Content uploaded by Efstratios Kontopoulos
Author content
All content in this area was uploaded by Efstratios Kontopoulos on Mar 22, 2018
Content may be subject to copyright.
Kontopoulos et al.
Ontology-based Representation of Crisis Management
WiPe/CoRe Paper – Track Name
Proceedings of the 15th ISCRAM Conference – Rochester, NY, USA May 2018
Kees Boersma and Brian Tomaszewski, eds.
Ontology-based Representation of
Crisis Management Procedures for
Climate Events
Efstratios Kontopoulos
CERTH-ITI, Information Technologies
Institute, Thessaloniki, Greece
skontopo@iti.gr
Panagiotis Mitzias
CERTH-ITI, Information Technologies
Institute, Thessaloniki, Greece
pmitzias@iti.gr
Jürgen Moßgraber
Fraunhofer IOSB, Karlsruhe, Germany
juergen.mossgraber@iosb.fraunhofer.de
Philipp Hertweck
Fraunhofer IOSB, Karlsruhe, Germany
philipp.hertweck@iosb.fraunhofer.de
Hylke van der Schaaf
Fraunhofer IOSB, Karlsruhe, Germany
hylke.vanderschaaf@iosb.fraunhofer.de
Désirée Hilbring
Fraunhofer IOSB, Karlsruhe, Germany
desiree.hilbring@iosb.fraunhofer.de
Francesca Lombardo
Alto Adriatico Water Authority, Italy
francesca.lombardo@adbve.it
Daniele Norbiato
Alto Adriatico Water Authority, Italy
daniele.norbiato@adbve.it
Michele Ferri
Alto Adriatico Water Authority, Italy
michele.ferri@adbve.it
Anastasios Karakostas
CERTH-ITI, Information Technologies
Institute, Thessaloniki, Greece
akarakos@iti.gr
Stefanos Vrochidis
CERTH-ITI, Information Technologies
Institute, Thessaloniki, Greece
stefanos@iti.gr
Ioannis Kompatsiaris
CERTH-ITI, Information Technologies
Institute, Thessaloniki, Greece
ikom@iti.gr
ABSTRACT
One of the most critical challenges faced by authorities during the management of a climate-related crisis is the
overwhelming flow of heterogeneous information coming from humans and deployed sensors (e.g. cameras,
temperature measurements, etc.), which has to be processed in order to filter meaningful items and provide crisis
decision support. Towards addressing this challenge, ontologies can provide a semantically unified
representation of the domain, along with superior capabilities in querying and information retrieval.
Nevertheless, the recently proposed ontologies only cover subsets of the relevant concepts. This paper proposes
a more “all-around” lightweight ontology for climate crisis management, which greatly facilitates decision
support and merges several pertinent aspects: representation of a crisis, climate parameters that may cause
climate crises, sensor analysis, crisis incidents and related impacts, first responder unit allocations. The ontology
could constitute the backbone of the decision support systems for crisis management.
Keywords
Crisis management, ontology, semantic integration, decision support, description logics.
Kontopoulos et al.
Ontology-based Representation of Crisis Management
WiPe/CoRe Paper – Track Name
Proceedings of the 15th ISCRAM Conference – Rochester, NY, USA May 2018
Kees Boersma and Brian Tomaszewski, eds.
INTRODUCTION
The effective management of a climate-related crisis (e.g. flood, earthquake, forest fire, etc.) entails serious
challenges for the authorities, the efficient handling of which is a key aspect for public security. One of the most
critical challenges is the overwhelming flow of incoming information from artificial and human sensors
(Babitski et al., 2011). The former type of sensors includes e.g. video footage from static cameras, water level
and temperature measurements from deployed devices, while the latter type mostly includes social media posts,
a rapidly increasing means for conveying information as a crisis incident unravels (Reuter & Kaufhold, 2017).
All this vastly heterogeneous information has to be processed by the authorities and the numerous organizations
involved in a crisis, in order to filter any meaningful items that could facilitate crisis management.
Towards addressing this challenge, recent trends in Crisis Information Management Systems (CIMS) turn to the
use of ontologies for facilitating decision support during a crisis (Liu et al., 2013). Ontologies serve as the
foundation for providing a semantically unified representation of concepts and relationships that is shareable by
different users and is processable by machines (Grimm et al., 2011). Furthermore, ontologies are often
associated with state-of-the-art logical reasoning services, which provide superior capabilities in querying and
information retrieval, as opposed to standard SQL-based applications (Babitski et al., 2011). Finally, since
nowadays a non-trivial subset of the knowledge and data useful to support a decision is available (in
heterogeneous formats) in the Web, a further advantage of using an ontology-based representation is that it
facilitates the integration of structured knowledge and data available on the Web (Rospocher & Serafini, 2012).
This trait is also very useful with regards to information streams coming from social media.
The need to address the interoperability challenge in crisis management has led to the development of a diverse
variety of relevant ontologies that provide interoperability in specific scenarios. A thorough overview of recent
existing approaches is given in (Liu et al., 2013). However, although crisis management pertains several aspects
(climate conditions, unit assignments, incidents and impacts, etc.), and, despite the variety in modelling
approaches, the drawback with the proposed ontologies is that they cover only specific aspects relevant to their
use case. Consequently, the resulting ontology-based systems have a narrow practical focus and provide only
limited decision support to the authorities.
In this context, this paper proposes a lightweight ontology for climate crisis management, which adopts features
from the most prominent existing models, but is more “all-around” and complete, merging all pertinent aspects
of crisis management: representation of a crisis (along with climate parameters that may cause climate crises),
sensor analysis, crisis incidents and related impacts, first responder unit allocations. The ontology constitutes the
backbone of the decision support system developed in the context of the beAWARE EU-funded project
1
focusing on crisis management of climate events.
The rest of the paper is structured as follows: The next section presents existing prominent ontologies for crisis
management, and discusses the comparative advantages of our proposed ontology. Next, an overview of the
project’s user requirements is given, mapping the latter to ontology functional requirements. The ontology is
presented in full detail in the next section, followed by a respective evaluation. The paper is concluded with
final remarks and directions for future research.
RELATED WORK
The advent of semantic technologies (Hendler, 2009) has led to the widespread adoption of ontology-based
approaches in numerous domains, including crisis management, amongst others. Several relevant ontologies
have been proposed in literature, e.g. SOFERS (Liu et al., 2014), ISyCri (Truptil et al., 2008), and the
approaches by Lauras et al. (2015), Mescherin et al. (2013), and Zavarella et al. (2014). A recent thorough
review of the state of the art in crisis management ontologies is given in (Liu et al., 2013).
Besides the above, two of the most prominent approaches in crisis management and response are MOAC
(Limbu, 2012) and SoKNOS (Babitski et al., 2011). MOAC (Management of a Crisis Vocabulary), is a
lightweight vocabulary that provides terms for linking crisis information from three different sources: (a)
traditional humanitarian agencies, (b) volunteer and technical committees, (c) disaster affected communities.
The vocabulary has been developed based on contributions from various key stakeholders, like the Inter Agency
Standing Committee (IASC)
2
, the Global Shelter Cluster
3
, and the Ushahidi platform
4
, who were also involved
1
http://beaware-project.eu/
2
https://interagencystandingcommittee.org/iasc/
3
https://www.sheltercluster.org/
4
https://www.ushahidi.com/
Kontopoulos et al.
Ontology-based Representation of Crisis Management
WiPe/CoRe Paper – Track Name
Proceedings of the 15th ISCRAM Conference – Rochester, NY, USA May 2018
Kees Boersma and Brian Tomaszewski, eds.
in assessing MOAC’s usability, functionality, and structure.
SoKNOS, on the other hand, is a set of ontologies ensuring that newly created information, as well as integrated
sensor information, is semantically characterized, supporting the goal of a shared and semantically unambiguous
information basis across organizations managing crisis incidents. The central SoKNOS ontology is a core
domain ontology defining the basic vocabulary of the emergency management domain. Additional dedicated
ontologies are used for representing resources and damages, and deployment regulations defining the relations
between resources and damages. Furthermore, for the definition of system components, ontologies of user
interfaces and interactions as well as geo sensors have been developed. Based on the aforementioned ontologies,
additional specialized application ontologies can be defined for each application used in the disaster scenario.
As indicated by the authors, all ontologies in SoKNOS have been developed in close cooperation with domain
experts, such as fire brigade officers.
Finally, another highly relevant approach, albeit rather outdated, is the BACAREX ontology (de la Asunción et
al., 2005), which is part of the SIADEX framework for facilitating the design of plans for fighting forest fires.
More specifically, BACAREX is a heavyweight ontology of planning objects and activities related to the forest
fighting plan in the Andalusian regional government. For every object stored, the ontology records both
operational (e.g. geographic coordinates of the object) and informational metadata (i.e. information that may be
needed by the technical staff during a forest fire incident, e.g. the radio channel of the responder responsible for
a specific forest sector).
Overall, the ontologies reported above share the drawback of covering only a subset of the notions involved in
climate-related crisis management (climate conditions, unit assignments, incidents and impacts). Contrary to
these existing approaches, our proposed ontology consists of modules for representing all aspects pertinent to
crisis management. Nevertheless, and as described in more detail in the next section, our proposed model adopts
concepts from some of the existing ontologies as well, predominantly from MOAC and SoKNOS.
USER REQUIREMENTS AND ONTOLOGY COMPETENCY QUESTIONS
The basis for the creation of a climate-related crisis management ontology are the needs and requirements of the
domain experts. The user requirements were extracted by domain experts in the context of an EU project, which
aims at developing a framework providing various services before, during, and after the occurrence of natural
disasters. A common methodology has been used to define the use cases and requirements of the system,
starting with the identification of the status of available tools through an existing situation analysis, in order to
clarify the current digital landscape concerning emergency service requirements. The requirements of the pilot
cases at hand were studied by identifying and interviewing stakeholders concerned with integrated risk
management (municipalities, regional/local civil protection agencies, etc.), focusing on their needs and the
current gaps both in the situational awareness and command and control aspect of the disaster response.
Table 1. Subset of the user requirements.
UR#
Requirement name
Requirement description
UR_107
Localize video, audio
and images
Provide authorities with the ability to localize videos, audio and
images sent by citizens from their mobile phones.
UR_108
Localize task status
Provide authorities with the ability to localize first responders’ reports
regarding the status of their assigned tasks.
UR_109
Localize tweets
Provide authorities with the ability to localize Twitter messages
concerning a crisis event.
UR_110
Localize calls
Provide authorities with the ability to localize phone calls to an
emergency number concerning a crisis event.
UR_111
Detect elements at risk
from video
Provide authorities with the ability to detect and count elements at
risk (e.g. cars and people) from video and images sent from mobile
phones and social media.
UR_120
Map of rescue teams and
task evaluation
Display to authorities the location in time of first responder teams and
provide the ability to evaluate in real time the execution of the
assigned tasks with a global visualization of the activities performed.
First, a common structure and a related terminology were established; as a consequence, a general emergency
Kontopoulos et al.
Ontology-based Representation of Crisis Management
WiPe/CoRe Paper – Track Name
Proceedings of the 15th ISCRAM Conference – Rochester, NY, USA May 2018
Kees Boersma and Brian Tomaszewski, eds.
situation is subdivided in scenarios, use cases and requirements. In detail, an operational scenario is defined as
the environmental and ecological context of the natural disaster and its impact on the elements at risk and
stakeholder assets. Furthermore, a use case is defined as a conceptual description of intended or expected
utilization of the system to prepare for, respond to, or act upon the occurrence of the scenario or various aspects
therein. This use case is defined and specified from an operational user’s point-of-view. Finally, user
requirements describe expectations, requests, and guidelines for functionalities, capabilities, and features of the
system that would facilitate successful completion of the use cases. In the following, the list of requirements
extracted from all the use case descriptions was clarified and shared among the domain experts (i.e. rescue
teams, water management authorities). Table 1 contains an indicative subset of the user requirements, which the
ontology can currently respond to; the full list of user requirements can be found in beAWARE deliverable D2.1
(Norbiato et al., 2017). These user requirements are catalogued as [UR_xzz], where x is the identifier of the
scenario in which the requirement originated, and zz is the serial number of the requirement.
The user requirements are mapped to the ontology’s Competency Questions (CQs). A competency question is a
natural language sentence that expresses a pattern for a type of question people expect an ontology to answer
(Uschold & Gruninger, 1996). The answerability of CQs hence becomes a functional requirement of the
ontology. Based on the list of user requirements above, the ontology is able to respond to several CQs, such as
providing the location of a specific media item (e.g. a tweet, video, image etc.), or indicate the number and type
of vulnerable objects detected from videos.
THE PROPOSED ONTOLOGY
As already discussed previously, the proposed ontology semantically represents three key aspects of climate
management: (a) climate-related natural disasters and associated climate conditions, (b) analyses of data coming
from human and artificial sensors, (c) unit assignments and mission status. This section delves deeper into the
respective representations.
Ontology Language
The ontology language deployed for developing the proposed ontology is OWL 2 (Web Ontology Language), a
declarative knowledge representation language for formally describing a domain of interest, representing
ontologies with formally defined meaning and semantics (W3C, 2012). OWL 2 is a W3C recommendation
based on the solid mathematical background of Description Logics (Baader, 2003), and, thus, it currently
constitutes the most popular ontology language.
For representing a given domain via OWL 2, one has to come up with a set of core terms, and to agree on their
meaning as well as on their interrelations. The vocabulary (terminology), together with the interrelationships,
constitutes the main context of an OWL 2 document. OWL 2 offers the following modelling building blocks:
● Classes provide an abstraction mechanism for grouping objects with similar characteristics, and denote
the set of objects comprised by a concept. There may be diverse criteria for grouping objects/individuals
and one individual may simultaneously belong to several classes. Classes can also form a hierarchy of
more generic (superclasses) and more specific (subclasses) notions.
● Individuals of an OWL class (also referred to as class instances) are the objects belonging to this class.
● Properties, which are further categorized into: (a) Object properties that describe single individuals, class
memberships, and how classes and individuals can relate to each other based on their instances; (b) Data
properties that describe single individuals by asserting specific data values, either from pre-defined data
types (e.g. string, integer, boolean, etc.) or within a data range expression defined by the user; (c)
Annotation properties that give additional description to the domain being modelled, without having any
effect on the logical aspects of the ontology.
Representing Natural Disasters
The representation of climate-related natural disasters in the proposed ontology is illustrated in Figure 1. Class
“Natural Disaster Type” represents the various types of disasters, e.g. floods, forest fires, storms or earthquakes
etc. Disasters may lead to other disasters (via property “leads to”); for instance, a heat wave may lead to forest
fires, or storms may lead to floods. Each type of disaster is characterized by certain climate parameters,
represented via class “Parameter”; for example, solar radiation and temperature are two parameters that
characterize a heat wave. The scheme for representing environmental and meteorological conditions is based to
some extent on the PESCaDO ontologies (Rospocher & Serafini, 2012), and, more specifically, we adopted a
number of related properties from classes EnvironmentalData and EnvironmentalNode.
Kontopoulos et al.
Ontology-based Representation of Crisis Management
WiPe/CoRe Paper – Track Name
Proceedings of the 15th ISCRAM Conference – Rochester, NY, USA May 2018
Kees Boersma and Brian Tomaszewski, eds.
Figure 1: Representation of climate-related natural disasters in the proposed ontology.
Moreover, the actual manifestation of a natural disaster is represented via class “Natural Disaster”, an instance
of which has specific climate conditions with specific values. The figure displays a sample temperature
measurement, which was recorded during the 2017 UK heatwave
5
(17-22 June). Note that in Figure 1 and in the
following two figures, data properties are omitted for reasons of brevity.
Representing Analyzed Data
Besides the representation of climate-related natural disasters and pertinent notions, the proposed ontology also
encompasses information relevant to the analysis of input data coming from the various sensors of the
framework. This information is fed to the ontology from the various analysis components; the core constructs in
the ontology are illustrated in Figure 2.
Figure 2: Representation of analyzed data in the proposed ontology.
Class “Media Item” represents an item of analyzed data, which is related to some analysis task (via class
“Task”). Media items can be pieces of text, images, videos, or even social media posts, all of them submitted
during the occurrence of a climate crisis. The analysis of the respective items (text analysis, image analysis or
video analysis) produces a “Detection” dataset containing all relevant information (e.g., an object detection task
may produce a dataset of detected incidents, objects, and confidence scores). The figure also demonstrates an
example of a video analysis instance, where a potentially injured person is detected in the flood.
Note that the ontology already contains a complete typology of vulnerable objects (e.g. assets, stakeholders,
infrastructure, buildings etc.), impacts and incidents, as well as various other properties (e.g. severity levels,
confidence scores, detection timestamps etc.), that are not displayed in Figure 2 for reasons of brevity. Part of
this scheme for representing disaster impacts is inspired by MOAC (Limbu, 2012), mainly classes
AffectedPopulation, CollapsedStructure, CompromisedBridge, Deaths, InfrastructureDamage and properties
affectedby and impact. Moreover, for categorizing damages and resources we were based on SoKNOS (Babitski
et al., 2011), and, more specifically on the SoKNOS approach for representing damages and their association to
resources (Babitski et al., 2009).
Representing Unit Assignments
The third component of the proposed ontology is responsible for semantically representing response unit
assignments. The adopted representation is based on the approach proposed by the OASIS project (Couturier &
Wilkinson, 2005), mainly the part for representing mission assignments to units and associating missions to
5
http://www.bbc.com/news/uk-40353118
Kontopoulos et al.
Ontology-based Representation of Crisis Management
WiPe/CoRe Paper – Track Name
Proceedings of the 15th ISCRAM Conference – Rochester, NY, USA May 2018
Kees Boersma and Brian Tomaszewski, eds.
incidents taking place during the climate crisis.
Figure 3 displays the respective concepts in the proposed ontology. First responders (class “Responder”) are
assigned one or more missions (class “Mission”), which in turn relate to incidents that involve participating
entities (class “Vulnerable Object”). A mission is also characterized by start and end time, status and mission
priority; these properties are omitted from the figure but reside in the ontology.
Figure 3: Representation of mission assignments to first responder units in the proposed ontology.
Finally, Figure 3 also displays a specific unit, which has been assigned the rescue mission of the injured person
trapped in the flood (see Figure 2).
ONTOLOGY EVALUATION
This section presents an evaluation of the ontology, as far as quality and structure are concerned.
Evaluating the Consistency and Quality
For evaluating the consistency and overall quality of the ontology we used OOPS! (OntOlogy Pitfall Scanner),
an online tool for detecting the most common pitfalls
6
in ontologies (Poveda-Villalón et al., 2014). After
analyzing the ontology, OOPS! provides an indicator for each pitfall detected, according to their possible
negative consequences, and suggests modifications in order to improve the ontology quality. The system detects
(a) critical pitfalls affecting the ontology’s consistency, which are crucial to be corrected; (b) important pitfalls,
which are not equally critical, but are considered as important to be corrected; (c) minor pitfalls, which do not
cause any critical problems, but correcting them will improve the quality of the ontology. Table 2 presents the
pitfalls detected by OOPS! while evaluating our ontology, along with a brief description of their meaning and
the number of cases for which they were specified.
Table 2. Ontology pitfalls detected by OOPS!.
No
Pitfall description
Results
1
Missing annotations (Minor): Ontology terms lack annotation properties that would
improve the ontology understanding and usability from a user point of view.
76 cases
2
Missing disjointness (Important): The ontology lacks disjoint axioms between
classes or between properties that should be defined as disjoint.
Applies to
whole ontology
3
Inverse relationships not explicitly declared (Minor): This pitfall appears when any
relationship (except for symmetric properties) does not have an inverse relationship
defined within the ontology.
19 cases
4
Symmetric or transitive object properties (Suggestion): The domain and range
axioms are equal for each of the following object properties. Could they be
symmetric or transitive? “leadsTo”, “relatesTo”.
2 cases
Regarding pitfall #1, OOPS! detected 76 cases where annotations and descriptions were missing. To overcome
this pitfall and to improve the ontology’s expressiveness, we assigned human readable annotations to every
defined concept in the ontology, with the adoption of properties rdfs:label and rdfs:comment.
Concerning pitfall #2, the tool warned on the absence of disjoint axioms. Specifying that classes are disjoint
enables a system to validate the ontology more efficiently. We fixed this shortcoming by introducing
disjointness between subclasses of the “Task” and “Vulnerable Object” classes.
6
A catalogue of common pitfalls is given at http://oops.linkeddata.es/catalogue.jsp
Kontopoulos et al.
Ontology-based Representation of Crisis Management
WiPe/CoRe Paper – Track Name
Proceedings of the 15th ISCRAM Conference – Rochester, NY, USA May 2018
Kees Boersma and Brian Tomaszewski, eds.
Pitfall #3 issued a warning on the absence of pairs of inverse properties for all of the object properties in the
ontology (19 cases in total). The pitfall was resolved by introducing the corresponding pairs of inverse
properties into the ontology, in order to improve its completeness.
Pitfall #4 consisted of a suggestion about two specific object properties (see Figure 1 and Figure 2,
respectively). In order to resolve this issue, we made the former property symmetric and the latter transitive.
Evaluating the Structure
For evaluating the structure, we relied on OntoMetrics
7
, an online framework that validates ontologies based on
established metrics. Table 3 presents the results derived from the analysis by OntoMetrics. Base Metrics
comprise of simple metrics, like the counting of classes, axioms, objects etc.; these metrics show the quantity of
ontology elements. Schema metrics, on the other hand, address the design of the ontology; metrics in this
category indicate the richness, width, depth, and inheritance of an ontology schema design.
Starting with the base metrics, the total count of classes and properties indicates that the proposed ontology is
rather a lightweight model, which could be easily adopted by various applications, contrary to heavier
“monolithic” ontologies that pose significant challenges in integration. Furthermore, DL expressivity refers to
the Description Logics variant the ontology belongs to (see also section “Ontology Language”). SI(D) indicates a
simple ontology (universal restrictions, limited existential quantification) with inverse, transitive, and datatype
properties.
Table 3. Ontology metrics produced by the OntoMetrics tool.
Base Metrics
Class count
38
Object property count
37
Data property count
22
SubClassOf axioms count
21
Disjoint classes axioms count
2
Inverse object properties axioms count
18
Transitive object property axioms count
2
Symmetric object property axioms count
1
DL expressivity
SI(D)
Schema
Metrics
Attribute richness
0.578947
Inheritance richness
0.657895
Relationship richness
0.609375
Axiom/class ratio
10.184211
Class/relation ratio
0.59375
Regarding schema metrics, the measurements in the table are adopted from (Gangemi et al., 2005) and (Tartir et
al., 2010). Attribute richness is defined as the average number of attributes per class and can indicate both the
quality of ontology design and the amount of information pertaining to instance data. The more attributes that
are defined the more knowledge the ontology conveys. Inheritance richness is defined as the average number of
subclasses per class and is a good indicator of how well knowledge is grouped into different categories and
subcategories in the ontology. This measure can distinguish a horizontal ontology (where classes have a large
number of direct subclasses) from a vertical ontology (where classes have a small number of direct subclasses).
The respective value in the table indicates that the proposed ontology is somewhere in between; this is
reasonable, since the ontology covers many aspects (breadth) while thoroughly modelling some of them (depth).
Relationship richness refers to the ratio of the number of non-inheritance relationships (i.e. object properties,
equivalent classes, disjoint classes) divided by the total number of inheritance (i.e. subclass relations) and non-
inheritance relationships defined in the ontology. This metric reflects the diversity of the types of relations in the
ontology. Finally, axiom/class ratio and class/relation ratio describe the ratio between axioms-classes and
7
https://ontometrics.informatik.uni-rostock.de
Kontopoulos et al.
Ontology-based Representation of Crisis Management
WiPe/CoRe Paper – Track Name
Proceedings of the 15th ISCRAM Conference – Rochester, NY, USA May 2018
Kees Boersma and Brian Tomaszewski, eds.
classes-relations, respectively, and are indications of the ontology’s transparency.
Compliance with User Requirements
As discussed in section “User Requirements and Ontology Competency Questions”, user requirements are
mapped to CQs that the ontology is expected to answer. Following the methodology proposed in (Zemmouchi-
Ghomari & Ghomari, 2013), we translated the CQs into SPARQL queries (Harris & Prud’hommeaux, 2013) and
evaluated the retrieved results. Table 4 includes an indicative set of CQs, along with their SPARQL translation
and an evaluation of the retrieved result sets.
Table 4. Indicative CQs and SPARQL translation.
Competency Question
SPARQL query
Correct?
What is the location of
each media item?
SELECT ?item ?location WHERE {
?item rdf:type :MediaItem .
?location rdf:type :Location .
?item :hasMediaLocation ?location .
}
Yes
What is the location
and mission status of
each rescue team?
SELECT ?team ?location ?status WHERE {
?team rdf:type :Responder .
?location rdf:type :Location .
?mission rdf:type :Mission .
?team :hasResponderLocation ?location .
?team :isAssignedMission ?mission .
?mission :hasMissionStatus ?status .
}
Yes
Which affected
vulnerable objects were
detected in a specific
video?
SELECT ?object WHERE {
?object rdf:type :VulnerableObject .
?dataset rdf:type :Dataset .
?task rdf:type :Task .
?dataset :detectedParticipant ?object .
?task :producesDataset ?dataset .
?task :relatesToMediaItem :video_1 .
}
Yes
What is the impact and
affected vulnerable
objects of a specific
incident?
SELECT ?incident ?impact ?object WHERE {
?incident rdf:type :Incident .
?impact rdf:type :Impact .
?object rdf:type :VulnerableObject .
?incident :hasIncidentImpact ?impact .
?incident :involvesParticipant ?object .
}
Yes
The set of CQs currently includes 29 queries translated into SPARQL, all of which have been evaluated
positively. Nevertheless, as the project progresses, the ontology will naturally further expand, resulting in
additional CQs being added to the original set of queries.
Publicly Available Version of the Ontology
A publicly available version of the ontology will shortly be released on the project’s website, along with the
respective project deliverable which is due June 2018, containing the ontology documentation and sample
instantiations of the notions discussed.
CONCLUSIONS AND FUTURE WORK
This paper argued that existing ontologies for climate-related crisis management only cover subsets of the
pertinent concepts, and proposed a lightweight ontology for semantically integrating all the relevant notions:
Kontopoulos et al.
Ontology-based Representation of Crisis Management
WiPe/CoRe Paper – Track Name
Proceedings of the 15th ISCRAM Conference – Rochester, NY, USA May 2018
Kees Boersma and Brian Tomaszewski, eds.
representation of a crisis, along with associated climate parameters, sensor analysis, crisis incidents and impacts,
and first responder unit allocations. The paper also presented how the proposed ontology satisfies the
requirements of the end users and discussed on the validation of the ontology. The ontology is already being
deployed in the context of an EU-funded project, and can potentially serve as the underlying knowledge base for
any crisis management system, providing authorities with superior decision support capabilities.
Regarding directions for future work, and besides iterative refinements to the model that will take place as the
project progresses, further research will focus on the reasoning techniques, which will be applied on top of the
ontology, in order to facilitate decision support. Our first aim is to provide mechanisms for generating
automated warnings (including reports) based on the current situation and respective context stored in the
knowledge base. Another imminent step is to have the end-users evaluate the ontology-based decision support
and the recommendations provided by it. This assessment will take place in a few months’ time, when the first
pilot deployments will be evaluated in the field, and our findings will then be publicly released.
ACKNOWLEDGMENTS
This work has received funding by the European Commission under contracts H2020-700475 beAWARE and
H2020-776019 EOPEN.
REFERENCES
de la Asunción, M., Castillo, L., Fdez-Olivares, J., García-Pérez, Ó., González, A., & Palao, F. (2005).
SIADEX: An interactive knowledge-based planner for decision support in forest fire fighting. AI
Communications, 18(4), 257-268.
Baader, F. (Ed.). (2003). The description logic handbook: Theory, implementation and applications. Cambridge
university press.
Babitski, G., Bergweiler, S., Grebner, O., Oberle, D., Paulheim, H., & Probst, F. (2011, May). SoKNOS–using
semantic technologies in disaster management software. In Extended Semantic Web Conference (pp. 183-
197). Springer, Berlin, Heidelberg.
Babitski, G., Probst, F., Hoffmann, J., & Oberle, D. (2009). Ontology Design for Information Integration in
Disaster Management. GI Jahrestagung, 154, 3120-3134.
Couturier, M., & Wilkinson, E. (2005, April). Open advanced system for improved crisis management (OASIS).
In Proceedings of the 2nd International Information Systems for Crisis Response and Management
(ISCRAM) Conference, Brussels, Belgium.
Gangemi, A., Catenacci, C., Ciaramita, M., & Lehmann, J. (2005). Ontology evaluation and validation: an
integrated formal model for the quality diagnostic task. Technical Report. Laboratory for Applied Ontology,
ISTC-CNR, Rome/Trento, Italy.
Grimm S., Abecker A., Völker J., & Studer R. (2011). Ontologies and the Semantic Web. In: Domingue J.,
Fensel D., Hendler J.A. (eds) Handbook of Semantic Web Technologies. Springer, Berlin, Heidelberg
Harris, S., Seaborne, A., & Prud’hommeaux, E. (2013). SPARQL 1.1 query language. W3C
recommendation, 21(10).
Hendler, J. (2009). Web 3.0 Emerging. Computer, 42(1).
Lauras, M., Truptil, S., & Bénaben, F. (2015). Towards a better management of complex emergencies through
crisis management meta‐modelling. Disasters, 39(4), 687-714.
Limbu, M. (2012). Management of a Crisis (MOAC) Vocabulary Specification. Available online:
http://www.observedchange.com/moac/ns/, last accessed: Mar’18.
Liu, S., Brewster, C., & Shaw, D. (2013). Ontologies for crisis management: a review of state of the art in
ontology design and usability. In: Proceedings of the 10th International ISCRAM Conference – Baden-
Baden, Germany, May 2013, pp. 349–359.
Liu, Y., Chen, S., & Wang, Y. (2014). SOFERS: Scenario Ontology for Emergency Response
System. JNW, 9(9), 2529-2535.
Mescherin, S. A., Kirillov, I., & Klimenko, S. (2013, October). Ontology of emergency shared situation
awareness and crisis interoperability. In Cyberworlds (CW), 2013 International Conference on (pp. 159-
162). IEEE.
Kontopoulos et al.
Ontology-based Representation of Crisis Management
WiPe/CoRe Paper – Track Name
Proceedings of the 15th ISCRAM Conference – Rochester, NY, USA May 2018
Kees Boersma and Brian Tomaszewski, eds.
Norbiato, D., Castro, C., Vourvachis, I., Janniche, A., Karppinen, A., Karakostas, A., & Ferri, M (2017). D2.1
Use cases and initial user requirements. beAWARE H2020 project deliverable. Available online:
http://beaware-project.eu/wp-content/uploads/2017/07/D2.1_beAWARE.pdf, last accessed: Mar’18.
Poveda-Villalón, M., Gómez-Pérez, A., & Suárez-Figueroa, M. C. (2014). Oops!(ontology pitfall scanner!): An
on-line tool for ontology evaluation. International Journal on Semantic Web and Information Systems
(IJSWIS), 10(2), 7-34.
Reuter, C., & Kaufhold, M. A. (2017). Fifteen years of social media in emergencies: a retrospective review and
future directions for crisis informatics. Journal of Contingencies and Crisis Management.
Rospocher, M., & Serafini, L. (2012, December). An ontological framework for decision support. In Joint
International Semantic Technology Conference (pp. 239-254). Springer, Berlin, Heidelberg.
Tartir, S., Arpinar, I. B., & Sheth, A. P. (2010). Ontological evaluation and validation. In Theory and
applications of ontology: Computer applications (pp. 115-130). Springer Netherlands.
Truptil, S., Bénaben, F., Couget, P., Lauras, M., Chapurlat, V., & Pingaud, H. (2008). Interoperability of
information systems in crisis management: Crisis modeling and metamodeling. Enterprise Interoperability
III, 583-594.
Uschold, M., & Gruninger, M. (1996). Ontologies: Principles, methods and applications. The knowledge
engineering review, 11(2), 93-136.
W3C OWL Working Group. (2012). OWL 2 Web Ontology Language Document Overview (2nd Edition). W3C
Recommendation, available online: http://www.w3.org/TR/owl2-overview/, last accessed: Dec’17.
Zavarella, V., Tanev, H., Steinberger, R., & Van der Goot, E. (2014). An Ontology-Based Approach to Social
Media Mining for Crisis Management. In SSA-SMILE@ ESWC (pp. 55-66).
Zemmouchi-Ghomari, L., & Ghomari, A. R. (2013). Translating natural language competency questions into
SPARQL Queries: a case study. In 1st International Conference on Building and Exploring Web Based
Environments (pp. 81-86).
