=Paper=
{{Paper
|id=Vol-2397/preface
|storemode=property
|title=Workshop on the Control of Systemic Risks in Global Networks (SysRisk2019 - as part of the 14th International Conference Wirtschaftsinformatik) (preface)
|pdfUrl=https://ceur-ws.org/Vol-2397/preface.pdf
|volume=Vol-2397
|authors=Fabian Lorig,Ingo J. Timm,Peter Mertens
|dblpUrl=https://dblp.org/rec/conf/wirtschaftsinformatik/X19
}}
==Workshop on the Control of Systemic Risks in Global Networks (SysRisk2019 - as part of the 14th International Conference Wirtschaftsinformatik) (preface)==
https://ceur-ws.org/Vol-2397/preface.pdf
Workshop on the Control of
Systemic Risks in Global Networks
(SysRisk2019 - as part of the 14th International
Conference Wirtschaftsinformatik)
Fabian Lorig1, Ingo J. Timm1, and Peter Mertens2
1
Trier University, Center for Informatics Research and Technology, Trier, Germany
{lorigf,itimm}@uni-trier.de
2
Friedrich-Alexander-University Erlangen-Nürnberg,
Wirtschaftsinformatik I, Nürnberg, Germany
peter.mertens@fau.de
Abstract. The emergence of global networks also results in the occurrence of
systemic risks that might affect the stability of the overall system. To cope with
these risks, this workshop on the “Control of Systemic Risks in Global Networks”
(SysRisk2019) provides a platform for the collection and discussion of
innovative approaches, methods, and theories but also of practical problems from
the areas of simulation, artificial intelligence, operations research, and statistics.
This enables the exchange of experiences and methods between scientists and
practitioners. The SysRisk2019 workshop has taken place as part of the 14th
International Conference Wirtschaftsinformatik (WI2019) in Siegen, Germany
on February 24th 2019.
Keywords: Grand Challenge, Systemic Risk, Reference Framework.
1 Introduction
Modern communication networks lead to a stronger coupling of and interdependency
between social and economic areas. Examples are electronic marketplaces, which
enable ever faster transactions, worldwide production networks, which allow for higher
specialization with increasing efficiency, and smart grids, which facilitate the provision
of energy in the European Single Market by means of flexible control. The resulting
worldwide and interconnected networks increasingly decide on the competitiveness of
enterprises.
On the one hand, this development is promoted by a strong demand pull for
innovative technologies that emanates from companies. This results from the
companies’ endeavor to take advantage of environmental differences in a “globalized
world”. Examples are increasing sales opportunities in emerging countries, low labor
costs, special competences in the development and production of electronic
components or software products, discoveries of raw materials, and tax conditions.
1
On the other hand, there is an increasing technology pressure. This is due to an
increasing performance-cost ratio of data management as well as from the fact that
modern multi and manycore systems accelerate or initially enable the solving of
sophisticated planning, disposition, and control algorithms. Moreover, the
advancement of traditional methods, e.g., artificial neural networks and deep learning,
allows for the discovery of patterns and the investigation of systems that remained
hidden or were inaccessible before.
Along with these worldwide networks, systemic risks emerge which affect the
stability of the overall system [1]. Examples of potential failures are flash crashes in
high-frequency trading, production downtime due to delivery delays, or blackouts in
energy networks. For instance, on September 28th, 2003, power plant failures in Italy
lead to disruptions of the Internet infrastructure, which relied on energy supply and at
the same time was required to control other power plants. This resulted in a cascade of
failures and has nearly caused the collapse of the entire Italian energy supply [2,3].
Obviously, not all risks are equivalent with respect to their probability of occurrence
and of the consequences. Thus, those systemic risks must be identified, which – as
illustrated by the example – affect the stability of the overall system and are not
considered as part of the risk assessment of the independent subsystems. Here, the
extent of the risk must be considered as well as the probability of finding an adequate
countermeasure with reasonable effort.
In a joint initiative, which is steered by the German Informatics Society (Gesellschaft
für Informatik e.V.; GI), Information Systems Research and Computer Science have
selected the control of systemic risks in global networks as one of the five most
important Grand Challenges for the future [4]. From an information system research
perspective, two major interests can be identified: On the one hand, the availability as
well as the situational aggregation and interpretation of decision-relevant information
and on the other hand the autonomous identification, quantitative estimation, and
flexible reaction to risks. To provide a forum for the presentation and discussion of
respective approaches, the SysRisk2019 workshop has been arranged as part of the 14th
International Conference Wirtschaftsinformatik (WI2019) in Siegen, Germany on
February 24th 2019. In this paper, current challenges and trends that lead to this
workshop as well as the reference framework was used for the classification and
discussion of approaches are outlined [5]. Moreover, the structure of the workshop is
presented and all contributions that were made are briefly introduced and related to
each other.
2 Current Technology Pressure
In information system as well as computer science research, there are ongoing
discussions whether networks can be designed or dynamically emerge from the
interaction of devices with network technologies: Worldwide networks are not designed
as part of an “engineering process”, they are created through the interaction of
interconnected systems as emergent phenomenon and must be described and
understood [6].
2
The need for a development of methods for the design of such networks can be
identified when investigating the current technology pressures. Developments that can
contribute to the control of systemic risks include but are not limited to:
1. Communication Networks: Advances in communication networks, e.g.,
an increasing performance-cost ratio of communication channels
(hardware) and greater flexibility in routing (software), which allow for
prioritized communication in case of emergency.
2. Simulation: Recent developments in simulation from a tool for planning
support to a real-time assistance for decision support through the
development of innovative formalisms, e.g., system dynamics or agent-
based simulation, and due to the immediate availability of current data.
3. Machine Learning: Revolutionary progress in machine learning that is
facilitated by the increasing availability and amount of (training) data as
well as shift from multi to multi and manycore computing. This allows for
the use of deep learning, convolutional neural networks as well as data, text,
and opinion mining techniques.
4. Decentralized Control: The availability of approaches for decentralized
and adaptive control with autonomous software agents, multiagent systems,
and organic computing promotes the high-tech strategy “Industry 4.0”.
5. Transaction Processing Systems (Blockchain): New forms of transaction
processing systems, e.g., blockchain, allow for the tamper-resistant and
decentralized organization and logging of safety-critical operations in
processes such as access or updates of sensitive data.
6. Multilayer and Multiplex Networks: A shift from the analysis of isolated
and homogenous networks to the investigation of multilayer and multiplex
networks (interdependent networks).
7. Convergence: The convergence of technical systems and processes leads
to the unification of business models and technologies across sectors.
Through this, technical and economic success of one domain might
dominate another domain, e.g., successful business models of internet
giants can compete with stationary trade in the physical world even though
the horizon of experience is considerably lower.
Due to disciplinary barriers, the aforementioned technology areas are not yet
sufficiently developed, applied, or transferred for controlling systemic risks. This limits
the opportunities for action that can be undertaken to prevent the potentially dramatic
consequences of systemic risks. Still, these technologies have a high potential to
contribute as component of a solution for controlling systemic risks.
Considering disaster management strategies, for instance, it can be illustrated how
disciplines can learn from each other and benefit from the experiences of other
disciplines. Insurance companies make use of reinsurances to handle major claims
which could result in their insolvency. Such approaches are also applicable to supply
chain management as protection against supply shortages that might result in
disruptions of the own production of goods. In this regard, supply chain management
3
can also learn from insurances as systemic risks emerge from networks of reinsurances
which can potentially result in uncontrollable chain effects that lead to global crises.
3 Reference Framework
Suitable technologies and methods for controlling systemic risks are diverse. Thus, to
classify and distinguish different approaches, we suggest the use of a morphological
box. It serves as a reference framework for discussion within the workshop as
approaches can be classified and assessed according to different dimensions. In Figure
1, the morphological box is illustrated that is used for the assessment of the approaches
that are presented as part of this workshop. For each approach, the aspects of networks,
risks, and decision situation are focused.
To this end, the domain focus of the workshop lies on logistics, finance &
insurances, and public services, yet, also contributions from other domains are
welcome. With respect to the type of risk that is addressed by the approaches, it can be
differentiated into five types, according to the domain the risk is related to: production,
market, finance, institution, and nature. In addition, also the occurrence of the risk is
classified as regularly, periodically, or rarely. Finally, the decision situation of the risk
can be specified according to the risk’s predictability as well as by the authority which
is the decision maker.
Figure 1: Reference framework for the classification and discussion of approaches.
4 Contributions to the Workshop
To address the Grand Challenge of controlling systemic risks in global networks,
this workshop aims at both the collection and discussion of innovative approaches,
methods, and theories but also practical problems from the areas of simulation, artificial
intelligence, operations research, and statistics. To this end, the goal of the workshop is
to provide a platform for the exchange of experiences and methods between scientists
and practitioners. Moreover, the development of a medium-term research agenda shall
be promoted for targeting this Grand Challenge.
4
As part of the SysRisk2019 workshop, 17 presentations were given providing
different perspectives on systemic risks in global networks. According to their focus,
the contributions were assigned to three different sessions: “Analyzing Worldwide
Networks”, “Design of Processes and Networks”, and “Risk and Crisis Management”.
Finally, in a fourth final session, the approaches and results from all previous sessions
were consolidated and related in a group discussion.
4.1 Analyzing Worldwide Networks
As part of the session “Analyzing Worldwide Networks”, four papers have been
presented. Most of the presented papers discuss the use of computer simulation as
method to investigate and analyze different aspects of global networks. In their paper
“Towards Systematic Testing of Complex Interacting Systems”, René Schumann and
Caroline Taramarcaz introduce a notion of adaptive systems, which can change their
behavior at run-time. They outline that such systems create a new type of error-behavior
for which conventional techniques cannot be applied and propose a structured
simulation framework to test the behavior of adaptive systems.
Due to the complexity and close connection between networks used in daily life,
systemic risks might occur. To prevent negative effects of these risks, Sören Bergmann,
Niclas Feldkamp, and Steffen Straßburger suggest in their paper “Wissensentdeckung
und Robustheitsanalyse für Simulationsmodelle weltweiter Netze” (“Knowledge
Discovery and Robustness Analysis for Simulation Models of Global Networks”) to
ensure robustness already during the design of networks. Therefore, the authors propose
a data-farming based method for conducting robustness analysis in the domain of
manufacturing. For adapting the proposed method to complex networks, various
research needs are presented.
The paper “Behavior Mining Methods for Dynamic Risk Analysis in Social Media
Communication” by Jan Ole Berndt considers systemic risks of social media with
respect to crisis and reputation management. Behavior mining is introduced for
analyzing communication processes which result from individual behaviors of
interconnected users.
Finally, the paper “Cryptocurrency Crashes: A Dataset for Measuring the Effect of
Regulatory News in Online Media” by Achim Klein, Lyubomir Kirilov, and Martin
Riekert extend prior research on effects of regulatory news on cryptocurrency markets,
to analyze the effects of restrictions of usage or even complete bans of cryptocurrencies.
To measure the effect of regulatory news on cryptocurrencies, a dataset of online media
news for application to empirically study the effects on Bitcoin pricing is presented.
4.2 Design of Processes and Networks
As part of the second session on “Design of Processes and Networks”, five papers
were presented. In their paper “Structural Change in Insurance: The Emergence of
Comprehensive Value Networks” Albrecht Fritzsche and Alexander Bohnert
investigate the convergences of business activities across different sectors of critical
infrastructure and identify two distinctive patterns of convergence. The focus of their
5
investigation is set to insurance industry, which has proven openness for new types of
offerings and business models as part of the digital transformation.
Sebastian Lehnhoff and Astrid Nieße propose an approach to derive relevant solution
parameters for optimizing distributed business processes and integrating the parameters
into existing supervisory automation and control concepts. In their paper “Event-driven
Reorganization of Distributed Business Processes in Electrical Energy Systems”, they
apply event-driven reorganization and multi-agent systems to find a solution in time
and to handle a large number of conflicting objectives.
Together with Frank Eggert, Astrid Nieße and Sebastian Lehnhoff contributed a
second paper to the workshop: “Managing Conflicting Interests in Socio-technical
Energy Systems – How to Identify and Mitigate Intra-actor Interests as Risk Factors.,
Here, they discuss two contradictory paradigms, i.e., complexity reducing and
complexity increasing measures, to ensure stable operations in global networks. It is
shown why intra-actor conflicts arise from adding both complexity and reducing
transparency at the same time. The authors propose a research agenda and present
existing approaches and open issues regarding an abstract model of decision conflicts,
a dynamic model to evaluate the effect of transparency changes during runtime, and
metrics to evaluate degrees of autonomy and transparency in the context of energy
systems.
The paper “Ein hierarchischer Ansatz des Risikomanagements zur Gestaltung
robuster Liefer- und Transportnetzwerke” (“A Hierarchical Risk Management
Approach for the Design of Supply and Transport Networks”) by Patricia Rogetzer and
Stefan Minner proposes a two-step hierarchical planning approach to design supply and
transport networks. By this means, more resilient networks can be designed as proactive
and reactive methods prevent the network from disturbances and interruptions.
Gilbert Fridgen and Martin Weibelzahl discuss the potential contributions of
Blockchain technology to systemic risk management in global supply chains and
networks in their paper “(How) Can Blockchain Contribute to the Management of
Systemic Risks in Global Supply Networks?”. The authors argue that distributed
ledgers like Blockchains in combination with secure multiparty computation could help
to detect and manage systemic risks in large supply networks. Therefore, Blockchain
could take the role of a central authority, which currently does not exist in large supply
networks, and could ensure the access of data to anamnesis, diagnose or therapy
systemic risks.
4.3 Risk and Crisis Management
As part of the third session on “Risk and Crisis Management”, three presentations
were given. The paper „Managing Systemic Risks: Opening up Public Crisis
Management in Global Networks” by Moreen Heine describes design options and areas
of activity based on system-theoretical foundations for collaborative and post-
bureaucratic crisis management in global networks. Related work and research needs
with a focus on empirical research are outlined and a research framework for public
crisis management in global networks is proposed.
6
Increasing requirements of more volatile markets and the digital transformation of
business are challenges for supply network planning and extant approaches are not
capable of dealing with these challenges. Therefore, the paper “Risk-based Planning in
Smart Supply Networks: The Merit of Multi-model Analytics” of Gerd J. Hahn
investigates the merit of multi-model-based analytics approaches using a risk-based
planning perspective along three lines: planning scope, conceptual framework, and
methodological approaches.
Peter Fettke and Peter Loos describe in their paper “Prädiktives Monitoring von
Geschäftsprozessen zur Beherrschung von Risiken in weltweiten Netzen auf Basis von
Process Mining und Simulation” (“Predictive Monitoring of Business Processes for
Controlling Systemtic Risks in Global Networks”) how simulation and process mining
can be used for predictive monitoring of business processes. To this end, their work
proposes how artificially generated data can be used for this purpose.
5 Acknowledgements
We would like to thank all those that contributed to the initiative on the “Control of
Systemic Risks in Global Networks” and all participants of the SysRisk2019 workshop
for their fruitful comments and innovative ideas. In particular, we would like to thank
Matthias Schulte-Althoff (Freie Universität Berlin), Amit Patil and Hermann de Meer
(Universität Passau), and Peter Bradl (Hochschule für angewandte Wissenschaften
Würzburg-Schweinfurt) for their interesting presentations and their participation in the
workshop. Finally, we would like to express our gratitude to Fabian Mirz for his
vigorous support throughout the preparation of these CEUR workshop proceedings.
References
1. Buldyrev, S. V., Parshani, R., Paul, G., Stanley, H. E., & Havlin, S. (2010). Catastrophic
cascade of failures in interdependent networks. In: Nature, 464 (pp. 1025-1028).
2. UCTE (2004). FINAL REPORT of the Investigation Committee on the 28 September 2003
Blackout in Italy. UCTE Report – April 2004.
3. Johnson, C. W. (2007). Analysing the causes of the Italian and Swiss blackout, 28th
September 2003. In Proceedings of the twelfth Australian workshop on Safety critical
systems and software and safety-related programmable systems. Volume 86 (pp. 21-30).
Australian Computer Society, Inc.
4. Mertens, P. & Barbian, D. (2015). Researching “Grand Challenges” – A “Grand Challenge”.
In: Business & Information Systems Engineering (2015) 57:6 (pp. 391-403).
5. Lorig, F., Timm, I. J., & Mertens, P. (2019). Control of Systemic Risks in Global Networks
– A Grand Challenge to Information Systems Research. In: Tagungsband der 14. Internati-
onalen Tagung Wirtschaftsinformatik, WI2019 (pp. 2034-2038).
6. Gkantsidis, C., Mihail, M., & Saberi, A. (2003). Conductance and congestion in power law
graphs. In: ACM SIGMET-RICS Performance Evaluation Review Vol. 31, No. 1 (pp. 148-
159), ACM.
7