=Paper=
{{Paper
|id=Vol-2991/paper09
|storemode=property
|title=SismaDL: an ontology to represent post-disaster regulation
|pdfUrl=https://ceur-ws.org/Vol-2991/paper09.pdf
|volume=Vol-2991
|authors=Francesca Caroccia,Damiano D'Agostino,Giordano d'Aloisio,Antinisca Di Marco,Giovanni Stilo
|dblpUrl=https://dblp.org/rec/conf/bir/CarocciaDdMS21
}}
==SismaDL: an ontology to represent post-disaster regulation==
https://ceur-ws.org/Vol-2991/paper09.pdf
SismaDL: an ontology to represent post-disaster
regulation ?
Francesca Caroccia2 , Damiano D’Agostino1 , Giordano d’Aloisio1 ,
Antinisca Di Marco1?? , and Giovanni Stilo1
1
DISIM Department, University of L’Aquila, Italy
damiano.dagostino777@gmail.com, giordano.daloisio@student.univaq.it
giovanni.stilo@univaq.it, antinisca.dimarco@univaq.it
2
DIIIE Department, University of L’Aquila, Italy
francesca.caroccia@univaq.it
Abstract. The emergency caused by a natural disaster must be tackled
promptly by public institutions. In this situation, Governments enact
specific laws (i.e., decrees) to handle the emergency and the reconstruc-
tion of destroyed areas. As it happened in 2009 and 2016 when the Italian
Government issued several, very different, decrees to face respectively the
earthquakes of L’Aquila and Centro Italia.
In this work, we propose SismaDL, a LKIF based ontology, that mod-
els the laws in the domain of natural disasters. SismaDL has been used
to model the aforementioned laws to build a knowledge base useful to
reason about why one regulation is less effective and efficient than the
other. SismaDL is the first step of a wider project whose aims are: i)
compare laws in the domain of natural disaster; ii) integrate such laws in
the Semantic Web; iii) evaluate the effectiveness of a post-disaster recon-
struction law; iv) identify good practices to build a reference normative
model of the natural disaster regulation. This project is a founding step
towards the development of accurate and timely IT systems for efficient
and high quality disaster management and reconstruction services to
support Governments and local institutions in case of natural disasters.
Keywords: Ontology · Regulation and law · Reasoning · Analysis of
laws · Semantic Web · LKIF
1 Introduction
Natural disasters have a big impact on human beings. The arisen emergency must
be tackled promptly by local and national institutions. Even if the emergency
management has been already planned, at both operative and normative levels,
governments must promptly provide the details needed to manage the specificity
of the event.
?
This work is partially supported by Territori Aperti a project funded by Fondo
Territori Lavoro e Conoscenza CGIL CISL UIL and by SoBigData-PlusPlus H2020-
INFRAIA-2019-1 EU project, contract number 871042.
??
Contact person for SismaDL project.
Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License
Attribution 4.0 International (CC BY 4.0)
99
One of the tools used by Governments could be to enact specific laws to
handle the emergency and the reconstruction of involved places, as in the Italian
Government case, which manages post-disaster and reconstruction by issuing
a series of normative acts, namely law-decrees (“decreti-legge” in Italian). It is
worth noting that the Italian Government can be, ” reasons of necessity and
urgency”, exceptionally authorized to exercise the legislative power through the
executive one. As it happened in 2009 and 2016 when the Italian Government is-
sued several decrees to face respectively the earthquakes of L’Aquila and Centro
Italia ([1], and [2] are respectively the principal ones). Those decrees aim to reg-
ulate: i) repair interventions; ii) reconstruction; iii) assistance to the population;
iv) economic recovery.
If we consider the socio-economic context and the natural disaster type, the
two events are comparable. But analyzing the related reports[3] and data[4], it is
possible to notice that the reconstruction related to the 2016 event proceeds at
a speed different than the one of 2009. As an example, the percentage of private
buildings reconstructed after 42 months from the event (2009 and 2016) are
25% and 3%, respectively3 .
Another unexpected difference, as noticed by Caroccia (co-author and law
professor), was that the two main decrees differ deeply in their structure and
content: the number of articles present is very different and the 2016’s decree
details the functions that each authority must perform.
Even if the reconstruction speed difference can be precisely analyzed quanti-
tatively, it is needed to carry out a semantic analysis to understand the hidden
causes.
The depicted scenario poses the basis to investigate which is the impact of
a law on the effectiveness of its actuation. Furthermore, we recognized a lack of
knowledge and good practice in the specific domain of the legal aspects of nat-
ural disasters. Considering the highlighted necessities, we drafted a long term
project which wants to provide support for the following activities: i) to com-
pare laws in the natural disaster domain; ii) to integrate the domain of natural
disaster-related laws in the Semantic Web; iii) to evaluate the effectiveness of a
post-disaster reconstruction law; iv) to identify good practices to produce effec-
tive reconstructions’ laws; v) to build a normative conceptual model for natural
disaster emergencies.
This long term project will provide governments and local institutions of
accurate and timely information supply for efficient and high quality disaster
management and reconstruction processes and services that are customized on
the specific context represented by the damaged area together with its social, eco-
nomical and environmental infrastructure. Moreover, using technology to map
and understand disaster-related laws can be useful to help law-scholars to un-
derstand inferences, to identify conflicting statutes and to infer hidden causes of
such conflicts.
3
The percentages are obtained by considering the number of reconstructed buildings
on the total number of buildings, normalized to involved area.
100
As a grounded step, in this paper, we present an ontology (namely SismaDL)
based on LKIF [8], a dedicated legal ontology. We extend LKIF to model the
semantic content of the decrees. To define SismaDL ontology, we employed a
top-down approach where we specialized the emergency concepts, specific of the
considered domain, on the LKIF abstract ones.
We chose to adopt the ontology formalism since it offers: i) to structure
the laws providing a clear representation of them to promote the dissemination
of knowledge according to the basis of the semantic web; ii) to capture and
highlight the differences of existing decree-laws, by using queries and inference
tools.
The ontology can be used to map and clarify disaster-related laws. The on-
tological representation of a decree should be done with the support of legal
experts and should be published with the decree itself.
The paper proceeds as follows: in Section 2 we report the related work, Sec-
tion 3 provides details about the SismaDL ontology. Section 4 describes the
analysis that has been made on the modeled decrees. Finally, in Section 5, we
discuss final remarks and future works.
2 Related Work
The related work is organised into three sections which discuss the following as-
pects: i) Legal Foundation Ontology, Classification and comparison; ii) Domain-
Specific Regulation; iii) Ontology Specification Process.
Legal Foundation Ontology, Classification and Comparison. In [8], a
legal domain ontology, namely LKIF, is presented. LKIF characterizes the ele-
ments of the legal domain in a very detailed way. Figure 1 describes the modules
and the relations defined in LKIF to model legal concepts. Top and Mere-
ology modules describe fundamentals elements like Abstract and Physical
Entities, Atom, Part, Whole etc. Time and Place modules extend the previ-
ous ones defining space (Place) and time concepts like Temporal Occurrence.
Process, Role and Action modules are useful to describe dynamic processes, like
causal changes, introducing the concepts of Change, Organization, Person,
Process and so on. Legal Role extends the Role module introducing more spe-
cific legal roles and finally the Expression and Norm modules introduce concepts
useful to describe the mental process of an actor while doing an action, and to
describe legal sources, respectively. In this paper we extended LKIF to define
SismaDL ontology. Since our domain is particular, we decided to include only
a few classes from the last two modules as the others may result misleading
in our context. For example, the Expression module starts from the definition
of the Proposition and Propositional Attitude concepts, useful to describe
the purpose of an action performed by an agent. Since in our contest the pur-
pose of the actions (described by the measures in the decrees) is always to deal
with the emergency, in SismaDL we decided to not include Proposition and
Propositional Attitude concepts.
101
Fig. 1: LKIF modules relations
The work presented in [12] and [13] deal with the comparison and classifi-
cation of ontologies in the legal domain. [12] provides an overview of the legal
domain ontologies by making a distinction between i) semantically oriented ap-
proaches that focus on the semantic interpretation of a representation of ele-
ments, ii) epistemically oriented approaches that focus on knowledge in a do-
main, and iii) ontologically oriented approaches that emphasize the entities and
relationships constituting a domain. [12] presents many examples of ontologies
that, as our one, are based on a top-down approach starting from very abstract
concepts and trying to apply them on concrete domains. It also discusses on-
tology’s applications to several domains helping us to determine the purpose
of SismaDL ontology. [13] analyzes legal ontologies by conceptualising the ba-
sic characterizing elements. It illustrates an analysis and comparison between
conceptualizations useful to represent the legal domain (such as McCarty’s lan-
guage [11], and Van Kralingen’s ontology [10]). Moreover, it implicitly exposes
critical issues related to the representation of the concepts expressed in natural
language.
Domain Specific Regulation. In literature, there exist ontologies dealing with
Privacy and Protection Regulation and Emergency Management Regulation. For
what concern Privacy and Protection Regulation, [5,7] deal with the GDPR issue
and the approach to this problem through IT ontologies. [5] presents IPROnto,
an ontology for GDPR, using the Semantic Web approach. The discussion illus-
trates the structure of ontology by going into the details of the description of
102
some key elements and describing some scenarios. It also discusses a correlation
between ontology and its applications. [7] presents the ontology for the process-
ing of personal data and privacy. It aims to support organizations in solving
personal data processing and privacy, providing a knowledge base on ontology-
based data protection. It highlights the interdependence between GDPR and
information security. Then it illustrates a methodology, similar to that applied
for SismaDL, which starts from the study of the legal field, then proceeds to
study the issued regulation, identifying the requirements and representing the
main concepts. For what concern the Emergency Management Regulation, two
relevant papers [6,9] are discussed. In [6], preliminary work is done to create an
ontology for emergency management. This paper assumes that the realization of
an ontology for emergencies includes the representation of any information useful
for representing an emergency, and should be based on an information ontology
capable of acquiring different types of data from different types of sources. It also
highlights the problems related to the cataloguing and representation of reality
within the ontology showing the example of the words used in emergency cases
and illustrating the ambiguity of natural language for this type of representa-
tion. It puts the base of any ontology concerning emergencies the need for a
shared basic vocabulary. Finally, [9] offers an overview of the construction of an
ontology for emergencies and uses the AFM methodology for the construction of
an ontology for avian influenza. The AFM methodology presented is described
through these 5 steps: 1. Select one emergency document and divide them into
knowledge pieces; 2. Verify relevant main topics of the knowledge pieces; 3. Ex-
tract relevant concepts from the knowledge pieces; 4. Extract relations among
these concepts; 5. Extract restrictions from these relations. The scope of [9] is
to provide support to decision-makers in case of emergencies occur.
Ontology Specification Methodology. One of the works that have most in-
spired this work is reported in [14] where the Superior Court case of Popov v.
Hayashi is modelled. The aforementioned work provides a cue on a possible rep-
resentation’s methodology. It highlights the motivations and choices that led to
creating an ontology focused on a representation useful for the fruition and dis-
semination of information. The parallelism between that research and SismaDL
concerns in addition to the choice in the use of the Protégé software to have a
support tool in the implementation of ontology.
3 SismaDL Ontology
This section reports on the SismaDL Ontology specification. We recall that the
main elements that must be defined into an ontology are:
– Concepts: provide general information about the objects of the domain de-
scribed by the ontology. They are identified as sets of individuals and are
modelled by Classes which are the key elements characterizing the domain.
– Individuals: are the smallest units of information that describe specific ob-
jects of the real world. They are instances of the classes.
103
– Properties: express links between individuals. They specify whether an indi-
vidual belongs to a class or connects an individual to a type of information.
Starting from the regulatory context definition (Section 3.1), we move to
illustrate the ontology conceptual model (in terms of classes and properties)
(Section 3.2) and finally to show some individuals (Section 3.3). In defining Sis-
maDL ontology, we used a top-down approach, starting from abstract concepts
and applying them to the emergency concrete domain. Moreover, We integrate
LKIF ontology [8] in SismaDL.
3.1 Regulatory Context Definition
The first step to build a legal ontology is the definition of the regulatory context
in which the research shall be carried out.
The Italian regulation has a hierarchical structure of laws. This means that
different legal bases may be situated at a different hierarchical level. This hi-
erarchical background allowing to derive one law from the others. Creating an
ordered chain of priority which allows determining which rules must prevail ac-
cording to the lex superior derogat inferiori (also referred to as kelsenian model).
Considering the strict hierarchy allows solving conflicts and deciding which law
is the most appropriate one to apply among several. Considering such a model,
the Italian Constitution expressly establishes that ”in extraordinary situations
of necessity and emergency” the government could adopt under its own respon-
sibility ”provisional measures having the force of law”, named Decreti-Legge
(law-decrees, art. 77 It. Const.). The law-decrees are normative acts having the
force of law but issued by the executive power. The earthquake (or natural dis-
aster) typically allows the governments to enact law-decrees. Thus, in 2009 and
2016, the Italian Government approved law decrees (earthquake of L’Aquila,
DL 28/04/2009, n. 39; the earthquake of Centro Italia, DL 17/10/2016, n. 189)
containing urgent provisions and interventions in favour of the areas affected
by seismic events. These normative acts were converted with (minimal) changes
into law (respectively, L. 24/06/2009, n. 77 and L. 15/12/2016, n. 229). Thus,
within the hierarchy of Italian legal sources, law-decrees are situated at the same
level as laws, after the Constitution and before secondary sources (regulations
etc.).
SismaDL ontology conceptual model embeds the regulatory context and the
hierarchy of the Italian legal sources.
3.2 SismaDL Conceptual Model
Figure 2 reports the ontology’s logical schema, where ovals represent classes and
arrows represent properties. An exception is made by the arrows labelled with
the prefix rdfs which indicates sub-classes. The blue colour depicts entities and
relations of LKIF (such as Agent class that play a Role class), while in yellow
are drafted the new proposed classes and relationships (such as, Measure class is
described in DecreeLaw class). Many of the entities in our ontology are defined
104
by LKIF [8]. SismaDL extends those entities to fit better the domain of interest
(examples are Task and Measure that specialize Action). Figure 3 highlights
this concept by showing the most important SismaDL entities’ hierarchy, also in
relation to LKIF. As before, the blue rectangles represent LKIF classes, while
yellow rectangles depict SismaDL classes. The leaves are the entities used in the
final ontology. As seen from the picture, all of our custom entities derive from
one or more LKIF classes.
perform
Role
perform
Task
plays
enforceMeasure rdfs:subclassOf
actor_in taskProvidedBy
Agent Action
rdfs:subclassOf
enforceMeasure
Measure described DecreeLaw
in
forTheBenefitOf
Fig. 2: SismaDL logical schema.
Mental Physical
Occurrence Medium Change Agent
Entity Entity
Fixed Infrastructure
Movable Property
Legal Document
Natural Object
Mental Object
Legal Source
Organisation
Occurrence
Temporal
Artifact
Action
Place
Role
Legal IT Source
Commission
Corporation
Conference
Reaction
Measure
Creation
Person
Task
Fig. 3: Entity hierarchy
One of the central classes in SismaDL is the Agent class, which is defined
as the set of individuals that can play a Role and have an actor in relation
105
with the Action class, meaning that they are part of an action, both in per-
forming or receiving it. Figure 4a shows the hierarchy of the Agent class: first
of all, we have decided to keep as in LKIF the distinction between Person
and Organization. An Organization is defined as an individual that has at
least a member which is in relation with an instance of Person. We introduced
two classes under Organization to better modelling the considered domain:
Corporation and, its subclass CommissionConference. The first contains all
the agents identified from the decrees as companies, foundations, associations,
committees and organizations. CommissionConference, instead, contains the in-
dividuals representing committees appointed by the decrees to carry out actions
as part of the measures issued to deal with the emergency (e.g. Special Offices).
Agents take part to actions, as specified by actor in relationship in Figure 2.
Figure 4b shows the specialization of the Action entity. An Action - defined as
a sub-class of the Process class which is in turn a sub-class of the Change class
- is modelled as a change brought about by a single agent playing a specific role.
Different kinds of actions are specified: Reaction and Creation are native of
LKIF while Task and Measure are classes introduced to model the specific do-
main. A Task is defined as an action provided by a decree-law and performed by
an agent which plays a particular role. Instead, a Measure is an action directly
described in a decree-law and enforced by an agent. The semantic difference be-
tween Measure and Task is that while a measure is a particular action, which
can be classified as Administrative, Economic, Infrastructural or Social, directly
described in the decree, a task is a duty given to an agent by a decree in func-
tion of his role. To emphasize this difference, we have introduced two object
properties enforceMeasure and perform which respectively relates an agent or
a role to a measure and a task. To clarify better this distinction, the individ-
ual ConcessioneGratuitaDiGaranzieSuFinanziamentiBancari (in English: Free
Granting Of Guarantees On Bank Loans) is an example of a SocialMeasure i.e.
a specific action described in the Art10Comma1L’Aquila (Article 10 paragraph
1 L’Aquila) and effected by the MinisteroDelloSviluppoEconomico (Ministry of
Economic Development) for the benefit of small and medium-sized enterprises.
Instead, AssegnazioneAlloggi (Housing Assignment) is an example of a Task
i.e. a duty assigned by Art43Comma2Amatrice (Article 43 Paragraph 2 Ama-
trice) to people with the role of Major. Figure 4c shows the hierarchy of the
Role entity. A Role is defined as a specification of default behaviour and ac-
companying expectations of the thing ’playing’ the role. Inside this class and
his sub-classes, we included some individuals defined as Passive Subjects or Ac-
tive Subjects in the decrees. It is worth noting that a Function is defined as a
particular kind of Role, meaning the purpose of some object as used in some
context. As stated in section 3.1, a central point in building a legal ontology is
the definition of the regulatory context. Figure 4d shows the hierarchy of Italian
legal sources. This class has been modeled as a child of the Medium class, which
contains all the individuals that are bearer of expressions. Legal Sources IT is
specified by five subclasses: CommunityLaw, Constitution, Statute BookLaw,
Regulations and OtherMeasure, all disjoint between each other. Within the
106
StatuteBookLaw there are DecreeLaw, DecreeSection (articles, defined as child
of DecreeSection), RegionalLaw and StateLaw sub-classes. This representa-
tion fully reflects the law theory on the sources hierarchy. Sometimes in a
DecreeSection, the issuance of an OtherMeasure is regulated. OtherMeasure is
a LegalSource having the property articleProvidesOther Measure. Finally
figure 4e describes the structure of physical entities and in particular of phys-
ical objects. Among the physical objects we distinguish Artifacts, which are
objects created as a consequence of an action and have a specific Function (i.e.
Role) and Person as a sub-class of Natural Object. FixedInfrastructure and
MovableProperty are instead two classed that we introduced to better charac-
terize objects mentioned in the decrees.
3.3 SismaDL individuals
We use SismaDL to model the two considered decrees-laws and then we insert
the individuals of ontology.
As an example of individual insertion, we describe the individual proget-
tazioneERealizzazioneModuliAbitativiEOpereUrbanizzazione (in English: Design
And Construction Of Housing Modules And Urbanisation Works). This individ-
ual is considered as belonging to the InfrastructuralMeasure class. It is described
by Art2Comma1LAquila (Article 1 paragraph 2 L’Aquila), which shows in the
description of the full text of the related paragraph of the decree-law. It has a
relationship with individuals belonging to the FixedInfrastructure class modu-
liAbitativi (in English: housing modules) and with individuals belonging to the
Plan (a sub-class of Mental Entity) class OpereDiUrbanizzazione (in English:
Urbanization works). It is effected by some actors having a specific Legal Role,
and therefore, it has a relationship with this class, particularly with Commissar-
ioDelegato (in English: Delegate Commissioner). Furthermore, it is connected
with the class of the Person Role since the measure has as beneficiaries the In-
dividuals PersoneFisicheResidenti (in English: Resident Physical Persons) and
DimorantiInAbitazioniInagibili (in English: people in uninhabitable dwellings).
In the future, we will expand the SismaDL ontology and its individuals by con-
sidering other decrees or other legal sources referred from the text of the decrees.
3.4 SismaDL implementation
SismaDL has been encoded with the OWL Web Ontology Language (OWL)
through the Protégé tool4 . OWL is a markup language used to represent knowl-
edge like ontologies and essentially is a declaration of 1) entities, object prop-
erties, individuals and annotations; 2) axioms describing sub-classes, equivalent
classes, equivalent object properties and sub-object properties; 3) class asser-
tions relating a class to an individual; 4) object property assertions relating an
object property to one or more individuals; 5) object properties characteristics
such as specific properties, domain, range and so on.
4
https://protege.stanford.edu/
107
(a) Hierarchy of agents
(b) Action hierarchy
(c) Role hierarchy
(d) Hierarchy of legal sources
(e) Hierarchy of physical entities
Fig. 4: Hierarchies of entities. Orange circles represent definite classes while yel-
low circles are primitive classes.
108
The SismaDL OWL’s version can be downloaded at the following link: https:
//doi.org/10.6084/m9.figshare.13853468.v2
4 Analysis and experimental results
The decree-laws articles included in the ontology have been subject to a prelim-
inary analysis to formalize the differences expressed in the two regulations. The
variations detected concern the regulatory model, the social measures, and the
financing mechanism. To verify the SismaDL expressiveness with respect to our
preliminary analysis, we query the ontology using SPARQL query language5 . In
the following we report on the queries we implement that highlight the most
interesting differences among the two considered decrees: the type of measures
issued, the subject of the functions assigned to the active subjects, and the fi-
nancing mechanism adopted to deal with the financial costs arising from the
reconstruction. For the sake of space, we omit the full results of the queries and
show only the query that extracts social measure for L’Aquila decree-law, since
all the other queries follow the same structure.
Fig. 5: Query to extract SocialMeasure.
The first analysis concerns the used modalities to reconstruct the social fab-
ric and aims to compare all the social measures described in the two consid-
ered decrees. To this purpose, we defined two queries, one for the 2009 decree-
law and the other for the 2016 decree-law. Figure 5 shows the query specifica-
tion that allows extracting SocialMeasures for L’Aquila. The first three lines,
characterized by PREFIX, contain abbreviations useful to avoid reporting the
entire URIs to refer to elements present in RDF and RDFS concerning the
structure of the ontology, and to elements defined within SismaDL. The sec-
ond part of the query refers to the selection. The SELECT keyword indicates
the elements resulting from the query. In this case, there are three variables:
?Measure ?DecreeSection ?TypeofMeasure which respectively represent the
measure, the item (decree section) describing the measure and the type of the
measure. In this way, we can extract measures for social purposes and the clas-
sification regarding the direct or indirect government intervention. The third
5
https://www.w3.org/TR/rdf-sparql-query/
109
part, indicated by the keyword WHERE, is used to express the Query Pattern
that creates a subgraph of the ontology and assigns a meaning to the variables
object of the selection, that is SocialMeasures (both direct and indirect ones) for
the Aquila decree (SismaDL:DL27-06-2009) . The results of the social measures
queries (i.e., the ones for earthquake 2009 and 2016) are organized in columns
according to the SELECT section. As it is observed during the analysis phase,
one of the important differences between the decrees concerns the direct or non-
direct involvement of the State in interventions in favor of the population: as
we have seen from the results of the queries, in the case of the earthquake in
L’Aquila, no direct social measures have been defined. This means that the in-
tervention of the State to promote the welfare of the affected population, the
workers, and in general the resumption of social activity was always exercised
through other types of intervention such as, for example, the granting of funds.
Instead, for what concerns the management of the Centro Italia emergency, the
social impact has been taken much more into account and regulated through
direct and indirect measures and articles entirely dedicated to this purpose. An-
other difference that has been observed is that in the Centro Italia decree, there
are definitions of functions whose purpose is still largely addressed to the so-
cial sphere, while they are completely missing in the L’Aquila decree. As an
example, there are functions for public and cultural works (namely Interven-
tiAlleOperePubblicheEBeniCulturali ) or to assign temporary accommodations
(AssegnazioneAlloggi ). The functions play an important role in the definition of
the rehabilitation strategy of the post-earthquake emergency, therefore the ab-
sence of functions in the decree for L’Aquila offers us further information about
the overall plan of interventions: both decrees provide for flooding interventions,
but then, for the understanding of the management model, it would be necessary
to represent in the ontology every normative act that appeared in the decrees. It
is possible, however, to infer, in the measure for Centro Italia, greater clarity of
the general design implemented and described through measures and functions
represented in the ontology. As a last analysis, we have studied the economic
measures grouped by Law decrees. The query results highlight differences in the
financing mechanism of the funds allocated. While bank credit institutions and
Fintecna S.p.a. appear in the measures for L’Aquila, they are not for the one
for Centro Italia, where all the charges are owned by the State. The mechanism
designed for L’Aquila afforded for loans from banks and credit institutions to
guarantee immediate liquidity, but they needed to be guaranteed in turn by the
Cassa Depositi e Prestiti, hence, the State. In this way, it was possible to find the
funds needed to finance the interventions, and Fintecna S.p.a. was also involved
to support the loan procedures and the identification of the beneficiaries of the
loans. The financing mechanism for Centro Italia was concerned only with the
intervention of the state and the establishment of donations to the solidarity
number 45500. Table 1 summarizes the differences between L’Aquila and Centro
Italia decrees, founded running the SPARQL queries on SismaDL ontology.
Finally, comparing our ontology to the related works described in section 2,
we can position SismaDL among the second and the third category defined by
110
17/10/2016 n.189 Amatrice 27/06/2009 n.77 L’Aquila
Legal Model Defined, Legal Function Flood Model
Social Measure Direct and Indirect Only Indirect
Financing Mechanism Ordinary means FINTECNA
Table 1: Differences in the decrees resulting from the analysis.
[12]. Differently from [12,13], our work does not discuss ontology classification
or comparison criteria, but instead focuses on aspects relevant to the specific
domain and identify key elements like the ontologies presented in [5,7]. In par-
ticular, a parallelism can be found with [6,9], since, as for SismaDL, the cited
approaches starts from the regulations issued to deal with emergencies, as ”...Ac-
cording to the experiences in practice, we found that most of such knowledge and
information is written in emergency documentation dispersedly.” (quoted by [9]).
5 Conclusion and Future Work
This paper presented SismaDL, a novel ontology that allows representing decrees
issued for an emergency caused by natural disasters. To this aim, we analyzed two
law decrees which face the post-disaster reconstruction of earthquakes. SismaDL
has been crafted with a top-down methodology, integrating the abstract legal
concepts of LKIF with a set of more concrete ones. Those specific concepts were
identified through the study of real decrees. We modeled the two decrees using
SismaDL in a way that they can be available for future semantic analysis. As
future work, we plan to analyze further the modeled decrees using reasoning
tools other than SPARQL.
The ontology aim is to provide reasoning support to legal questions. In the
future, we plan to create an answering tool that determines entailment between
different legal laws. The proposed system does not address the range of NLP
challenging issues as polysemy, legal named entity recognition, and implicit in-
formation in the legal text. Nevertheless, the ontology could be a handy tool for
jurists. The research outcome was primarily focused on giving legal interpreters
and lawmakers an instrument to help them understand better the conditions
that make a law more effective.
Until now we did not fully address a range of issues as the legal automated
reasoning and the linking, and the representation to the referenced legal texts.
We plan to address them in future development.
SismaDL ontology, currently, allows reasoning on qualitative aspects, i.e. rea-
soning on the normative differences of two or more decrees. In the future, we
will include the modeling of quantitative aspects to reason on the efficiency and
effectiveness of the decrees in the ontology. The ontology will be populated with
other decrees to favor comparisons between several instances. Moreover, we will
insert the data related to the normative texts referred to within the decrees
studied to reach a higher detail level.
Through the extension of SismaDL, we aim to identify a functional scheme for
legal regulation and specify guidelines for an efficient regulation for emergencies.
111
References
1. Law decree 77/2009 (2009), https://www.camera.it/parlam/leggi/09077l.htm
2. Law decree 189/2016 (2016), https://www.gazzettaufficiale.it/eli/id/2016/10/18/
16G00205/sg
3. Rapporto sulla ricostruzione post sisma 2016 (2016), https://sisma2016.gov.it/
wp-content/uploads/2020/08/Rapporto sulla Ricostruzione Giugno2020 REV 2.
pdf”
4. USRA Open Data (2021), https://bde.comuneaq.usra.it/bdeTrasparente/
openData/openDataSet/
5. Delgado, J., Gallego, I., Llorente, S., Garcı́a, R.: Ipronto: An ontology for digital
rights management. In: 16th Annual Conference on Legal Knowledge and Infor-
mation Systems, JURIX. vol. 106. Citeseer (2003)
6. Galton, A., Worboys, M.F.: An ontology of information for emergency manage-
ment. In: ISCRAM (2011)
7. Geko, M., Tjoa, S.: An ontology capturing the interdependence of the general
data protection regulation (gdpr) and information security. In: Proceedings of the
Central European Cybersecurity Conference 2018. pp. 1–6 (2018)
8. Hoekstra, R., Breuker, J., Di Bello, M., Boer, A., et al.: The LKIF core ontology
of basic legal concepts. LOAIT 321, 43–63 (2007)
9. Kai Yu, Qingquan Wang, Lili Rong: Emergency ontology construction in emer-
gency decision support system. In: 2008 IEEE International Conference on Ser-
vice Operations and Logistics, and Informatics. vol. 1, pp. 801–805 (2008).
https://doi.org/10.1109/SOLI.2008.4686508
10. van Kralingen, R.: A conceptual frame-based ontology for the law. In: Proceedings
of the first international workshop on legal ontologies. pp. 6–17. Citeseer (1997)
11. McCarty, L.T.: A language for legal discourse i. basic features. In: Proceedings
of the 2nd international conference on Artificial intelligence and law. pp. 180–189
(1989)
12. Poli, R., Seibt, J.: Theory and applications of ontology: Philosophical perspectives.
Springer (2010)
13. Visser, P.R., Bench-Capon, T.J.: A comparison of four ontologies for the design of
legal knowledge systems. Artificial Intelligence and Law 6(1), 27–57 (1998)
14. Wyner, A., Hoekstra, R.: A legal case owl ontology with an instantiation of Popov
v. Hayashi. Artificial Intelligence and Law 20(1), 83–107 (2012)
112