Transcript Folien - Humor in der Informatik

Kooperation erforschen – Kooperationen leben
Management-by-Acronym (MBA)
A New Paradigm for Evaluating and Running
ICT Research Projects
Univ.-Prof. Dr. Peter Reichl
Forschungsgruppe “Cooperative Systems” (COSY)
Fakultät für Informatik der Universität Wien
Universität Dortmund, 05.12.2014
So weich, dass man
es blind erkennt…
Management-by-Acronym (MBA)
A New Paradigm for Evaluating and Running
ICT Research Projects
Univ.-Prof. Dr. Peter Reichl
Forschungsgruppe “Cooperative Systems” (COSY)
Fakultät für Informatik der Universität Wien
Universität Dortmund, 05.12.2014
So weich, dass man
es blind erkennt…
Management-by-Acronym (MBA)
A New Paradigm for Evaluating and Running
ICT Research Projects
Univ.-Prof. Dr. Peter Reichl
Forschungsgruppe “Cooperative Systems” (COSY)
Fakultät für Informatik der Universität Wien
Universität Dortmund, 05.12.2014
Background and Motivation
 Starting point: Lehrstuhl Informatik 4, RWTH Aachen
 Project ANNABELLE (Anomalies in Normaluser Network Applications
Based on Experiments in a Lovely Lehrstuhl-Environment), 1996-1998
M. Berger, O. Kubitz, K. Jakobs, P. Reichl, A. Koch: Project ANNABELLE: Anomalies in
Normal-User Network Applications Based on Experiments in a Lovely Lehrstuhl Environment.
Proc. 3rd Biennial SSE European Meeting, p. 26, Freiburg, Germany, Oct. 1996.
Background and Motivation
 Starting point: Lehrstuhl Informatik 4, RWTH Aachen
 Project ANNABELLE (Anomalies in Normaluser Network Applications
Based on Experiments in a Lovely Lehrstuhl-Environment), 1996-1998
M. Berger, O. Kubitz, K. Jakobs, P. Reichl, A. Koch: Project ANNABELLE: Anomalies in
Normal-User Network Applications Based on Experiments in a Lovely Lehrstuhl Environment.
Proc. 3rd Biennial SSE European Meeting, p. 26, Freiburg, Germany, Oct. 1996.
„Annabelle, ach Annabelle,
Du bist so herrlich intellektuell,
Du bist so wunderbar negativ
und so erfrischend destruktiv...“
Background and Motivation
 Starting point: Lehrstuhl Informatik 4, RWTH Aachen
 Project ANNABELLE (Anomalies in Normaluser Network Applications
Based on Experiments in a Lovely Lehrstuhl-Environment), 1996-1998
M. Berger, O. Kubitz, K. Jakobs, P. Reichl, A. Koch: Project ANNABELLE: Anomalies in
Normal-User Network Applications Based on Experiments in a Lovely Lehrstuhl Environment.
Proc. 3rd Biennial SSE European Meeting, p. 26, Freiburg, Germany, Oct. 1996.
Background and Motivation
 Starting point: Lehrstuhl Informatik 4, RWTH Aachen
 Project ANNABELLE (Anomalies in Normaluser Network Applications
Based on Experiments in a Lovely Lehrstuhl-Environment), 1996-1998
M. Berger, O. Kubitz, K. Jakobs, P. Reichl, A. Koch: Project ANNABELLE: Anomalies in
Normal-User Network Applications Based on Experiments in a Lovely Lehrstuhl Environment.
Proc. 3rd Biennial SSE European Meeting, p. 26, Freiburg, Germany, Oct. 1996.
Background and Motivation
 Starting point: Lehrstuhl Informatik 4, RWTH Aachen
 Project ANNABELLE (Anomalies in Normaluser Network Applications
Based on Experiments in a Lovely Lehrstuhl-Environment), 1996-1998
M. Berger, O. Kubitz, K. Jakobs, P. Reichl, A. Koch: Project ANNABELLE: Anomalies in
Normal-User Network Applications Based on Experiments in a Lovely Lehrstuhl Environment.
Proc. 3rd Biennial SSE European Meeting, p. 26, Freiburg, Germany, Oct. 1996.
 After the PhD: Forschungszentrum Telekommunikation Wien (FTW)
 WISQY (Wireless Inter-System Quality of Service), 2002-2004
P. Reichl, M. Umlauft, J. Fabini et al.: Project WISQY: A Measurement-Based End-to-End Performance Comparison of 2.5G and 3G Networks. Proc. 5th Wireless Telecommunications Symposium, Pomona, CA, April 2005.
Background and Motivation
 Starting point: Lehrstuhl Informatik 4, RWTH Aachen
 Project ANNABELLE (Anomalies in Normaluser Network Applications
Based on Experiments in a Lovely Lehrstuhl-Environment), 1996-1998
M. Berger, O. Kubitz, K. Jakobs, P. Reichl, A. Koch: Project ANNABELLE: Anomalies in
Normal-User Network Applications Based on Experiments in a Lovely Lehrstuhl Environment.
Proc. 3rd Biennial SSE European Meeting, p. 26, Freiburg, Germany, Oct. 1996.
 After the PhD: Forschungszentrum Telekommunikation Wien (FTW)
 WISQY (Wireless Inter-System Quality of Service), 2002-2004
P. Reichl, M. Umlauft, J. Fabini et al.: Project WISQY: A Measurement-Based End-to-End Performance Comparison of 2.5G and 3G Networks. Proc. 5th Wireless Telecommunications Symposium, Pomona, CA, April 2005.
 CAMPARI (Configuration, Architecture, Migration, Performance Analysis
and Requirements of 3G IMS), 2004-2006
P. Reichl, J. Fabini, N. Jordan et al.: Quo Vadis, IMS? Das Projekt CAMPARI als mobilnetz-orientierte
Standortbestimmung für das IP Multimedia Subsystem. e&i 123 no. 7/8, pp. 315–322. Springer, Aug. 2006.
Background and Motivation
 Successor Projects:
 CAIPIRINA (Converging towards All-IP: IMS Realization Issues
for NGN Applications), 2006-2008
Background and Motivation
 Successor Projects:
 CAIPIRINA (Converging towards All-IP: IMS Realization Issues
for NGN Applications), 2006-2008
 BACCARDI (Beyond Architectural Convergence – Charging, Applications, Realization and Demonstration of 3G IMS), 2008-2010
Background and Motivation
 Successor Projects:
 CAIPIRINA (Converging towards All-IP: IMS Realization Issues
for NGN Applications), 2006-2008
 BACCARDI (Beyond Architectural Convergence – Charging, Applications, Realization and Demonstration of 3G IMS), 2008-2010
 BIQINI (BACCARDI’s IMS QoS Implementation Initiative), 2010
C. Egger, M. Happenhofer, J. Fabini, P. Reichl: BIQINI – A Flow-based QoS Enforcement
Architecture for NGN Services. Proc. TRIDENTCOM 2010, Berlin, Germany, May 2010
Background and Motivation
 Successor Projects:
 CAIPIRINA (Converging towards All-IP: IMS Realization Issues
for NGN Applications), 2006-2008
 BACCARDI (Beyond Architectural Convergence – Charging, Applications, Realization and Demonstration of 3G IMS), 2008-2010
 BIQINI (BACCARDI’s IMS QoS Implementation Initiative), 2010
C. Egger, M. Happenhofer, J. Fabini, P. Reichl: BIQINI – A Flow-based QoS Enforcement
Architecture for NGN Services. Proc. TRIDENTCOM 2010, Berlin, Germany, May 2010
 Empirical observation: obvious correlation between “efficiency level” of the
acronym [in %] and acquired project budget [MEUR]
Background and Motivation
 Successor Projects:
 CAIPIRINA (Converging towards All-IP: IMS Realization Issues
for NGN Applications), 2006-2008
 BACCARDI (Beyond Architectural Convergence – Charging, Applications, Realization and Demonstration of 3G IMS), 2008-2010
 BIQINI (BACCARDI’s IMS QoS Implementation Initiative), 2010
C. Egger, M. Happenhofer, J. Fabini, P. Reichl: BIQINI – A Flow-based QoS Enforcement
Architecture for NGN Services. Proc. TRIDENTCOM 2010, Berlin, Germany, May 2010
 Empirical observation: obvious correlation between “efficiency level” of the
acronym [in %] and acquired project budget [MEUR]
 EU FP7: similar, but much more hesitant: CHIANTI, SCAMPI, DUPLO …
Background and Motivation
 Successor Projects:
 CAIPIRINA (Converging towards All-IP: IMS Realization Issues
for NGN Applications), 2006-2008
 BACCARDI (Beyond Architectural Convergence – Charging, Applications, Realization and Demonstration of 3G IMS), 2008-2010
 BIQINI (BACCARDI’s IMS QoS Implementation Initiative), 2010
C. Egger, M. Happenhofer, J. Fabini, P. Reichl: BIQINI – A Flow-based QoS Enforcement
Architecture for NGN Services. Proc. TRIDENTCOM 2010, Berlin, Germany, May 2010
 Empirical observation: obvious correlation between “efficiency level” of the
acronym [in %] and acquired project budget [MEUR]
 EU FP7: similar, but much more hesitant: CHIANTI, SCAMPI, DUPLO …
 Research question: can we generalize that? underlying models / metrics?
From Evolutionary to Evaluationary Game Theory
 Idea: theoretical analysis
 Established Framework:
Evolutionary Game Theory  “survival of the fittest”
From Evolutionary to Evaluationary Game Theory
 Idea: theoretical analysis
 Established Framework:
Evolutionary Game Theory  “survival of the fittest”
 Problem: Nash equilibria exist, but asymmetry between
applying project consortia and evaluation body (a.k.a. EU)
From Evolutionary to Evaluationary Game Theory
 Idea: theoretical analysis
 Established Framework:
Evolutionary Game Theory  “survival of the fittest”
 Problem: Nash equilibria exist, but asymmetry between
applying project consortia and evaluation body (a.k.a. EU)
 Idea: use “Nash-Potton Equilibrium”
From Evolutionary to Evaluationary Game Theory
 Idea: theoretical analysis
 Established Framework:
Evolutionary Game Theory  “survival of the fittest”
 Problem: Nash equilibria exist, but asymmetry between
applying project consortia and evaluation body (a.k.a. EU)
 Idea: use “Nash-Potton Equilibrium”
 original approach:
informatics and religion
(catholics/protestants vs GOD)
From Evolutionary to Evaluationary Game Theory
 Idea: theoretical analysis
 Established Framework:
Evolutionary Game Theory  “survival of the fittest”
 Problem: Nash equilibria exist, but asymmetry between
applying project consortia and evaluation body (a.k.a. EU)
 Idea: use “Nash-Potton Equilibrium”
 original approach:
informatics and religion
(catholics/protestants vs GOD)
 here: project consortia vs. EU
From Evolutionary to Evaluationary Game Theory
 Idea: theoretical analysis
 Established Framework:
Evolutionary Game Theory  “survival of the fittest”
 Problem: Nash equilibria exist, but asymmetry between
applying project consortia and evaluation body (a.k.a. EU)
 Idea: use “Nash-Potton Equilibrium”
 original approach:
informatics and religion
(catholics/protestants vs GOD)
 here: project consortia vs. EU
 Result: Evaluationary Game Theory – Management-by-Acronym (MBA)
Fundamental Metric
 First idea: Evaluation result E proportional to acronym matching ratio m
Fundamental Metric
 First idea: Evaluation result E proportional to acronym matching ratio m

E = m  c2
Fundamental Metric
 First idea: Evaluation result E proportional to acronym matching ratio m

E = m  c2
 Problem 1: formula already invented and heavily used
 Problem 2: no Greek letters ( scientific value???)
Fundamental Metric
 First idea: Evaluation result E proportional to acronym matching ratio m

E = m  c2
 Problem 1: formula already invented and heavily used
 Problem 2: no Greek letters ( scientific value???)
 Much better: queuing theoretic approach
• r = /m = degree of efficiency [%]
• n = funding arrival rate ( 1/runtime, Little’s Law!)
• ACK = “Formalprüfungsindikator” (FBI, = 1 if proposal has been
properly submitted in time & space etc.)
Fundamental Metric
 First idea: Evaluation result E proportional to acronym matching ratio m

E = m  c2
 Problem 1: formula already invented and heavily used
 Problem 2: no Greek letters ( scientific value???)
 Much better: queuing theoretic approach
• r = /m = degree of efficiency [%]
• n = funding arrival rate ( 1/runtime, Little’s Law!)
• ACK = “Formalprüfungsindikator” (FBI, = 1 if proposal has been
properly submitted in time & space etc.)
 Resulting metric:
E = ACK  r  n  m
Empirical Validation
27
Empirical Validation
r=25%
m=1
r=40%
m<1
28
r=24%
m1
r=37.5%
m=1
Simplifying assumption:
constant funding arrival rate n
Empirical Validation
r=25%
m=1
r=40%
m<1
r=37.5%
m=1
Simplifying assumption:
constant funding arrival rate n
r=24%
m1
 Conclusion: no need to worry any more about project content etc.
29
Empirical Validation
r=25%
m=1
r=40%
m<1
r=37.5%
m=1
r=24%
m1
 Conclusion: no need to worry any more about project content etc.
 Open issue: limiting behaviour, e.g. what about other drugs?
30
Empirical Validation
r=25%
m=1
r=40%
m<1
r=37.5%
m=1
r=24%
m1
 Conclusion: no need to worry any more about project content etc.
 Open issue: limiting behaviour, e.g. what about other drugs?
 MARIUANA (Monitoring, Analysis and Reporting of Intrusive,
© Fabio Ricciato
Unwanted and Anomalous Network Activities)???
31
Empirical Validation
r=25%
m=1
r=40%
m<1
r=37.5%
m=1
r=24%
m1
 Conclusion: no need to worry any more about project content etc.
 Open issue: limiting behaviour, e.g. what about other drugs?
 MARIUANA (Monitoring, Analysis and Reporting of Intrusive,
© Fabio Ricciato
Unwanted and Anomalous Network Activities)???
© Toni Fuhrmann
32 ViaGra (Via donau Graphics)???
Summary and Conclusions
Summary and Conclusions
Ode an das Akronym
Summary and Conclusions
Ode an das Akronym
Wie herrlich ist es, ein Projekt zu planen,
sodann zu leiten auf kühne Bahnen,
Summary and Conclusions
Ode an das Akronym
Wie herrlich ist es, ein Projekt zu planen,
sodann zu leiten auf kühne Bahnen,
sich nie zu kümmern um Budget und Zahlen,
stattdessen visionär voranzustrahlen.
Summary and Conclusions
Ode an das Akronym
Wie herrlich ist es, ein Projekt zu planen,
sodann zu leiten auf kühne Bahnen,
sich nie zu kümmern um Budget und Zahlen,
stattdessen visionär voranzustrahlen.
O Kohle mio, du kommst daher
von ganz alleine und ohn‘ Beschwer;
o Kohle, o Kohle mio,
nur Akronyme, es braucht nicht mehr.
BACCARDI, CHIANTI, WISQY, CAIPIRINA,
und auch CAMPARI brachten viel Gewinn, ja!
BACCARDI, CHIANTI, WISQY, CAIPIRINA,
und auch CAMPARI brachten viel Gewinn, ja!
MARIUANA…? Man muss nicht viel wagen,
um umso mehr Budget davonzutragen.
BACCARDI, CHIANTI, WISQY, CAIPIRINA,
und auch CAMPARI brachten viel Gewinn, ja!
MARIUANA…? Man muss nicht viel wagen,
um umso mehr Budget davonzutragen.
O Kohle mio, vorbei der Stress
der Antragsstellung: wie man bemess‘
die Kohle für das Projekt,
das man so mühsam da ausgeheckt.
BACCARDI, CHIANTI, WISQY, CAIPIRINA,
und auch CAMPARI brachten viel Gewinn, ja!
MARIUANA…? Man muss nicht viel wagen,
um umso mehr Budget davonzutragen.
O Kohle mio, vorbei der Stress
der Antragsstellung: wie man bemess‘
die Kohle für das Projekt,
das man so mühsam da ausgeheckt.
O Kohle mio, du kommst daher
von ganz alleine, ’s ist nicht so schwer;
o Kohle, o Kohle mio:
die Akronyme, die helfen sehr.
The End
Univ.-Prof. Dr. Peter Reichl
Full Professor and Head of Research Group
“Cooperative Systems (COSY)”
Faculty of Informatics
University of Vienna
Visiting address:
Währinger Strasse 29 / 3.38
1090 Vienna
[email protected]
Thanks for your attention!
Any feedback welcome!