=Paper=
{{Paper
|id=Vol-2228/short8
|storemode=property
|title=Towards a Conceptual Model for Brazilian Popular Music Representation
|pdfUrl=https://ceur-ws.org/Vol-2228/short8.pdf
|volume=Vol-2228
|authors=Marcos Fragomeni Padron,Fernando William Cruz,Juliana Rocha F. Silva
|dblpUrl=https://dblp.org/rec/conf/ontobras/PadronCS18
}}
==Towards a Conceptual Model for Brazilian Popular Music Representation==
<pdf width="1500px">https://ceur-ws.org/Vol-2228/short8.pdf</pdf>
<pre>
                     Towards a conceptual model
              for Brazilian popular music representation
  Marcos Fragomeni Padron1 , Fernando William Cruz1 , Juliana Rocha F. Silva1
                          1
                          Faculdade de Ciência da Informação
                 Universidade de Brası́lia (UnB) – Brası́lia, DF – Brazil
     marcos.fragomeni@gmail.com, fwcruz@unb.br, jurfsilva@gmail.com

    Abstract. Popular music is a valued component of Brazilian national heritage
    because it reflects local traditions and produces records of current social and
    cultural organization. Despite the importance of popular music to Brazilian
    heritage, many sources of information regarding this music have interoperability
    problems and are unable to meet all users’ information needs. Most systems for
    classifying musical information applied to Brazilian popular music are adapta-
    tions that hinder the management of this repertoire. In this article it is discussed
    a specific representation for Brazilian popular music in the form of a concep-
    tual model in order to communicate its nuances and enable the derivation of
    application models for different areas, such as Museology and Librarianship.

1. Introduction
The Brazilian popular music is one of the most listened music styles in Brazil due to
its strong connection with the local culture. Through this music it is possible to meet
the several Brazilian ethnic aspects, either by its lyrics, or by the origins of sounds and
rhythms. Regardless of its importance, there are no formal guidelines for Brazilian music
heritage preservation. Therefore, several information sources of Brazilian music lack
strategies dedicated to serve different information needs and quite often adopt cataloguing
standards that prevent these musical objects to be put in a context of digital curation.
        Since late 1990’s, some multi-entity bibliographic conceptual models emerged.
Those models had a strong influence over musical resources organization. The first one re-
ceived the FRBR acronym [IFLA 2009], which in practice synthesizes a family of concep-
tual models later consolidated on a single model known as IFLA LRM [IFLA 2017]. The
FRBR model inaugurated a new look over the process of intellectual creation, providing
an entity-relationship model whose core entities, WEMI (Work, Expression, Manifes-
tation and Item), separate intellectual from physical aspects of bibliographic resources.
Other model that have influenced the musical information organization is the FRBRoo
[IFLA 2015], that builds on CIDOC-CRM [CSIG 2011] to offer a common view of li-
brary and museum information under a temporal, event-oriented perspective. Those bib-
liographic models are agnostic about the kind of catalogued resources and generic in
relation to subject information and the cultural contextualization of works. The Music
Ontology [Raimond et al. 2007], used to publish musical resources in the Semantic Web,
represents another conceptualization for organizing musical information, however it does
not consider the particularities of different cultures and musical traditions.
        This article presents a specific conceptual model for Brazilian popular music rep-
resentation, which aims to communicate its nuances and enable the derivation of appli-
cation models for different areas, such as Museology and Librarianship. The proposed
model expresses the nature of musical works in Brazilian popular music tradition and real
world situations in which those works are conceived, performed and recorded. The rep-
resented conceptualization is built based on the study of the aforementioned conceptual
models and on a previous research [Silva 2017], that proposed guidelines for organizing
Brazilian popular music information in accordance with the characteristics of this type
of music as identified in the literature and in the perceptions of 22 interviewed Brazilian
musicians.
       For better expressivity, the model is represented in the well-founded OntoUML
language [Guizzardi and Wagner 2010] whose metamodel is supported by the Unified
Foundational Ontology (UFO) categories. It was used the 2.0 version of OntoUML as
introduced in [Guizzardi et al. 2018] which is founded on the UFO’s reviewed theory of
endurant types and taxonomic structures.

2. The conceptual model
The conceptual model is presented in the form of OntoUML diagrams followed by tex-
tual explanations of the represented conceptualization. In order to reduce the number of
elements and ease visualization, mediation and formal relation stereotypes are not shown.
Not stereotyped relations linked to relatorKind classes and its specializations are media-
tions, while the remaining not stereotyped relations are formal.


                      Figure 1. Diagram focusing on musical work.

        Figure 1 presents a diagram with the arguably most important concepts in the
model, the Work and its specializations. To explain the concept of Work, we borrow
a fragment of F1 Work specification of FRBRoo model: “(the work) comprises distinct
concepts or combinations of concepts identified in artistic and intellectual expressions,
such as poems, stories or musical compositions. Such concepts may appear in the course
of the coherent evolution of an original idea into one or more expressions that are domi-
nated by the original idea” [IFLA 2015].
        A MusicalWork relates to musical expressions, while a TextualWork relates to
textual expressions. Musical works can be seen as “musical ideas”, something like a
musical sound not expressed, as if someone could imagine it or play it internally in her
mind. TextualWork was modeled as a kind and not a category in order to abstract the
taxonomy of textual works, considered not important for this conceptualization.
        It is important to draw a distinction between MusicalComposition, Song and
Arrangement. A musical composition is a musical work which is considered original, or
authorial in users’ point of view. It is mainly identified by a melody, that, even with some
variations, can be used to identify its expressions (e.g., musical sounds, scores, etc.) as
instances of the same “music”, even with no lyrics. Musical compositions and songs are
considered original musical ideas (OriginalMusicalIdea).
        A Song is a musical work which links a MusicalComposition to a TextualWork.
A song is a music with lyrics. “A Garota de Ipanema”1 is a song with music by Antônio
Carlos Jobim and Portuguese lyrics by Vinicius de Moraes. The Lyrics was modeled as
a role of a textual work in a song considering that different sorts of textual works can be
used as lyrics. Usually the lyrics is written for the song, but it is not rare for a poem to
be used as lyrics. One good example is the song “Canteiros” where Fagner composed a
music for a poem by Cecı́lia Meireles. The music in the song was modelled as the role
MusicalCompositionInSong. There is also the case where a song is made by adding a
lyrics to an existing musical composition. In this conceptualization the song “A Garota
de Ipanema” with Portuguese lyrics and “The Girl of Ipanema” with English lyrics by
Norman Gimbel are different songs that share the same music (MusicalComposition).
It is important to notice that a music to be considered a song must privilege the text.
Instrumental music can use the human voice to compose the musical texture on the same
level of other musical instruments, even when emitting words or isolated phrases, and still
not be considered a song.
        An Arrangement is a musical work whose author elaborates one (or more) orig-
inal musical idea, in order to represent it in a given context or in a certain form. That
context or form can be as diverse as an album recording of a band, a public performance,
a representation in form of a musical notation for a specific orchestration, or a film sound-
track. The class BasicMusicalComponents represents the intrinsic musical elements in
musical works (e.g., harmony, melody, rhythm, timbre, texture, etc.). Those elements
are not modeled separately because they usually overlap. CompositionMusicalCompo-
nents consist primarily of the composition’s melody. They carry the minimum set of
elements needed to group all arrangements made for that composition. On the other hand,
the ArrangementMusicalComponents contain all musical components intended for its
execution.
        The arrangement is a central piece in the Brazilian popular music scenario. The
expression of any musical work requires an arrangement, being this expression in the
form of sound or any kind of musical notation. That conceptualization is represented by
the category ArrangementExpression and its specializations MusicalSound and Writ-
tenMusicalExpression. Musical compositions and songs need an arrangement to be ex-
pressed. For example, when someone makes a song and records a video for the Internet
singing and playing a guitar, this person is expressing a song with its lyrics and musical
composition through an arrangement for voice and guitar.
       An arrangement is not always expressed in the form of musical notation (a com-
mon situation in this domain), nor its authorship is always registered or recognized. Many
arrangements are made in the context of a recording session in a studio where the ar-

   1
       Brazilian name for the original song “The Girl From Ipanema”.
ranger and the performers interact informally. An arrangement exists even in the context
of a group of performers playing together without previous preparation. In that case, the
performers themselves are the arrangers.


                       Figure 2. Diagram focusing on arrangement.

        Other concepts related to arrangements are shown in Figure 2. The first arrange-
ment by which a song or a musical composition is made known is an original version.
Other arrangements of the same original musical idea are considered derived versions.
Most arrangements are made for one single original musical idea, however some oth-
ers concatenate more than one original musical idea in the same arrangement (Medley).
These medley versions are always derived. If an arrangement is made for at least one
song, it is considered to be a sung version, otherwise it is an instrumental version.
          The category MediumOfPerformance (MOP) represents any sort of media used
by an individual performer artist to express sound when playing in a performance. Usu-
ally it is a musical instrument or the human voice, but it can be anything that can produce
sound as when Hermeto Pascoal plays his heartbeat or the whistle of a kettle. An ar-
rangement defines an Orchestration, which states the number and types of medium of
performances must be used (e.g., one singer and two guitars). The powerType Mediu-
mOfPerformanceType represents all the possible types of MOP, so the orchestration can
prescribe kinds of MOP instead of MOP instances. An arrangement is made for any piano,
not a specific piano instance. This powerType plays the role IntendedMediumOfPerfor-
manceType in the orchestration. The attribute numberOfMOP indicates the number of
MOPs of that type should be used. The relation between MediumOfPerformanceType
and MOP follows the UML notation proposed by [Carvalho et al. 2016] for representing
the case of powerType partitioning the base type, following their multi-level conceptual
modelling theory (MLT).
        The musical Performance is shown in the diagram on Figure 3. A performance
must be an IndividualPerformance (performed by only one IndividualPerformer) or a
ColectivePerformance (when there are more than one musician playing). A collective
performance is always composed by two or more individual performances. A Band is
formed by two or more BandMember that can change over time. Its principle of identity
does not depend on the variation of its members in time. One Person plays the role
of an IndivdualPerformer when playing one or more MediumOfPerformance in an
IndividualPerformance. The role PerformingBandMember is an individual performer
that performs on behalf of a band.
                   Figure 3. Diagram focusing on musical performance.


        The model represents the reality in such a way it can be more expressive and serve
to a broader spectrum of applications. One example of that expressivity is the use of
Powertypes to differentiate the instances of played instruments from their types. One ap-
plication can be interested in the actual instance of the guitar played by João Gilberto at
the Carnegie Hall in 1962, while, for other application, knowing the kind of the played
instrument is enough. Museum studies are interested in object instances, whereas biblio-
graphical applications ignore this information.
        A performance is held in one Place. It can occur in the context of a MusicalEvent.
The PerformedSound is considered inherent to the performance. That sound can option-
ally be recorded to produce musical tracks (Track). Those tracks can be published in
diverse media and formats, however concepts related to these publications are not cov-
ered by this model.


                   Figure 4. Diagram focusing on authorship relations.


        The diagram in Figure 4 depicts all authorship relations. All of those relations
specialize the relatorKind Authorship. There is an authorship relation for each work
with one or more authors, and each author can produce several works. The role Author
is specialized according to the conceived work type.
3. Final considerations
The presented model has already shown its value by supporting better communication and
analysis of the domain and should evolve to cover concepts such as bibliographical items,
cultural context and musical genres. Other usages for this model may include the ade-
quacy evaluation of other music information organization initiatives to this domain, and
its use as a common conceptualization for semantic interoperability among heterogeneous
data sources.
       This research focus on Brazilian popular music, although several of its concepts
seem to be common to other musical traditions. The adequacy of this model to other
musical contexts will be object of future studies.
         It is expected that this work might serve as a first formal description about Brazil-
ian popular music and support the development of software applications, such as music
digital libraries, in consonance with Brazilian culture and traditions.

References
Carvalho, V. A., Almeida, J. P. A., and Guizzardi, G. (2016). Using a well-founded multi-
  level theory to support the analysis and representation of the powertype pattern in con-
  ceptual modeling. In Proceedings of the 28th International Conference on Advanced
  Information Systems Engineering, pages 309–324. Springer.
CSIG (2011). Definition of the CIDOC Conceptual Reference Model, Version 5.0.4.
  CIDOC CRM Special Interest Group.
Guizzardi, G., Fonseca, C. M., Benevides, A. B., Almeida, J. P. A., Porello, D., and
  Sales, T. P. (2018). Endurant types in ontology-driven conceptual modeling: Towards
  ontouml 2.0. In Proceedings of the 37th International Conference on Conceptual Mod-
  eling (ER 2018), Xi’an, China.
Guizzardi, G. and Wagner, G. (2010). Using the unified foundational ontology (ufo) as a
  foundation for general conceptual modeling languages. In Theory and Applications of
  Ontology: Computer Applications, pages 175–196. Springer.
IFLA (2009). Functional Requirements for Bibliographic Records, Final Report. Inter-
  national Federation of Library Associations and Institutions.
IFLA (2015). Definition of FRBRoo: A Conceptual Model for Bibliographic Informa-
  tion in Object-Oriented Formalism, Version 2.4. International Federation of Library
  Associations and Institutions.
IFLA (2017). IFLA Library Reference Model: A Conceptual Model for Bibliographic
  Information, Final Report. International Federation of Library Associations and Insti-
  tutions.
Raimond, Y., Abdallah, S. A., Sandler, M. B., and Giasson, F. (2007). The music ontol-
  ogy. In Proceedings of the 8th Annual International Symposium on Music Information
  Retrieval.
Silva, J. R. d. F. (2017). Diretrizes para a Organização da Informação Musical Brasileira.
   Doutorado em ciência da informação, Universidade de Brası́lia, Brası́lia.

</pre>