Iustinianus LOD

A Linked Open Data project exploring the Byzantine Emperor Justinian and his cultural legacy across law, art, and history

EXPLORE THE PROJECT

PROJECT

This ontology-based project was developed for the Information Science and Cultural Heritage course at the University of Bologna, taught by Professor Francesca Tomasi and Professor Marilena Daquino.

It focuses on the Byzantine Emperor Justinian I, collecting, analyzing, and semantically interlinking a curated set of cultural heritage items —including mosaics, buildings, historical documents, artifacts, audiovisual resources, and digital media—related to his reign and legacy.

Through Linked Open Data methodologies, the project constructs a dynamic knowledge graph, in which entities, drawn from diverse sources, are interconnected across disciplines and institutions. This not only enriches academic research but also ensures that Justinian’s multifaceted heritage — from the Corpus Iuris Civilis to the Hagia Sophia, from artistic representations to modern cultural reinterpretations — is preserved, accessible, and constantly recontextualized.

LOD are the foundation of the Semantic Web, an evolution of the traditional web that aims to make data understandable to machines as well. The goal is to transform the web into a network of semantically interconnected data.

To achieve this, the metadata provided by the institutions holding the selected items were transformed into semantic triples composed of a subject, a predicate, and an object. This is a key step in our project, and it is where RDF becomes essential. RDF is one of the fundamental technologies of the Semantic Web, used to structure and link data on the web. In addition to RDF, other technologies such as TEI/XML were used in the text analysis section.

STUDY OF THE DOMAIN

IDEA

IT
"Cesare fui e son Iustiniano,
che, per voler del primo amor ch'i' sento,
d'entro le leggi trassi il troppo e 'l vano.
[...]
Tosto che con la Chiesa mossi i piedi,
a Dio per grazia piacque di spirarmi
l’alto lavoro, e tutto ‘n lui mi diedi;

e al mio Belisar commendai l’armi,
cui la destra del ciel fu sì congiunta,
che segno fu ch’i’ dovessi posarmi."
— Dante Alighieri, Paradiso VI
EN
"Caesar I was, and am Justinian,
Who, by the will of primal Love I feel,
Took from the laws the useless and redundant.
[...]
As soon as with the Church I moved my feet,
God in his grace it pleased with this high task
To inspire me, and I gave me wholly to it,

And to my Belisarius I commended
The arms, to which was heaven's right hand so joined
It was a signal that I should repose."
— Dante Alighieri, Paradiso VI

The project stems from a collective reflection on the role of figures and monuments that serve as a junction between different cultural traditions.

Justinian I was chosen because, more than many other late antique rulers, he embodies the tension and encounter between East and West: from the legal reforms of the Corpus Iuris Civilis to grandiose architectural undertakings such as the Hagia Sophia, his political and symbolic action produced material and textual traces that cross borders and languages.

The popularity and resonance of Justinian’s image did not end with his lifetime. From Dante’s celebrated invocation in Paradiso to modern historiography, cinema, music and games, Justinian’s figure has been continuously reinterpreted. Contemporary media continue to reframe his achievements and controversies, showing how a sixth-century emperor still informs debates about law, empire, faith, and the interchange between East and West.

ITEMS

We conducted targeted research to identify ten key items directly connected to Justinian’s reign and later reception. The selection spans media and periods: mosaics and architecture, numismatics, manuscripts and legal codices, visual art, film and music, and contemporary games.

Sources combine institutional catalogues and museum collections (Catalogo Beni Culturali, Louvre, British Museum), digital archives and platforms (Internet Archive, YouTube, IMDb, BoardGameGeek), producing a compact dataset ready for semantic conversion and LOD analysis.

Mosaic from the Basilica di San Vitale in Ravenna

Basilica di San Vitale in Ravenna, a 6th-century Byzantine church, renowned for its mosaics including the iconic portrait of Emperor Justinian, exemplifying the union of political power and sacred art.

Mosaic 6th century
See more

Diptych Barberini

6th century ivory diptych depicting the triumphant Emperor Justinian on horseback.

Ivory Diptych 525-550 A.D.
See more

Procopius: History of the Wars

Loeb Classical Library edition (Greek text & English translation by H. B. Dewing). The book corresponds to Volume I, which contains Books I-II (The Persian War). This volume is also significant because it introduces for the first time the figure of Justinian I.

Book 1961
See more

Codex Florentinus (part of Corpus Iuris Civilis)

Manuscript dating back to the 6th century which reports almost entirely the most relevant part of the Digest, that is, a part of the Corpus Iuris Civilis, the collection of normative and jurisprudential material of Roman law commissioned by the emperor Justinian.

Manuscript 6th century
See more

Coin

Gold Coin representing the bust of Justinian I. The obverse features a bust of Justinian, wearing a helmet and cuirass, facing forward, holding a spear and shield.
The reverse features an angel facing forward, holding a long cross and a cross on a globe (globus cruciger).

Coin 527-565 A.D.
See more

Teodora, imperatrice di Bisanzio

This movie, also known as "Theodora, Slave Empress," tells the story of Teodora, a former courtesan who marries Emperor Justinian I and becomes the Empress of Byzantium. The film depicts her journey from her humble origins to the throne, where she displays great political acumen to save her husband's rule and deeply reform the state.

Movie 1954
See more

Hagia Sophia

Historic architectural marvel in Istanbul, originally built as a Christian church and completed in 537 AD under the Byzantine Emperor Justinian I.

Historical Building 532-537 A.D.
See more

Justinian Epic Symphony

A symphony that recounts the fundamental events that occurred under the empire of Justinian I.

Music 2023
See more

Civilization (Videogame)

Videogame that offers a modern retelling of Justinian's reign, letting players rebuild the Byzantine Empire through lawmaking, diplomacy and large-scale military campaigns.

Videogame 2016
See more

Justinian: Intrigue at the Emperor's Court (Board Game)

Board game in which players bribe influential courtiers to gain the emperor’s favor. Bribes come in four colors, but only three are scored during the game. At the end of each turn, one color is scored, and players earn points based on their followers’ influence at court.

Board Game 2006
See more

KNOWLEDGE ORGANIZATION

Metadata Analysis

The first activity carried out by the team was the identification of the metadata standard adopted by the providers of the selected items, with the goal of ensuring interoperability. In cases where the institution had not explicitly stated the standard in use, the team decided to apply either the standard commonly adopted by similar institutions or the one deemed most suitable and relevant to the nature of the object, according to semantic accuracy and the context of usage of the formal grammar chosen.

Theoretical Model

The Theoretical Model was developed by analysing the natural language descriptions of each cultural heritage item, as provided by the holding institutions. The goal was to capture the core knowledge about Justinian’s legacy in a structured yet human readable form. All statements were expressed as subject–>predicate–>object triples, following the pattern of Linked Open Data. This made it possible to explicitly show not only how each item relates to the central theme, being the multifaceted heritage of Emperor Justinian I, but also the network of connections among the items themselves.

Because not every institution supplied metadata according to a formal standard, we carefully aligned each predicate to the most appropriate vocabulary. Where necessary, predicates were adapted to maintain readability and to ensure consistency with authority files. Additional entities—such as historical figures (Theodora, Belisarius), places (Constantinople, Ravenna), and concepts (Corpus Iuris Civilis, Byzantine architecture) were introduced to enrich the dataset and to reveal further links across the items.

In the visualisation, different colours and shapes serve to distinguish the types of information, and to make it easier to navigate the map. Items are represented with distinct shapes and colors, while entities are shown in rectangular white and green boxes. Entities with a literal and/or numeric nature are shown with oval shaped green boxes. A legend in the diagram explains the visual encoding in detail.

Theoretical model (click on image to enlarge)

Conceptual Model

The Conceptual Model formalises the domain knowledge by defining the classes, properties, and relationships that structure the dataset. Starting from the theoretical model, each natural language triple was elevated to an abstract level and assigned a suitable class or property from established ontologies. A custom base URI was created to uniquely identify the resources of the project: https://github.com/Iustinianus-LOD/Justinian-LOD/.

The selection of ontologies was driven by the nature of each item. For the Codex Florentinus and other bibliographic resources, the team adopted BIBFRAME, the standard for describing library materials. Museum objects—such as the Diptych Barberini, the San Vitale Mosaic, and the gold solidus—were modelled using CIDOC‑CRM, which offers the necessary granularity for cultural heritage artefacts. Schema.org and FOAF were employed for people, places, and organisations, ensuring interoperability with general purpose web applications. Nomisma.org ontologies provided the required specificity for the numismatic item (coin)

All classes and properties were selected on a case by case basis to guarantee semantic precision while preserving coherence across heterogeneous data sources. The resulting model thus reflects a careful balance between the richness of the original descriptions and the formal requirements of the Semantic Web.

Conceptual model diagram

Conceptual model (click on image to enlarge)

The full list of namespace prefixes used in our ontology includes:

  • @prefix just: <https://iustinianus-lod.github.io/Justinian-LOD//>
  • @prefix aat: <http://vocab.getty.edu/aat/>
  • @prefix bf: <http://id.loc.gov/ontologies/bibframe/>
  • @prefix bibo: <http://purl.org/ontology/bibo/>
  • @prefix crm: <http://www.cidoc-crm.org/cidoc-crm/>
  • @prefix dbo: <http://dbpedia.org/ontology/>
  • @prefix dcterms: <http://purl.org/dc/terms/>
  • @prefix nm: <http://nomisma.org/id/>
  • @prefix nmo: <http://nomisma.org/ontology#>
  • @prefix owl: <http://www.w3.org/2002/07/owl#>
  • @prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
  • @prefix schema: <https://schema.org/>
  • @prefix skos: <http://www.w3.org/2004/02/skos/core#>
  • @prefix wd: <https://www.wikidata.org/wiki/>

KNOWLEDGE REPRESENTATION

CSV FILES

Below is a collection of .csv files (one table per item) with the full natural-language description of our selected cultural heritage items.

The predicates were aligned to each source whenever possible; when not available, the mapping was completed by the team to balance precision and consistency for a later formal mapping.

Diptych
Mosaic
Codex
Coin
Movie
Building
Book
Music
Game
Board Game
Additional

Diptych Barberini

Subject Predicate Object
Diptych Barberini taken from https://collections.louvre.fr/ark:/53355/cl010114082
Diptych Barberini has title Feuillet de diptyque en cinq parties : l'Empereur triomphant (Justinien ?)
Diptych Barberini has material Ivory
Diptych Barberini created in 525 - 550 A.D.
Diptych Barberini has type Diptych
Diptych Barberini landed by Musée du Louvre, Département des Antiquités grecques, étrusques et romaines
Diptych Barberini has height 34,2 cm
Diptych Barberini has subject Justinian I
Diptych Barberini originating from Constantinople

San Vitale Mosaic

Subject Predicate Object
San Vitale Mosaic taken from https://catalogo.beniculturali.it/detail/PhotographicHeritage/0800649118
San Vitale Mosaic has title Ravenna - San Vitale - Mosaico con Giustiniano e il suo seguito
San Vitale Mosaic created in 6th century
San Vitale Mosaic has time period Byzantine Empire
San Vitale Mosaic has type Mosaic
San Vitale Mosaic has material glass paste
San Vitale Mosaic has subject Justinian I and his entourage
San Vitale Mosaic portrays Justinian I
San Vitale Mosaic portrays Maximianus of Ravenna
San Vitale Mosaic portrays Belisarius
San Vitale Mosaic located in Basilica of San Vitale, Ravenna

Codex Florentinus

Subject Predicate Object
Codex Florentinus Taken from https://tecabml.contentdm.oclc.org/digital/collection/pandette/id/4/rec/1
Codex Florentinus is part of Corpus Iuris Civilis
Codex Florentinus held by Biblioteca Medicea Laurenziana, Florence
Codex Florentinus has type Manuscript (Pandects)
Codex Florentinus has language Latin
Codex Florentinus has date 6th century CE
Codex Florentinus has origin Constantinople
Codex Florentinus has material Parchment
Codex Florentinus has script Greek uncial
Codex Florentinus describes Roman law

Coin

Subject Predicate Object
Coin taken from https://www.britishmuseum.org/collection/object/C_1852-0630-11
Coin has title Gold solidus of Justinian I
Coin has material gold
Coin has type coin
Coin has authority Justinian
Coin minted in Constantinople
Coin has date 527-565
Coin has die-axis 6 o'clock
Coin has diameter 21 millimetres (max)
Coin has weight 4.354 grammes
Coin is located in The British Museum
Coin has museum ID 1867,0101.1015

Theodora, slave empress

Subject Predicate Object
Theodora, slave empress taken from https://www.imdb.com/title/tt0047567/
Theodora, slave empress has original title Teodora, l'imperatrice di Bisanzio
Theodora, slave empress has release year 1954
Theodora, slave empress has type Film
Theodora, slave empress directed by Riccardo Freda
Theodora, slave empress produced in Italy
Theodora, slave empress has language Italian
Theodora, slave empress depicts Theodora
Theodora, slave empress has genre Historical drama
Theodora, slave empress has actor Gianna Maria Canale

Hagia Sophia

Subject Predicate Object
Hagia Sophia taken from https://www.archnet.org/sites/1989
Hagia Sophia has type Religious Architecture
Hagia Sophia also known as Ayasofya-i Kebir Cami-i Şerifi
Hagia Sophia has construction period 532–537 CE
Hagia Sophia commissioned by Justinian I
Hagia Sophia designed by Anthemius of Tralles; Isidore of Miletus
Hagia Sophia has architectural style Byzantine
Hagia Sophia original function Eastern Orthodox Cathedral
Hagia Sophia previous function Museum (1935–2020)
Hagia Sophia current function Mosque

History of the wars

Subject Predicate Object
History of the wars taken from https://archive.org/details/procopius0001proc/page/n5/mode/2up
History of the wars has type book
History of the wars has author Procopius
History of the wars has title History of the wars
History of the wars has publisher Cambridge : Harvard Univ. Press
History of the wars has publication date 1914
History of the wars has translator H.B. Bronson
History of the wars has identifier procopius0001proc
History of the wars has OCLC identifier 1256733539
History of the wars has subject Justinian I
History of the wars has subject Byzantine Empire
History of the wars has subject The Persian war
History of the wars has language English
History of the wars has language Greek

Justinian Epic Symphony

Subject Predicate Object
Justinian Epic Symphony Taken from https://www.youtube.com/watch?v=E6zzpya2FT4
Justinian Epic Symphony has title Justinian Epic Symphony
Justinian Epic Symphony has subject Justinian I
Justinian Epic Symphony has composer Farya Faraji
Justinian Epic Symphony has release date 2023
Justinian Epic Symphony has format Audio recording; Video upload
Justinian Epic Symphony has duration Approximately 1 hour 12 minutes
Justinian Epic Symphony has language Latin
Justinian Epic Symphony has language Greek
Justinian Epic Symphony has instrumentation note Period and court-style instruments (circa year 1000)
Justinian Epic Symphony distributed on YouTube; Apple Music; SoundCloud
Justinian Epic Symphony has YouTube ID E6zzpya2FT4

Civilization

Subject Predicate Object
Civilization taken from https://civilization.fandom.com/wiki/Justinian_I
Civilization has genre 4X Turn-Based Strategy
Civilization first released in 1991
Civilization created by Sid Meier
Civilization developed by Firaxis Games
Civilization published by 2K Games
Civilization latest main installment Civilization VII
Civilization has defining feature Progression through Ages (Antiquity, Exploration, Modern)
Civilization has victory condition Economic
Civilization has features leader Justinian I

Justinian: Intrigue at the Emperor’s Court

Subject Predicate Object
Justinian: Intrigue at the Emperor’s Court taken from https://boardgamegeek.com/boardgame/22938/justinian-intrigue-at-the-emperors-court
Justinian: Intrigue at the Emperor’s Court has title Justinian: Intrigue at the Emperor’s Court
Justinian: Intrigue at the Emperor's Court has subject Justinian I
Justinian: Intrigue at the Emperor’s Court has designers Leo Colovini; Alessandro Saragosa
Justinian: Intrigue at the Emperor’s Court has publisher Phalanx Games; Mayfair Games
Justinian: Intrigue at the Emperor’s Court has year published 2006
Justinian: Intrigue at the Emperor’s Court has artist Harald Lieske
Justinian: Intrigue at the Emperor’s Court has player count 2–4 players
Justinian: Intrigue at the Emperor’s Court has playing time approximately 45 minutes
Justinian: Intrigue at the Emperor’s Court has mechanics Bluffing; Influence; Betting
Justinian: Intrigue at the Emperor’s Court has components Game board; Tokens; Cards; Box

Additional CSV

Subject Predicate Object
Belisarius is part of Justinian Epic Simphony
Belisarius is part of Justinian: Intrigue at the Emperor's court
Belisarius is part of History of the Wars
Belisarius is magister militum of Justinian I
Theodora is wife of Justinian I
Theodora is part of History of the Wars
Theodora is part of Justinian: Intrigue at the Emperor's court
Louvre Museum has type museum
British Museum has type museum
Louvre Museum is located in France
Codex Florentinus also known as Codex Pisanus
Hagia Sophia has label Religious Architecture
Basilica di San Vitale has label Religious Architecture
Corpus Iuris Civilis has authority Justinian I
Greek uncial is a writing script
Byzantine Empire has capital Constantinople

RDF PRODUCTION

CSV Files in Formal Language

From these natural language CSV files, another set of CSV files is created using formal language. These files will be found under the folder “csv_files/Formal_language” within the project’s repository. The files will have the same structure as the natural language files, following the subject-predicate-object form, so that every statement could be traced back to its equivalent form in natural language and mapped into a semantic triple.

The formalization process consisted of two parallel operations:

  • For the predicates, every relationship identified in the natural language description was associated with a property from a predefined ontology or vocabulary. Dublin Core (dcterms) was used for generic descriptive metadata such as title, date, language, and source. BIBFRAME was used for bibliographic resources, CIDOC-CRM for cultural heritage objects and events, Schema.org for media items and generic descriptions, and the Nomisma Ontology (nmo) for the coin.
  • For the objects, every value was associated with a URI from an authority-controlled resource rather than being represented as a plain string. VIAF was used for people and authors, Wikidata for historical concepts and entities, GeoNames for geographical locations, and Getty AAT for object types and materials.

Entities that are central to the project but not held by any particular institution — such as Justinian I, Theodora, Belisarius, the Byzantine Empire, and the Corpus Iuris Civilis — were assigned custom URIs under the project's base namespace (just:).

Two additional CSV files were also created to describe relationships between entities that are not items themselves but are relevant to the dataset. These include relationships between historical figures and the items in which they appear, descriptions of the museums that hold the items, and the relationship between the Byzantine Empire and its capital Constantinople. owl:sameAs statements were also added to align the project's custom URIs with their equivalent Wikidata entries, ensuring interoperability with the broader Linked Open Data ecosystem.

Python Script for Transformation

The RDF dataset is produced programmatically by a Python script using the rdflib, csv and glob libraries. The script initializes an RDF graph, declares namespace prefixes, and processes every CSV file found in the csv_files/Formal_language folder. Files are parsed with csv.DictReader, which reads the header row automatically and yields subject–predicate–object rows.

For each row the script expands CURIEs to full URIs using the prefix map, accepts full http/https URIs directly, and normalizes casing when needed. Objects are handled according to their content:

  • Values wrapped in angle brackets (<>) are treated as URI references.
  • CURIEs are expanded using the prefix map.
  • Language-tagged values (e.g. @en, @la) become language-tagged literals.
  • Four-digit values within the historical year range are mapped to xsd:gYear.
  • Other numeric values are mapped to xsd:integer.
  • Remaining values are stored as plain string literals.

If a subject or predicate cannot be resolved to a known CURIE or URI, the row is skipped and reported on the console to make source correction easy. After processing all files the graph is serialized to Turtle as full_dataset.ttl, chosen for its readability and broad tooling support.

What the script does specifically:

  • Initialises a graph and binds namespace prefixes.
  • Reads all .csv files in csv_files/Formal_language using DictReader.
  • Expands subjects/predicates from CURIEs or accepts full URIs; parses objects into URIs or typed literals.
  • Adds triples to the graph and writes the resulting Turtle file (full_dataset.ttl).

RDF Dataset

Our final output:

RDF Visualization

RDF Visualization diagram

RDF Visualization (click to enlarge)

TEXT ANALYSIS

XML/TEI document

Our coding work focused on selected excerpts from Procopius's History of the Wars. Specifically, we coded the Introduction and Chapters XIII-XIV of Book I in Loeb's English translation as represented in the Internet Archive copy.

The excerpts were chosen because they explicitly situate the initial phase of Justinian's reign in the Persian (Mesopotamian) theater of operations and introduce key actors and events used in our LOD case study. Chapter XIII, in particular, introduces Justinian in the context of the war against the Persians and describes the appointment of Belisarius and the preparations for the battle. Chapter XIV continues the narrative of the campaign and contains further descriptions of troop deployments, commanders, and locations.

To validate the XML/TEI document, we used the XML/TEI plugin for VS Code.

Main sections of the XML/TEI document:

<teiHeader>

Contains all metadata relating to the work and its digital edition. It is divided into several subsections:

  • <fileDesc> — Describes the electronic resource, including the title of the digital edition, the author, the digital editor, and the publication and licensing information. It also provides a bibliographic description of the original source used for the encoding.

  • <encodingDesc> — Explains the editorial principles and encoding choices. For example, pb/ has been used to preserve the page breaks of the original edition.

  • <profileDesc> — Details non-bibliographic features of the text, such as language, text type and thematic classification.

    It also includes structured metadata such as lists of people and places relevant to the encoded text. In the body of the text, personal and place names are marked with @ref, which links each occurrence to the corresponding xml:id in the teiHeader. This structure makes it possible to connect the encoded entities to external authority identifiers, such as Wikidata.

    Furthermore, keywords considered meaningful for the project are included to highlight the central themes of the encoded text.

<text>

Contains the actual text of the work.

  • <front> — Includes the preliminary parts, such as the title page and part of the preface.

  • <body> — Contains the selected excerpts from the main body of the text. Proper names are enclosed in persName and place names in placeName.

From XML to HTML

We have produced a stylesheet with XSL extension, a file that contains the rules for identifying XML nodes in the source document (using XPath), specifying how to manipulate them, and saving them in a new HTML file.

In particular, XSLT processes the XML document by following its hierarchical tree structure, starting from the root (/tei:TEI) and recursively applying templates to all child elements. The XSL file defines templates that specify the rules for transforming specific XML elements into corresponding HTML elements.

The XSLT transformation process was structured to extract and reorganize the metadata from the TEI header of the XML document, converting them into dedicated and well-formatted HTML sections. All descriptive information has been extracted and presented: metadata from both the original 1865 edition and the current digital edition, as well as a specific section dedicated to displaying the list of cited people and places mentioned, linked to external resources such as Wikidata and GeoNames, along with the keywords highlighted in the text, all organized for clear and immediate consultation.

Regarding the processing of the actual text, specific templates were created to handle all structural elements (such as divisions into chapters and sections) and textual elements (e.g. paragraphs), preserving the original semantic attributes.

The adopted approach maintains a clear separation between content and presentation: the XSL stylesheet deals exclusively with structure and semantic transformation, while all visual formatting rules are delegated to embedded CSS, thus ensuring a well-structured and aesthetically coherent HTML output.

Once the XSL file was produced, both it and the XML file were uploaded to an online tool called Free Formatter to generate the corresponding HTML file, which was then converted into a browsable HTML page.

XML/TEI to RDF

In order to extend the Linked Open Data model of the project to include the encoded textual source, a Python script was developed to convert the XML/TEI document of the Procopius encoding into an RDF dataset.

The conversion process was designed to leverage the structured metadata already encoded in the TEI document. Entities recorded in the <teiHeader> — such as historical figures listed in the <listPerson> element and geographical locations in the <listPlace> element — were extracted and assigned URIs under the project's base namespace. Where the TEI encoding already contained references to external authority resources such as Wikidata or VIAF, these were reused directly as owl:sameAs statements in the resulting RDF dataset, ensuring alignment with the broader Linked Open Data ecosystem.

Descriptive properties were represented using the same ontologies employed throughout the project: dcterms for bibliographic metadata and foaf for describing individuals. The resulting graph was then serialized in Turtle (.ttl) format, producing an RDF file that complements the main dataset and can be queried and browsed alongside it. Validation was performed using the W3C RDF Validator to confirm the syntactic correctness of the file before publication.

Team

Information Science and Cultural Heritage Students

Maria Concetta De Matteis

Maria Concetta
De Matteis

Master's degree student of DHDK at the University of Bologna

Master of Archaeology

Rumana Mehboob

Rumana Mehboob

Master's degree student of DHDK at the University of Bologna

Bachelor of Humanities

Kourosh Shahbazi

Kourosh "Cyrus" Shahbazi

Master's degree student of DHDK at the University of Bologna

Bachelor of Teaching English as a Foreign Language

Information Science and Cultural Heritage - University of Bologna