Future directions

This document provides an overview of the United Nations Document Ontology (UNDO), i.e. an OWL 2 DL ontology that aims to provide a framework for the description of all the entities mentioned in United Nations documents, and the relations that can exist among them. The idea behind the development of this model was to provide a common framework to be used and, eventually, extended by the various agencies of the United Nations for sharing data about documents and their content in RDF format in an interchangeable way.

There are several actions that need to be taken in the future in the context of UNDO and related entities, so as to guarantee its broad usage within the United Nations system of organizations, as presented below.

Alignment with foundational ontologies

UNDO is already aligned to two important foundational ontologies, i.e. DOLCE and BFO. This alignment is totally inherited from ALLOT, which defines it in the first place. However, several terms introduced in UNDO could be aligned in a more granular way to the two aforementioned ontologies, and eventually even to other relevant ones – such as the Information Artifact Ontology (IAO).

Even if it could seem straightforward, these kinds of alignments are often quite complex to address, and needs extensive background knowledge and several discussions with experts to be introduced in an appropriate way.

Akoma Ntoso IRI resolver

The IRIs defined in Akoma Ntoso documents follows a specific naming convention, which does not allow to specify any HTTP URL for defining the individuals of the various Top Level Classes. According to the Akoma Ntoso design principles, the conversion between these Akoma Ntoso IRIs into appropriate Web IRIs should be handled by an appropriate resolver application, which should take care about the conversion.

Having such an application available within the United Nations is an essential step for associating the intended HTTP IRI of the resources available in an RDF-based knowledge base of the United Nations (compliant with UNDO). For instance, please consider the following XML tags in an exemplar Akoma Ntoso document:

<TLCPerson
eId="palmirani" showAs="Monica Palmirani"
href="/akn/ontology/person/it/palmirani" />

<TLCRole
eId="author" showAs="Author of Document"/>
href="/akn/ontology/roles/un/author"

<TLCRole
eId="editor" showAs="Editor of Document"/>
href="/akn/ontology/roles/un/editor"

<TLCOrganization eId="ga" showAs="General Assembly"
href="/akn/ontology/organizations/un/generalAssembly"/>

The IRIs specified in the “href” attribute that are used for identities in consideration could be resolved in an HTTP IRI in the following way:

https://w3id.org/un/data/person/palmirani

https://w3id.org/un/data/role/author

https://w3id.org/un/data/role/editor

https://w3id.org/un/data/organization/generalAssembly

Akoma Ntoso to UNDO converter

Having an Akoma Ntoso IRI resolver does not solve entirely the issue of creating a knowledge base starting from the Akoma Ntoso documents available in the United Nations. In fact, another crucial step to achieve is to study and implement an algorithm (e.g. based on XSLT) that takes Akoma Ntoso documents as input and returns a set of RDF statements compliant with UNDO by interpreting Akoma Ntoso IRIs and metadata sections. For instance, please consider the additional XML tags to the ones introduced in the previous section, shown as follows:

<FRBRWork>
<FRBRuri value="/akn/un/collection/publication/un-ga/2013-11-21/A-C3-68-L77"/>
<FRBRauthor href="#ga" as="#author"/>

</FRBRWork>

<FRBRExpression>
<FRBRuri value="
/akn/un/collection/publication/un-ga/2013-11-21/A-C3-68-L77/eng@"/>
<FRBRauthor href="#ga" as="#editor"/>
<FRBRlanguage language="eng"/>

</FRBRExpression>

A possible representation of the converted data in RDF is introduced in the following excerpt (in Turtle):

@prefix allot: <https://w3id.org/akn/ontology/allot/> .
@prefix concept: <https://w3id.org/un/data/concept/> .
@prefix dcterms: <http://purl.org/dc/terms/> .
@prefix expression: <https://w3id.org/un/data/doc/2013-11-21/A-C3-68-L77/> .
@prefix frbr: <http://purl.org/vocab/frbr/core#> .
@prefix lan: <http://id.loc.gov/vocabulary/iso639-1/> .
@prefix org: <https://w3id.org/un/data/organization/> .
@prefix role: <https://w3id.org/un/data/role/> .
@prefix tvc: <http://www.essepuntato.it/2012/04/tvc/> .
@prefix undo: <https://w3id.org/un/ontology/undo/> .
@prefix work: <https://w3id.org/un/data/doc/2013-11-21/> .

work:A-C3-68-L77 a undo:Document ;
allot:hasRepresentation expression:@eng .

expression:@eng a undo:DocumentVersion ;
dcterms:language lan:en .

org:generalAssembly a allot:Organization ;
tvc:hasValue
concept:ga-as-author ;
concept:ga-as-editor .

concept:ga-as-author a undo:ValueInTimeAndContext ;
tvc:withValue role:author ;
tvc:withinContext work:A-C3-68-L77 .

concept:ga-as-editor a undo:ValueInTimeAndContext ;
tvc:withValue role:editor ;
tvc:withinContext expression:eng@ .

Extending the scope of the ontology

While UNDO includes already several ontological entities for addressing the description of several parts of interest of the United Nations document domain, there are still other aspects that need further elucidations. In particular, future extensions of the ontology could consider the definition of:

organizations, e.g. in terms of industries, institutions, countries;

languages adopted in the documents written by all the agencies;

actions introduced in the United Nations documents;

document structures and their rhetoric function;

arguments introduced in the textual content of documents.