language wd:Q5287,
                                wd:Q652 ;
                            dc:title "Nana" ;
                            dct:format <format/slipcase> ;
                            dct:publisher wd:Q132093230 ;
                            schema:datePublished "2007" ;
                            schema:director wd:Q516966 ;
                            schema:isBasedOn <item/manga> ;
                            schema:itemLocation wd:Q3639614 ;
                            schema:musicBy wd:Q2568944 .
                    

ENCODING

100 Days of Active Resistance :
A Digital Edition

The XML file provides the encoding of the book 100 Days of Active Resistance following the TEI guidelines.The book contains what is known as The Active Resistance to Propaganda manifesto, that precedes the photos collected from September 2010 by Vivienne Westwood in collaboration with Lee Jeans for 100 days.

This manifesto is a dramatic text written by Vivienne Westwood, so it has been encoded following the TEI guidelines for performance texts (drama module). For further information please refer to the teiHeader.

In order to make the XML document browsable it has been transformed into a HTML document via XSLT language. The aim was to provide a visualization tool as similar as possibile to the original book presentational aspects.

To facilitate data exchange and semantic interoperability a Python script has been provided for transforming XML into RDF. Starting from the XML/TEI encoding of our source book, we used Python to parse the XML tree and construct an RDF graph. We first described the encoding itself, then the original bibliographic source (linked to our ontology via owl:sameAs), followed by the inferred bibliographic entities mentioned throughout the text. Finally, we identified and described each character who speaks, adding key metadata and linking them to relevant bibliographic references when applicable. The resulting RDF graph was serialized as a Turtle (.ttl) file.

Book Cover