Cloud Decision Support System for Risk Management
in Railway Transportation
Wojciech Górka
1
, Jacek Bagi
´
nski
1
, Michał Socha
1
, Tomasz Ste¸clik
1
, Dawid Le
´
sniak
2
, Marek Wojtas
2
,
Barbara Flisiuk
1 a
and Marcin Michalak
1 b
1
Research Network ŁUKASIEWICZ — Institute of Innovative Technologies EMAG,
ul. Leopolda 31, 40-189 Katowice, Poland
2
Telvis Sp. z o.o., ul. Karoliny 4, 40-189 Katowice, Poland
Keywords:
Decision Support System, Safety Management System, Risk Management, Software Support, Cloud Server.
Abstract:
The paper features the development of a decision support system for railway transportation based on risk
management. The implementation of risk management in this area is required by EU and national regulations.
At present, there are no dedicated systems that would provide complex support of this process and, at the
same time, enable to exchange experience and develop a common knowledge base not only on the level of
a particular railway company, but also the whole industry. The solution presented in this paper assists the
personnel responsible for risk management, allows to use a knowledge base and expertise, and supports data
exchange between particular organizations on the railway market.
1 INTRODUCTION
The risk management process is carried out in an or-
ganization by its managers and employees. It is re-
flected in the organization’s strategy and covers the
whole organization. The goal of risk management is
to identify events, or potential events, which could
impact the fulfilment of the organization’s objectives
and to keep the risk on an accepted level.
Each organization is vulnerable to certain risks
which might seriously disrupt its functioning. Rail-
way transportation is particularly important in this re-
spect because it is a part of the national critical infras-
tructure. Having this in mind and realizing the ben-
efits stemming from risk management, such as lower
maintenance cost due to deeper awareness of events
and security incidents, preparedness for crisis situa-
tions, and selection of security measures adequate to
potential losses, it is obvious to recognize risk man-
agement as an indispensable process in railway trans-
portation companies (a case study of traction break
due to icing and description of a decision making pro-
cess are provided in (Bagi
´
nski et al., 2019)).
This requirement results not only from the desire
to improve security and reduce the frequency and con-
a
https://orcid.org/0000-0002-0106-591X
b
https://orcid.org/0000-0001-9979-8208
sequences of events, but also from legal regulations
imposed by national–level authorities (e.g. in Poland
(Ministry of Infrastructure and Development, 2015))
and EU-level institutions (e.g. (European Commi-
sion, 2012)). Currently, there are several standards
where expert knowledge on risk management is col-
lected (Rausand, 2011).
The issues of risk management can be considered
in terms of making decisions which are appropriate
for the sake of the current environmental conditions
of the process and for the sake of expected future state
of the process.
Decision support systems (DSS) have been ap-
plied more and more often in different areas of hu-
man activity. The DSSes depending on the areas
and purposes of their application make use of ar-
tificial intelligence methods, knowledge engineering,
multi-criteria analysis, operational research, and deci-
sion theory.
Although there are known supporting systems for
public urban transport (Bellini et al., 2017) such solu-
tions cannot be applied directly in rail transport. It is
mostly due to the fact that rail transportation is much
more restricted: it is more difficult to get the permis-
sion for rail transportation, there are fewer carriers
and other companies, the rail traffic is more regular
and scheduled (similarly to air traffic).
So, in the railway industry, there are currently no
Górka, W., Bagi
´
nski, J., Socha, M., Steclik, T., Le
´
sniak, D., Wojtas, M., Flisiuk, B. and Michalak, M.
Cloud Decision Support System for Risk Management in Railway Transportation.
DOI: 10.5220/0007837904750482
In Proceedings of the 14th International Conference on Software Technologies (ICSOFT 2019), pages 475-482
ISBN: 978-989-758-379-7
Copyright
c
2019 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
475
systems that would allow continuous exchange of in-
formation on safety and security (e.g. information
regarding threats, existing vulnerabilities). Accord-
ing to national legal requirements, stakeholders of the
Polish railway system (railway companies, infrastruc-
ture managers, entities responsible for maintenance)
are obliged to immediately report accidents, serious
accidents and incidents to the National Railway Ac-
cident Investigation Commission and the national Of-
fice of Rail Transport (ORT).
The lack of any DSS dedicated to risk manage-
ment in the railway transportation branch was one of
the motivations to launch the project called Central
Threat Register (CTR): The management system of
information and expertise knowledge about threats for
railway transportation safety.
In the paper the selected elements and aspects of
developed CTR system are presented. They focus on
the risk management processes, cloud architecture of
the system, problems of data sharing and anonymiza-
tion. The necessity of such system development came
straight from rail companies. CTR is designed as a
cloud system, which comes from the business model
of its selling and licensing (see Section 5.1). More-
over, such an architecture improves data collecting
and sharing as well as performing statistical and risk
analyses.
The CRT system is dedicated to support security
representatives work in railway transport companies,
support activities in risk management, provision of
transport services and rail infrastructure maintenance.
Moreover, the risk analysis module and decision sup-
port module are more generic and may be used in dif-
ferent branches of industry too.
The system is still in the development phase, in
which rail transportation experts participate. During
this phase real acccidents and threats cases are ana-
lyzed.
The paper is organized as follows: it starts from
a description of the results of a questionnaire con-
ducted amongst 24 Polish railway companies; after-
wards, a short description of related works in the areas
of risk management and DSS systems is presented;
the following section brings results of some prelim-
inary studies in the CTR system requirements; next
section presents the selected elements of the system
architecture and is followed by a presentation of two
of the most important CTR system use cases; the pa-
per ends with some conclusions and perspectives of
further works and the system development.
2 RISK MANAGEMENT AND
DECISION SUPPORT IN
POLISH RAILWAY
TRANSPORTATION
Risk management is a relatively new and open is-
sue in Polish railway transportation: the appropri-
ate Polish standard comes from 2018 (Polish Com-
mittee for Standardization, 2018). As an element
of risk management, railway transportation authori-
ties require new fact sheets, reports and summary in-
formation reports (e.g. published by ORT). More-
over, railway companies do not exchange information
about threats, though they unanimously declare they
would be willing to use such data.
To better understand different approaches to risk
management performing, the authors conducted a
questionnaire among safety managers from 24 Polish
railway companies. Almost 80% of those questioned
said they did not use any IT systems supporting risk
management, security management (SMS Secu-
rity Management System) or maintenance (MMS
Maintenance Management System). Only 4 respon-
dents admitted they used IT management support sys-
tems. Most respondents used basic commonly avail-
able programs to maintain threats and events registers
and to prepare reports, i.e. text editors and spread-
sheets. Alternatively, these documents were produced
as paper ones. Fewer than 30% of those questioned
used IT systems to support the collection and analy-
sis of information about events. Only 2 respondents
declared they used IT support systems to exchange
information about the occurring events and threats.
Over 90% of the respondents admitted they had not
used any systems supporting mutual information ex-
change and 100% of them were willing to apply a sys-
tem for anonymous exchange of information, should
such a system be developed.
The above shows that the current level of infor-
mation exchange is not enough. This results from the
lack of knowledge, including the lack of awareness
about benefits coming from risk management and
possibility to get prepared for the threats before they
actually occur. Now most corrective actions and anal-
yses are made no sooner than after the event occurs or
as a result of the event. Moreover, there are only few
available risk management support tools dedicated to
particular industries or infrastructures. The existing
tools do not support information exchange or use of
data from the whole industry or infrastructure.
ICSOFT 2019 - 14th International Conference on Software Technologies
476
3 RELATED WORKS
In this section a short overview of the present ap-
proaches to risk management and existing applica-
tions of DSS will be presented.
3.1 Risk Management
There are a number of solutions on the Polish mar-
ket which offer risk management support for railway
transportation, e.g. RailSoft (Petrosoft, 2017). The
RailSoft software is an integrated security manage-
ment system for railway companies in accordance
with the requirements of MMS/SMS. The software
has an analysis module RAMS (Reliability, Avail-
ability, Maintainability, Safety) which enables to test
railway systems in terms of their reliability, mainte-
nance and application safety. RailSoft also allows
to generate reports about security events in the com-
pany, which can be sent to the national ORT. Still, the
system does not ensure the exchange of information
about security events between railway companies.
There are also applications for risk analysis and
assessment, though these solutions are not dedicated
to railway companies. The available software was de-
signed to assess the risk for classified information, oc-
cupational hazards and operational risk. For example,
the application Risk Analysis (F-tec, 2019) enables to
analyze classified information in accordance with the
requirements of the Internal Security Agency.
There is no obligation to immediately inform
about identified internal threats or vulnerabilities.
ORT provides a portal where people can report de-
tected threats, security issues. However, in practice,
even if railway companies report such information,
usually they do not do it on a regular basis. There
are more elaborate tools supporting analyses, e.g. Re-
liability Workbench (Isograph, 2019), but they do not
offer the threat information exchange option.
Basically, on the European market there are no
integrated systems or platforms for the management
of railway threats, yet there have been attempts to
develop and implement such platforms. For exam-
ple, European Railway Agency (ERA) launched the
Safety Alerts IT Tool (SAIT) platform to exchange in-
formation about events that can potentially lead to ac-
cidents. The obligation to exchange such information
is a new requirement imposed by a European directive
(European Parliament, 2016). ERA facilitates one
more application, ERA SMS, which enables carriers
and managers to assess the functioning of their secu-
rity management systems. The application allows to
check the conformance with new SMS requirements
(so called 4th Railway Package).
3.2 Decision Support Systems
Decision support systems have evolved with the de-
velopment of technology (Shim et al., 2002) and
as it was mentioned in the Introduction section,
there are several types of DSSes defined in literature
(Power et al., 2015), such as: data-driven, model-
driven, document-driven, communication-driven, and
knowledge-driven DSS. A decision support system
can be of one type, however, it can also contain sub-
systems representing different DSS types.
A characteristic feature of a data-driven DSS is ac-
cess to and processing of time series that can be inter-
nal or external company data (Power, 2008). The data
that are processed within the system can be stored in
various locations starting from a file system, where a
query and retrieval functionality is provided, up to a
data warehouse (Kimball and Ross, 2011; Poe et al.,
1997), where advanced data manipulation methods
are available. An appropriate data repository is par-
ticularly important for further analysis, inference and
decision support. Thanks to that, it is possible to in-
tegrate data from various sources and to clean and
prepare them. Such data-driven systems are used for
data originating, e.g., from sensors (Dong et al., 2018;
´
Sl˛ezak et al., 2018; Janusz et al., 2017).
The data stored in a system repository can be pro-
cessed by means of advanced data analysis methods
in order to derive a model assisting in decision mak-
ing. The systems where a model plays central role
are called model-driven DSSes (Power and Sharda,
2007). Such systems enable a non-technical user to
access a model by a dedicated interface. Additionally,
the created model is intended to be applied repeatedly
in the same or similar decision situation. The mod-
els utilized in the system can be of various types, e.g.,
differential equation models, analytical hierarchy pro-
cess based models, or forecasting models.
4 PRELIMINARY STUDIES
The support of risk management has to comprise
supporting tools (which implement certain knowable
practices), e.g. risk analysis methods, risk analyzers,
vulnerability analyzers, countermeasures. The sup-
port must be based on credible data on whose basis
(having already concrete quantitative data) it is pos-
sible to recommend certain solutions or to assess the
importance of defined threats. The data necessary for
risk analyses cover a wide spectrum of information,
in terms of range, structure, granulation, and quanti-
tative aspects. A part of the data which are the ba-
sis to conduct risk analyses are entered by the opera-
Cloud Decision Support System for Risk Management in Railway Transportation
477
tors of domain systems, a part come from online au-
tomatic sensors, and another part are secondary data,
i.e. results of partial analyses. Thus, it is necessary for
the system to acquire information about events (acci-
dents, incidents, failures of railway stock, equipment,
etc). The acquisition of these data will give access to
the statistics about the occurrence of certain phenom-
ena within the organization. This, in turn, allows to
follow the trends of identified risks. In addition, the
acquisition of data about events makes it possible to
examine the causes of certain phenomena and points
to interim methods to solve them. Finally, it is possi-
ble to reason about the efficiency of the applied solu-
tions to mitigate the consequences of the events or to
reduce their frequency. The causes, actions, threats,
and consequences are elements used in the analytical
part.
Apart from quantitative data (based on events in
the company) it is necessary, particularly during the
first stage of the system work, to support the users
with expert knowledge. This can be done by provid-
ing contact with a security expert. The users should
be able to contact the experts directly. Direct commu-
nication in the system will come down to running an
experts base and defining the range of their compe-
tence. Expert knowledge is not only the support from
experts while solving current problems but also the
knowledge saved and available in the system.
Among the prerequisites identified during prelimi-
nary analyses of the system design there is an assump-
tion that the system should have a modular structure.
Thanks to that it will be possible to use a part of the
developed system in other risk management systems
dedicated to domains other than railway transporta-
tion. The development of a domain system only for
one sector (railway transport) would raise the cost of
an individual IT system. This led to the development
of a modular system whose components can be used
to construct systems dedicated to other industrial sec-
tors. Though the PN-ISO 31000 (Polish Committee
for Standardization, 2018) standard remains the ba-
sis of the system, the specifics of particular sectors
(e.g. railway, telecommunications, finances, fuels,
road transport) suggest that making a universal sys-
tem for all domains would not be a good solution.
Thus the common part of the system will cover certain
definitions of data structures, algorithms implementa-
tion, program libraries, and editors (tools) supporting
knowledge acquisition.
All these assumptions and requirements laid down
the basis to carry out the requirements analysis and
the functional analysis of the system.
5 CLOUD DSS FOR RISK
MANAGEMENT IN RAILWAY
TRANSPORT
In the first stage of the system design the project
team focused on a data schema that would be general
enough to become a kind of a common point for risk
management systems. Such a universal data structure
allows not only to develop domain systems on the ba-
sis of the same data structure, but also to exchange
certain data elements between particular instances of
the system (this is important in the case of cloud so-
lutions). The first step was to systematize the terms
used in the system and to determine the context of
their use in particular parts of the system (Figure 1).
Figure 1: Dependencies and data flow in the system.
In risk management there are four basic terms:
threat, vulnerability (cause), consequence, and pre-
ventive action (security measure). Their mutual re-
lations can be seen in Figure 1. In addition, risk man-
agement covers the assessment of the current situa-
tion. Therefore the system should store the events that
would make the basis for the analyses. An event will
be assigned to an existing threat. The statistics of the
event occurrence can be used for the assessment of
threats, probability and potential consequences of the
threat materialization.
5.1 System Architecture
The system architecture should allow for easy access
to the system, data exchange between particular com-
panies of the railway sector and mechanisms enabling
to preserve and promote expert knowledge in the sys-
tem. The best and the most commonly used method
is to develop the system in a cloud as it is presented
in Figure 2.
Such a solution facilitates the system commercial-
ization, particularly when it comes to smaller compa-
nies of the railway sector: it is possible to adapt the
price to the client’s financial abilities, reduce mainte-
nance cost, eliminate the necessity to install the sys-
tem, and provide online update (Avram, 2014). An-
other argument is that the exchange of data between
ICSOFT 2019 - 14th International Conference on Software Technologies
478
Figure 2: System architecture.
particular users is easier. It is not necessary to make
complicated integration of systems, either between
each other or with the central node all data are stored
in one point and it is easy to use them, provided that
the given company gives consent to that. The access
to the system will be provided by a browser – the sys-
tem interface will have a WEB form.
The revision of existing practices and interviews
with the personnel responsible for keeping records of
risk management events showed that the events col-
lection and risk analysis were strictly related to each
other. Frequently, the information belonging to the
event description, e.g. consequences, causes or pre-
vention actions, was treated already as a resulting ele-
ment of risk analysis. In the developed system it was
proposed to separate these two areas. Thus the system
is logically divided into the supervision part related to
the collection of events and the analytical part where,
based on these events, threats are defined and risk in-
dicators calculated.
An important element of the system architecture
is the method of defining the knowledge base. The
system will have an independent data structure where
the patterns of threats, security controls, vulnerabili-
ties, and consequences will be stored. These data will
be constantly updated by the system experts based on
their experiences and observations. The users who
will define their own risk-related elements (threats,
security controls, etc) will refer to the patterns from
the knowledge base the patterns will be given as
original elements for users-defined elements.
This solution will allow to acquire global informa-
tion for the whole system, generate charts and statis-
tics based on data from the whole sector. If, defining
their own new threats, the users refer to a threat from
the knowledge base, the information or data about this
threat described by one user can be used as hints for
other users who refer to this element too. This aspect
of the system is presented in Figure 3.
5.2 Data Anonymization
The system architecture facilitates and promotes mu-
tual use of data by particular stakeholders of the rail-
way market who will make use of the developed sys-
tem. As it was mentioned before, this situation evokes
certain concern about possible disclosure of business
secrets. Therefore it is necessary to anomymize the
data, so that only this information could be provided
which does not subject the user to unauthorized dis-
closure of confidential data or business secrets. To
anonymize the data, it is possible to use generaliza-
tion or permutation (Cormode and Srivastava, 2010)
as well as, in certain cases, clustering-based or graph
modification approaches (Zhou et al., 2008). The de-
veloped system adopts data generalization with the
use of the knowledge base. Thus the anonymization
is reflected in the system architecture and in the data
model employed in the system. First of all, it was de-
cided to exchange aggregated quantitative data. For
example, it would be possible to get access to col-
lective data about the number of occurrences of the
given type of events (translated into certain threats)
but not to the contents of the events as such. Cer-
tain collective data will be collected for knowledge
base elements and the related elements defined by the
users. The technique of using the knowledge base as
an intermediary between data of different organiza-
tions guarantees that the anonymization will be ap-
plied.
6 SAMPLE USE CASES
In this subsection two use cases are presented. They
reflect one of the most important situations: the
first one represents a single risk analysis and the
second one illustrates data flow in data analyzing,
anonymization and sharing.
Cloud Decision Support System for Risk Management in Railway Transportation
479
Figure 3: Use of knowledge base.
The first use case deals with the single risk analy-
sis performance. The threat analysis will comprise, in
fact, several analyses. It will contain not only a risk
analysis for a certain threat but also the identification
of the threat causes and consequences as well as pos-
sible security measures. Different types of analyses
will support these operations. For the given threat the
final risk level will be determined with the use of one
of available analyses, e.g. FMEA.
A threat analysis, conducted by a railway com-
pany employee responsible for analyses, is done with
the use of available resources, such as the cata-
logues of previously collected data: consequences,
security measures, causes. For the purpose of the
analysis the user will have access to events asso-
ciated with the threat, so that, based on their de-
scriptions (including: recommendations of the com-
mission, quantitative consequences – victims, delays,
losses) he/she can draw proper conclusions and define
suitable causes, consequences and security measures.
An expert analysis, in turn, will be ordered by the
company’s representatives. Here the expert can be an
employee of a railway company, ORT or other institu-
tion. He/she must be familiar with the railway trans-
port specifics and have enough expert knowledge in
this domain. The expert will receive an order with
the description resulting from already analyzed data.
The description will have a form of a text paraphrase
of the collected data (from threat analysis), so the ex-
pert will not have direct access to threat-related data
(causes, security measures, consequences) but will be
provided with their text version.
Based on the description, the expert will be able
to examine the issue. The result of his/her work will
be either a description with remarks or his/her own
analyses. It will be possible for the expert to get fa-
miliar with the company’s own analyses in the form
of documents from analyzers.
The actors of the use case (security representative
and security expert) and the whole use case are pre-
sented in Figure 4.
The second use case deals with the data analysis
and sharing. A railway company employee who is re-
sponsible for security initiates the process of provid-
ing information about new threats, security measures,
etc. At this stage he/she decides which information
should be released to the remaining users of CTR. The
provided data are anonymized in the next stage.
The second actor of the process is the system ad-
ministrator whose task is to approve the anonymized
data provided by the employees of particular railway
companies. This step allows to ensure coherence and
suitable quality of data to be stored in the common
data base. The process itself is finalized by supple-
menting particular catalogues of the database.
The scheme of the use case is presented in Figure
5.
7 CONCLUSIONS AND FURTHER
WORKS
In the paper the new developed decision support sys-
tem for risk management in Polish railway trans-
portation is presented. The description consists of
an overview of most general aspects such as data
anonymization, data sharing, experts involvement,
decision support, and cloud architecture.
Big railway transportation companies, which took
part in the above mentioned questionnaire, have at
their disposal some tools supporting risk analyses.
Still, these are local-level solutions. Therefore the
project is not focused on the risk analysis as such
but on collecting information from the whole railway
transport domain (e.g. information how often secu-
rity incidents occur, etc). The risk analysis module is
extra support for smaller railway transportation com-
panies which have not used such tools so far. The
AHP-based decision support module, in turn, is an
added value for this domain as this method has not
been employed by railway carriers before.
The presented ideas for the system development
are implemented progressively within the project. Not
until the implementation stage has been concluded,
the system use by real users will allow to assess
whether the decisions about possible solutions were
ICSOFT 2019 - 14th International Conference on Software Technologies
480
Figure 4: The use case illustrating the single risk analysis.
Figure 5: The use case illustrating the single data analysis and sharing.
right. The system is built in the Agile methodology,
so that permanent contacts with domain experts could
allow to verify the correctness of successive system
sections. A threat to proper functioning of the sys-
tem may be the users’ lack of willingness to exchange
data between each other (in spite of the existing data
anonymization). A supporting action here will be a
system of incentives, i.e. if a user wants to access data
from other users, he/she has to give access to his/her
data in return.
An important element that will impact the success
of the system will be its status during the start–up
whether the system already has some data, which are
an added value, in the form of domain knowledge.
Thus, the contents of the data base should be prepared
by the domain experts already at the stage of the sys-
tem implementation. This will be done by means of a
special interface for knowledge base editing.
The developed common part for different risk
management systems in the organization will be veri-
fied by an attempt to apply it in a different sector. No
doubt, this will involve some changes or will require
to work out another common elements. It is consid-
ered to give public access to this part of the system
in the form of a framework for risk management sys-
tems.
On the basis of the feedback from the domain ex-
perts the modules od Knowledge Base and Experts
are now filled with data. In addition, the compa-
nies’ databases and the global anonymized database
are filled with the sample initial data. The most im-
portant part of the project the implementation in
selected rail companies — is still ahead.
Our future works will also focus on the system
Cloud Decision Support System for Risk Management in Railway Transportation
481
efectiveness and afficiency. Although the potential
number of system users might not be high (the access
to the rail transportation market is strictly supervised
and the total number of companies does not exceed
one hundred and several dozen) the next implementa-
tion phase may require to overcome some difficulties
in the system operation.
ACKNOWLEDGEMENTS
This work was supported by Polish National Cen-
tre for Research and Development (NCBiR) within
the Operational Programme Intelligent Development:
Grant No. POIR.04.01.02-00-0024/17-00 (Central
Threat Register — The management system of infor-
mation and expertise knowledge about threats for rail-
way transportation safety).
REFERENCES
Avram, M. (2014). Advantages and challenges of adopt-
ing cloud computing from an enterprise perspective.
Procedia Technology, 12:529 – 534.
Bagi
´
nski, J., Flisiuk, B., Górka, W., Rogowski, D., and
Ste¸clik, T. (2019). Multi–criteria decision analy-
sis in the railway risk management process. Com-
munications in Computer and Information Science,
1018:126–138.
Bellini, E., Ceravolo, P., and Nesi, P. (2017). Quan-
tify resilience enhancement of uts through exploit-
ing connected community and internet of everything
emerging technologies. ACM Trans. Internet Technol.,
18(1):7:1–7:34.
Cormode, G. and Srivastava, D. (2010). Anonymized data:
Generation, models, usage. In 2010 IEEE 26th In-
ternational Conference on Data Engineering (ICDE
2010), pages 1211–1212.
Dong, F., Zhang, G., Lu, J., and Li, K. (2018). Fuzzy com-
petence model drift detection for data-driven decision
support systems. Knowledge-Based Systems, 143:284
– 294.
European Commision (2012). Commision regulation no
1078/2012.
European Parliament (2016). Directive 2016/798 on railway
safety.
F-tec (2019). https://f-tec.pl/analiza-ryzyka/.
Isograph (2019). https://www.isograph.com/software/reliability-
workbench/.
Janusz, A., Grzegorowski, M., Michalak, M., Wróbel, L.,
Sikora, M., and
´
Sl˛ezak, D. (2017). Predicting seis-
mic events in coal mines based on underground sensor
measurements. Engineering Applications of Artificial
Intelligence, 64:83 – 94.
Kimball, R. and Ross, M. (2011). The data warehouse
toolkit: the complete guide to dimensional modeling.
John Wiley & Sons.
Ministry of Infrastructure and Development (2015). Com-
mon security indices decree. Dziennik Ustaw, 1061.
Petrosoft (2017). https://petrosoft.pl/produkty/oprogramowanie-
kolejowe/.
Poe, V., Brobst, S., and Klauer, P. (1997). Building a data
warehouse for decision support. Prentice-Hall, Inc.
Polish Committee for Standardization (2018). Polish stan-
dard PN-ISO 31000:2018–08: Risk management
guidelines.
Power, D. J. (2008). Understanding data-driven decision
support systems. Information Systems Management,
25(2):149–154.
Power, D. J. and Sharda, R. (2007). Model-driven deci-
sion support systems: Concepts and research direc-
tions. Decision Support Systems, 43(3):1044–1061.
Power, D. J., Sharda, R., and Frada, B. (2015). Decision
Support Systems, pages 1–4. American Cancer Soci-
ety.
Rausand, M. (2011). Risk Assessment: Theory, Methods,
and Applications. Wiley.
Shim, J. P., Warkentin, M., Courtney, J. F., Power, D. J.,
Sharda, R., and Carlsson, C. (2002). Past, present,
and future of decision support technology. Decision
support systems, 33(2):111–126.
´
Sl˛ezak, D., Grzegorowski, M., Janusz, A., Kozielski, M.,
Nguyen, S. H., Sikora, M., Stawicki, S., and Wróbel,
L. (2018). A framework for learning and embedding
multi-sensor forecasting models into a decision sup-
port system: A case study of methane concentration
in coal mines. Information Sciences, 451-452:112
133.
Zhou, B., Pei, J., and Luk, W. (2008). A brief survey on
anonymization techniques for privacy preserving pub-
lishing of social network data. ACM SIGKDD Explo-
rations Newsletter, 10(2):12 –22.
ICSOFT 2019 - 14th International Conference on Software Technologies
482