The Domain

Subject of the project, intentions and goals.

Figure 1. Charles, Grock... or both?

We chose the clown Grock as the subject of our project.

Grock is a fictional character created and played by Charles Adrien Wettach, a versatile artist and entertainer.

Our idea revolves around two main premises:

  • the representation of Wettach's versatility through the choice of certain items, created by the artist itself, and others, created by some other people;
  • the distinction between the person (Charles Adrien Wettach) and the character (Grock), for two main reasons: first, because Grock can be considered, from a certain point of view, a piece of work created by Wettach; secondliy, since years before his death Wettach decided to hang up the mask of Grock and retire to private life, we can safely assume, from a conceptual and modelling standpoint, that the two did not always coexist with each other.

E/R Model

Explanation of our idea through an Entity-Relationship model.

Figure 2. E/R model showing the relationships between Charles Adrien Wettach, Grock, the ten objects we chose and the metadata we are interested in.

The Items

The items we selected and their respective metadata analyses.

Metadata Analysis
Title Au-revoir Paree
Resource link LINK
Object type Music sheet
Provider Library of Congress
Schema MODS
Metadata Analysis Hover on the image or HERE
Metadata Analysis
Title La mia carriera di clown
Resource link LINK
Object type Book
Provider ICCU
Schema ISBD
MARC21
Metadata Analysis Hover on the image or HERE
Metadata Analysis
Title Clown: readings in theatre practice
Resource link LINK
Object type Book
Provider WorldCat
Schema ISBD
RDA
Metadata Analysis Hover on the image or HERE
Metadata Analysis
Title Etude de Grock - Clown
Resource link LINK
Object type Artwork
Provider Smithsonian American Art Museum
Schema CDWA
Metadata Analysis Hover on the image or HERE
Metadata Analysis
Title A la bastoche
Resource link LINK
Object type Music recording
Provider MusicBrainz
Schema MMD
Metadata Analysis Hover on the image or HERE
Metadata Analysis
Title Au-revoir M. Grock
Resource link LINK
Object type Video
Provider IMDb
Schema SOMA
CM
Metadata Analysis Hover on the image or HERE
Metadata Analysis
Title Milano - Ricordo di Grock
Resource link LINK
Object type Newsreel
Provider Europeana
Schema EDM
Metadata Analysis Hover on the image or HERE
Metadata Analysis
Title Portraitserie Grock
Resource link LINK
Object type Portraitserie
Provider Europeana
Schema EDM
Metadata Analysis Hover on the image or HERE
Metadata Analysis
Title Villa Grock
Resource link LINK
Object type Photograph
Provider Europeana
Schema EDM
Metadata Analysis Hover on the image or HERE
Metadata Analysis
Title Autoritratto di Adrian Wettach Grock
Resource link LINK
Object type Artwork
Provider Cultura Italia
Schema PICO
Metadata Analysis Hover on the image or HERE

Metadata Alignment

Mapping of metadata elements highlighted during the analysis phase.

We defined the alignment between the different used standards for metadata (related to people, place, date and concepts) by asking ourselves a series of questions related to such categories and by selecting the metadata elements we deemed to be more relevant for our model.

The questions we asked ourselves during the process of metadata mapping focused on the roles of people involved with the creation, distribution and conservation of the item. Regarding date and place, we chose not to delve too deep since it is information that our model will focus on only marginally. With regards to concepts, we chose to focus on general ideas related to some object properties, such as the title and the format.

Table 1. Series of questions we asked ourselves in order to identify the metadata mostly interesting to us.
Question
Who is the creator of the item?
Who contributed to the item's creation?
Who is the distributor of the item?
Who is the subject of the item?
Who is physically owning the item?
Who owns the intellectual property rights of the item?
When was the item created?
Where was the item created?
What is the title of the item?
What is the content of the item?
What is the format of the item?
What is the alias of the person?

Theoretical Model

Creation of a theoretical framework, modeled by exploiting FRBR conceptual structure.

We decided to base our theoretical model on the FRBR structure. What we did, for each object, was to conceptually treat it from the point of view of all the FRBR Group 1 entities at once and to insert each metadatum we deemed to be relevant to our task into its respective entity slot (either work, manifestation, expression or item). In this way, we were allowed to give a distinct and clear answer to each question we developed during the alignment phase.

Figure 3. In our theoretical model each metadatum is attached to the FRBR entity we deemed to be most appropriate.

We then applied this theoretical model to three of our objects:

Theoretical Model
Au-revoir Paree
Theoretical Model
Autoritratto
Theoretical Model
La mia carriera di clown

Conceptual Model

Creation of a conceptual model by reusing other existing models.

Starting from the E/R model, the metadata alignment and the theoretical model, the next step was to create an actual conceptual model able to adequately describe the data at our disposal.

With regard to the date format, we chose to use either the DMY format (in case we are dealing with a precisely defined date) or xsd:gYear (in case we are dealing with a vaguely defined date, without a precise day and/or month of that year).

Table 2. Conceptual model for Dates.
DD/MM/YYYY W3 Consortium
xsd:gYear RELAX NG
Table 3. Conceptual model for People.
Question Property Schema
What is the person's name? person:birthName Core Person Vocabulary
What is the person's alias? (optional) poder:alias Poder Vocabulary
Where was the person born? person:placeOfBirth Core Person Vocabulary
When was the person born? core:dateOfBirth Core Concept Ontology
Where did the person die? (optional) person:placeOfDeath Core Person Vocabulary
When did the person die? (optional) core:dateOfDeath Core Concept Ontology
What is the person's job/occupation? gvp:aat2312_perform Getty Vocabulary Program ontology
What is an item related to the person? (optional) gvp:aat2000_related_to Getty Vocabulary Program ontology
Who does the person know? (optional) foaf:knows Friend Of A Friend

With regards to people, we decided to use the Core Person Vocabulary to describe the identity of an individual, enhanced with properties from Friend Of A Friend, Poder Vocabulary, Getty Vocabulary Program ontology and Core Concepts Ontology. As for the last two properties aat2000_related_to and foaf:knows, we are aware of the vagueness of them (especially that of the former, since it is not specified whether the person is the item's creator, producer, distributor, etc...), but we were mostly interested in expressing some kind of relation with at least a person and, possibly, a cultural object of some sort. An interesting future development of this model could consist in implementing a more detailed system of relationships at this level, from the point of view of both objects and people related to the person.

With regards to places, we decided to keep the description to a minimum that would still be useful to us. We used the Location Core Vocabulary to describe a specific place in terms of its name, and the GeoNames ontology to add information about the country in which that place is located.

Table 4. Conceptual model for Places.
Question Property Schema
What is the name of the place? locn:geographicName Location Core Vocabulary
What is the country in which the place is located? gn:countryCode GeoNames
Table 5. Conceptual model for Concepts.
Question Property Schema
What is the occupation? rdfs:label RDFS
How could it be described by exploiting an uncontrolled description? rdfs:comment RDFS
How could it be described by exploiting a controlled vocabulary? dcterms:subject Dublin Core
Who performs this occupation? gvp:aat2311_performed_by Getty Vocabulary Program ontology

With regards to concepts, we decided to focus on people's occupations, which we managed to describe by using a mixture of general properties of RDF Schema and Dublin Core and the Getty Vocabulary Program ontology's "aat2311_performed_by" property (inversely related to "aat2312_perform"). Another interesting development could consist in implementing a more sophisticated and complex model in order to be able to describe concepts, other than occupation, related to people and their doing.

Data Description

Natural language tabulated description of some toy data we chose for each category (People, Places, Concepts).

People

Table 6. Natural language description of Charles Adrien Wettach.
Subject Property Object
[Charles Adrien Wettach]
person:birthName "Charles Adrien Wettach"
poder:alias "Grock"
person:placeOfBirth [Lovelesse]
core:dateOfBirth 10/01/1880
person:placeOfDeath [Imperia]
core:dateOfDeath 14/07/1959
gvp:aat2312_perform [Clown]
gvp:aat2312_perform [Illustrator]
gvp:aat2312_perform [Actor]
gvp:aat2312_perform [Musician]
aat2000_related_to [Autoritratto]
foaf:knows [Suzy Prim]
Table 7. Natural language description of Suzanne Mariette Arduini.
Subject Property Object
[Suzanne Mariette Arduini]
person:birthName "Suzanne Mariette Arduini"
poder:alias "Suzy Prim"
person:placeOfBirth [Paris]
core:dateOfBirth 11/10/1896
person:placeOfDeath [Boulogne-Billancourt]
core:dateOfDeath 07/07/1991
gvp:aat2312_perform [Actor]
gvp:aat2312_perform [Producer]
aat2000_related_to [Au-revoir M. Grock]
foaf:knows [Charles Adrien Wettach]
Table 8. Natural language description of Jon Davison.
Subject Property Object
[Jon Davison]
person:birthName "Jon Davison"
person:placeOfBirth [Bexley]
core:dateOfBirth 1962
gvp:aat2312_perform [Teacher]
gvp:aat2312_perform [Writer]
gvp:aat2312_perform [Clown]
gvp:aat2312_perform [Musician]
aat2000_related_to [Clown: readings in theatre practice]

Places

Table 9. Natural language description of Florence.
Subject Property Object
[Florence]
locn:geographicName "Firenze"@it
gn:countryCode "IT"
Table 10. Natural language description of Paris.
Subject Property Object
[Paris]
locn:geographicName "Paris"@fr
gn:countryCode "FR"
Table 11. Natural language description of Basingstoke.
Subject Property Object
[Basingstoke]
locn:geographicName "Basingstoke"@en
gn:countryCode "GB"

Concepts

Table 12. Natural language description of Actor.
Subject Property Object
[Actor]
rdfs:label "Actor"@en
rdfs:comment "An actor (or actress for female) is one who portrays a character in a performance. The actor performs "in the flesh" in the traditional medium of the theatre, and/or in modern mediums such as film, radio, and television. The analogous Greek term is ὑποκριτής (hupokritḗs), literally "one who answers". The actor's interpretation of their role pertains to the role played, whether based on a real person or fictional character. Interpretation occurs even when the actor is "playing themselves", as in some forms of experimental performance art, or, more commonly; to act, is to create, a character in performance."@en
dcterms:subject [Entertainment_occupations]
gvp:aat2311_performed_by [Charles Adrien Wettach]
gvp:aat2311_performed_by [Suzanne Mariette Arduini]
Table 13. Natural language description of Illustrator.
Subject Property Object
[Illustrator]
rdfs:label "Illustrator"@en
rdfs:comment "An illustrator is an artist who specializes in enhancing writing or elucidating concepts by providing a visual representation that corresponds to the content of the associated text or idea. The illustration may be intended to clarify complicated concepts or objects that are difficult to describe textually, which is the reason illustrations are often found in children's books."@en
dcterms:subject [Arts_occupations]
dcterms:subject [Aesthetics]
gvp:aat2311_performed_by [Charles Adrien Wettach]
Table 14. Natural language description of Clown.
Subject Property Object
[Clown]
rdfs:label "Clown"@en
rdfs:comment "Clowns are comic performers who employ slapstick or similar types of physical comedy, often in a mime style."@en
dcterms:subject [Comedy]
dcterms:subject [Performing_arts]
gvp:aat2311_performed_by [Charles Adrien Wettach]

URI, RDF and Semantic Connections

URI design and RDF description, enriched with semantic connections, for three instances we chose (one for each category).

Table 15. The three data instances we chose for URI and RDF modeling.
Item Person Place Concept
Au-revoir M. Grock Suzanne Mariette Arduini Paris Actor
Autoritratto Charles Adrien Wettach Florence Illustrator
Clown: readings in theatre practice Jon Davison Basingstoke Clown

We created three URIs for three instances (Suzanne Mariette Arduini for People, Florence for Places and Clown for Concepts) and described them by using the RDF Turtle serialization.

Data Instances

RDF Turtle
Suzanne Mariette Arduini
RDF Turtle
Florence
RDF Turtle
Clown

Table 16 shows the semantic relations we added to our RDF description in order to describe some connections our data could have with other resources that are effectively implemented on the Web in Linked Open Data. We used very simple yet meaningful relations (such as similarity, difference, hyponymy, meronymy, hyperonymy).

Table 16. The semantic relations we selected to create meaningful relations between some resources already existing on the Web.
Relation Property Schema
Similarity owl:sameAs OWL
Difference owl:differentFrom OWL
Hyponymy wn:hyponymOf WordNet
Meronymy wn:meronymOf WordNet
Hyperonymy wn:hyperonymOf WordNet