s ystems analysis laboratory helsinki university of technology 1 raimo p. hämäläinen systems...

81
1 S ystems Analysis Laboratory Helsinki University of Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology www.raimo.hut.fi JMCDA, Vol. 12 , No. 2-3, 2003, pp. 101-110. Aiding Decisions, Negotiating and Collecting Opinions on the Web www.decisionarium.hut .fi D E C I S I O N A R I U M v. 3.2006

Upload: kelley-chandler

Post on 25-Dec-2015

220 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

1

S ystemsAnalysis LaboratoryHelsinki University of Technology

Raimo P. HämäläinenSystems Analysis Laboratory

Helsinki University of Technologywww.raimo.hut.fi

JMCDA, Vol. 12 , No. 2-3, 2003, pp. 101-110.

Aiding Decisions, Negotiating and Collecting Opinions on the Web

www.decisionarium.hut.fi

D E C I S I O N A R I U M

v. 3.2006

Page 2: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

2

S ystemsAnalysis LaboratoryHelsinki University of Technology

selected publications J. Mustajoki, R.P. Hämäläinen and A. Salo: Decision support by interval SMART/SWING – Incorporating

imprecision in the SMART and SWING methods, Decision Sciences, 2005.H. Ehtamo, R.P. Hämäläinen and V. Koskinen: An e-learning module on negotiation analysis, Proc. of HICSS-37, 2004.

J. Mustajoki and R.P. Hämäläinen, Making the even swaps method even easier, Manuscript, 2004. R.P. Hämäläinen, Decisionarium - Aiding decisions, negotiating and collecting opinions on the Web, J. Multi-Crit. Dec. Anal., 2003.

H. Ehtamo, E. Kettunen and R.P. Hämäläinen: Searching for joint gains in multi-party negotiations, Eur. J. Oper. Res., 2001. J. Gustafsson, A. Salo and T. Gustafsson: PRIME Decisions - An interactive tool for value tree

analysis, Lecture Notes in Economics and Mathematical Systems, 2001.J. Mustajoki and R.P. Hämäläinen: Web-HIPRE - Global decision support by value tree and AHP analysis, INFOR, 2000.

D E C I S I O N A R I U M

PRIME DecisionsWINPRE

web-sites www.decisionarium.hut.fi www.dm.hut.fi

www.hipre.hut.fi www.jointgains.hut.fi www.opinions.hut.fi www.smart-swaps.hut.fi www.rich.hut.fiPRIME Decisions and WINPRE downloadable at www.sal.hut.fi/Downloadables

Web-HIPREvalue tree and AHP based decision support

Smart-Swaps

Opinions-Online platform for global participation, voting, surveys, and group decisions

Joint Gains

groupcollaboration decision

making

computer support

CSCW

multicriteriadecision analysis

internet

groupdecision making

GDSS, NSS

DSS

multi-party negotiation support with the method of improving directions

Windows software for decision analysis with imprecise ratio statements

g l o b a l s p a c e f o r d e c i s i o n s u p p o r t

elimination of criteria and alternatives by even swaps

preference programming, PAIRS

Updated 25.10.2004

SystemsAnalysis Laboratory

RICH Decisionsrank inclusion in criteria

hierarchies

Page 3: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

3

S ystemsAnalysis LaboratoryHelsinki University of Technology

Mission of Decisionarium

Provide resources for decision and negotiation support and advance the real and correct use of MCDA

History: HIPRE 3+ in 1992 MAVT/AHP for DOS systems

Today: e-learning modules provide help to learn the methods and global access to the software also for non OR/MS people

Page 4: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

4

S ystemsAnalysis LaboratoryHelsinki University of Technology

Opinions-Online (www.opinions.hut.fi)• Platform for global participation, voting, surveys, and group

decisions

Web-HIPRE (www.hipre.hut.fi)• Value tree based decision analysis and support

WINPRE and PRIME Decisions (for Windows)

• Interval AHP, interval SMART/SWING and PRIME methods

RICH Decisions (www.rich.hut.fi)• Preference programming in MAVT

Smart-Swaps (www.smart-swaps.hut.fi)• Multicriteria decision support with the even swaps method

Joint Gains (www.jointgains.hut.fi)• Negotiation support with the method of improving directions

Page 5: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

5

S ystemsAnalysis LaboratoryHelsinki University of Technology

• Possibility to compare different weighting and rating methods

• AHP/MAVT and different scales

• Preference programming in MAVT and in the Even Swaps procedure

• Jointly improving direction method for negotiations

New Methodological Features

Page 6: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

6

S ystemsAnalysis LaboratoryHelsinki University of Technology

SAL eLearning sites:

Multiple Criteria Decision Analysis www.mcda.hut.fiDecision Making Under UncertaintyNegotiation Analysis www.negotiation.hut.fi

eLearning Decision Makingwww.dm.hut.fi

Page 7: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

S ystemsAnalysis LaboratoryHelsinki University of Technology

Opinions-Online

Platform for Global Participation, Voting, Surveys and Group Decisions

Design: Raimo P. Hämäläinen

Programming: Reijo Kalenius

www.opinions.hut.fiwww.opinions-online.com

Systems Analysis LaboratoryHelsinki University of Technology

http://www.sal.hut.fi

Page 8: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

8

S ystemsAnalysis LaboratoryHelsinki University of Technology

Surveys on the web

• Fast, easy and cheap

• Hyperlinks to background information

• Easy access to results

• Results can be analyzed on-line

• Access control: registration, e-mail list, domain, password

Page 9: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

9

S ystemsAnalysis LaboratoryHelsinki University of Technology

Creating a new session

• Browser-based generation of new sessions

• Fast and simple

• Templates available

Page 10: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

10

S ystemsAnalysis LaboratoryHelsinki University of Technology

Possible questions

• Survey section Multiple/single choice• Best/worst• Ranking• Rating• Approval voting• Written comments

Page 11: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

11

S ystemsAnalysis LaboratoryHelsinki University of Technology

Viewing the results

• In real-time• By selected fields• Questionwise public

or restricted access• Barometer• Direct links

to results

Page 12: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

12

S ystemsAnalysis LaboratoryHelsinki University of Technology

Approval voting

• The user is asked to pick the alternatives that he/she can approve

• Often better than a simple “choose best” question when trying to reach a consensus

Page 13: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

13

S ystemsAnalysis LaboratoryHelsinki University of Technology

Advanced voting ruleswww.opinion.vote.hut.fi

• Condorcet criteria– Copeland’s methods, Dodgson’s method,

Maximin method

• Borda count– Nanson’s method, University method

• Black’s method• Plurality voting

– Coombs’ method, Hare system, Bishop method

Page 14: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

14

S ystemsAnalysis LaboratoryHelsinki University of Technology

Examples of use

• Teledemocracy – interactive citizens’ participation

• Group decision making

• Brainstorming

• Course evaluation in universities and schools

• Marketing research

• Organisational surveys and barometers

Page 15: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

S ystemsAnalysis LaboratoryHelsinki University of Technology

Global Multicriteria Decision Support by Web-HIPRE

A Java-applet for Value Tree and AHP Analysis

Raimo P. Hämäläinen

Jyri Mustajoki

www.hipre.hut.fi

Systems Analysis LaboratoryHelsinki University of Technology

http://www.sal.hut.fi

Page 16: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

16

S ystemsAnalysis LaboratoryHelsinki University of Technology

Web-HIPRE links can refer to any web-pages

Page 17: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

17

S ystemsAnalysis LaboratoryHelsinki University of Technology

Note: Weights in this example are her personal opinions

Direct Weighting

Page 18: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

18

S ystemsAnalysis LaboratoryHelsinki University of Technology

• SMARTER uses rankings only

SWING,SMART and SMARTER Methods

Page 19: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

19

S ystemsAnalysis LaboratoryHelsinki University of Technology

• Continuous scale 1-9

• Numerical, verbal or graphical approach

Pairwise Comparison - AHP

Page 20: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

20

S ystemsAnalysis LaboratoryHelsinki University of Technology

• Ratings of alternatives shown

• Any shape of the value function allowed

Value Function

Page 21: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

21

S ystemsAnalysis LaboratoryHelsinki University of Technology

• Bar graphs or numerical values

• Bars divided by the contribution of each criterion

Composite Priorities

Page 22: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

22

S ystemsAnalysis LaboratoryHelsinki University of Technology

• Group model is the weighted sum of individual decision makers’ composite priorities for the alternatives

Group Decision Support

Page 23: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

23

S ystemsAnalysis LaboratoryHelsinki University of Technology

• Individual value trees can be different

• Composite priorities of each group member

- obtained from their individual models

- shown in the definition phase

Defining Group Members

Page 24: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

24

S ystemsAnalysis LaboratoryHelsinki University of Technology

• Contribution of each group member indicated by segments

Aggregate Group Priorities

Page 25: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

25

S ystemsAnalysis LaboratoryHelsinki University of Technology

• Changes in the relative importance of decision makers can be analyzed

Sensitivity analysis

Page 26: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

26

S ystemsAnalysis LaboratoryHelsinki University of Technology

Future challenges

Web makes MCDA tools available to everybody - Should everybody use them?

It is the responsibility of the multicriteria decision analysis community to:

• Learn and teach the use different weighting methods

• Focus on the praxis and avoidance of behavioural biases

• Develop and identify “best practice” procedures

Page 27: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

27

S ystemsAnalysis LaboratoryHelsinki University of Technology

Number-of-attribute-levels effect in

conjoint analysis

Splitting bias withweighting methodsbased on ranking

Rank reversal inAHP

Averages over agroup yield even

weights

Normalization

Decisionmakers onlygive ordinalinformation

Division ofattributes changes

weightsRange effect

Hierarchicalweighting leads to

steeper weights

Weighting methodsyield different weights

Sources of biases and problems

Page 28: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

28

S ystemsAnalysis LaboratoryHelsinki University of Technology

Visits to Web-HIPRE

1617

53036826

9233

1315114216

16526

1998 1999 2000 2001 2002 2003 2004

Page 29: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

29

S ystemsAnalysis LaboratoryHelsinki University of Technology

Visitors’ top-level domains

Domain Visits Domain Visits Domain Visitsfi 14174 (36.8 %) ch 385 (1.0 %) sg 124 (0.3 %)

com 4758 (12.3 %) hu 348 (0.9 %) be 119 (0.3 %)net 4677 (12.1 %) jp 321 (0.8 %) is 112 (0.3 %)edu 2729 (7.1 %) br 284 (0.7 %) ru 97 (0.3 %)si 1606 (4.2 %) tr 251 (0.7 %) gov 92 (0.2 %)uk 981 (2.5 %) it 198 (0.5 %) mx 87 (0.2 %)ca 894 (2.3 %) pl 187 (0.5 %) il 82 (0.2 %)ee 849 (2.2 %) gr 171 (0.4 %) org 77 (0.2 %)es 834 (2.2 %) fr 165 (0.4 %) no 72 (0.2 %)nl 774 (2.0 %) pt 147 (0.4 %) za 65 (0.2 %)de 744 (1.9 %) mil 144 (0.4 %) ar 63 (0.2 %)au 571 (1.5 %) se 141 (0.4 %)at 449 (1.2 %) tw 132 (0.3 %)

Total 38557 visits (+ 28315 visits with only IP)

Page 30: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

30

S ystemsAnalysis LaboratoryHelsinki University of Technology

Visitors’ first-level domainsDomain Domain Domainhkkk.fi 4593 (11.9 %) aol.com 343 (0.9 %) bke.hu 181 (0.5 %)hut.fi 4453 (11.5 %) ac.at 324 (0.8 %) comcast.net 181 (0.5 %)uni-mb.si 1220 (3.2 %) edu.au 323 (0.8 %) carleton.ca 172 (0.4 %)duke.edu 1069 (2.8 %) siol.net 290 (0.8 %) te-keskus.fi 172 (0.4 %)ac.uk 776 (2.0 %) arnes.si 287 (0.7 %) uwaterloo.ca 160 (0.4 %)htv.fi 763 (2.0 %) omakaista.fi 262 (0.7 %) net.br 157 (0.4 %)inktomi.com 689 (1.8 %) abo.fi 215 (0.6 %) co.uk 156 (0.4 %)inet.fi 629 (1.6 %) estpak.ee 215 (0.6 %) edu.tr 151 (0.4 %)googlebot.com 608 (1.6 %) asu.edu 206 (0.5 %) rr.com 151 (0.4 %)helsinki.fi 497 (1.3 %) knology.net 205 (0.5 %) sympatico.ca 149 (0.4 %)ja.net 464 (1.2 %) uab.es 205 (0.5 %) ne.jp 142 (0.4 %)kolumbus.fi 417 (1.1 %) verizon.net 203 (0.5 %)t-dialin.net 376 (1.0 %) cesca.es 193 (0.5 %) (+ 28315 visits with only IP)

Total 38557 visits

Visits Visits Visits

Page 31: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

31

S ystemsAnalysis LaboratoryHelsinki University of Technology

Visits through sites linking to Web-HIPRESite Sitehut.fi 7375 (40.4 %) google.fi 190 (1.0 %)

duke.edu 1480 (8.1 %) man.ac.uk 174 (1.0 %)

google.com 1349 (7.4 %) unige.ch 165 (0.9 %)

dicksmart.net 589 (3.2 %) msn.com 162 (0.9 %)

cmu.edu 568 (3.1 %) colorado.edu 144 (0.8 %)

carleton.ca 527 (2.9 %) helsinki.fi 143 (0.8 %)

uni-mb.si 403 (2.2 %) clarolineserver.com 122 (0.7 %)

ttu.ee 383 (2.1 %) urjc.es 118 (0.6 %)

clarku.edu 324 (1.8 %) univie.ac.at 115 (0.6 %)

yahoo.com 272 (1.5 %) gov.uk 100 (0.5 %)

altavista.com 229 (1.3 %) ecohotelsbolivia.com 97 (0.5 %)

100gen.fi 222 (1.2 %) nus.edu.sg 96 (0.5 %)

informs.org 218 (1.2 %)

Visits Visits

Total 18261 visits

Page 32: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

32

S ystemsAnalysis LaboratoryHelsinki University of Technology

Literature

Mustajoki, J. and Hämäläinen, R.P.: Web-HIPRE: Global decision support by value tree and AHP analysis, INFOR, Vol. 38, No. 3, 2000, pp. 208-220.

Hämäläinen, R.P.: Reversing the perspective on the applications of decision analysis, Decision Analysis, Vol. 1, No. 1, pp. 26-31.

Mustajoki, J., Hämäläinen, R.P. and Marttunen, M.: Participatory multicriteria decision support with Web-HIPRE: A case of lake regulation policy. Environmental Modelling & Software, Vol. 19, No. 6, 2004, pp. 537-547.

Pöyhönen, M. and Hämäläinen, R.P.: There is hope in attribute weighting, INFOR, Vol. 38, No. 3, 2000, pp. 272-282.

Pöyhönen, M. and Hämäläinen, R.P.: On the Convergence of Multiattribute Weighting Methods, European Journal of Operational Research, Vol. 129, No. 3, 2001, pp. 569-585.

Pöyhönen, M., Vrolijk, H.C.J. and Hämäläinen, R.P.: Behavioral and Procedural Consequences of Structural Variation in Value Trees, European Journal of Operational Research, Vol. 134, No. 1, 2001, pp. 218-227.

Page 33: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

33

S ystemsAnalysis LaboratoryHelsinki University of Technology

New Theory: Preference programming

Analysis with incomplete preference statements (intervals):

”...attribute is at least 2 times as but no more than 3 times as important as...”

Windows software• WINPRE – Workbench for Interactive Preference Programming Interval AHP, interval SMART/SWING and PAIRS• PRIME-Preference Ratios in Multiattribute Evaluation Method Incomplete preference statements Web software• RICH Decisions – Rank Inclusion in Criteria Hierarchies

Page 34: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

34

S ystemsAnalysis LaboratoryHelsinki University of Technology

Preference Programming – The PAIRS method

• Imprecise statements with intervals on– Attribute weight ratios (e.g. 1/2 w1 / w2 3)

Feasible region for the weights

– Alternatives’ ratings (e.g. 0.6 v1(x1) 0.8)

Intervals for the overall values– Lower bound for the overall value of x:

– Upper bound correspondingly

n

iiii xvwxv

1

)(min)(

Page 35: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

35

S ystemsAnalysis LaboratoryHelsinki University of Technology

2

1

2

61

31

21

C

B

C

A

B

A

w

w

w

ww

w

Interval statements define a feasible region S for the weights

Page 36: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

36

S ystemsAnalysis LaboratoryHelsinki University of Technology

Uses of interval models

New generalized AHP and SMART/SWING methods

DM can also reply with intervals instead of exact point estimates – a new way to accommodate uncertainty

Interval sensitivity analysis

Variations allowed in several model parameters simultaneously - worst case analysis

Group decision making

All members´ opinions embedded in intervals = a joint common group model

Page 37: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

37

S ystemsAnalysis LaboratoryHelsinki University of Technology

Interval SMART/SWING• A as reference - A given 10 points

• Point intervals given to the other attributes:– 5-20 points to attribute B– 10-30 points to attribute C

• Weight ratio between B and C not explicitly given by the DM

Page 38: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

38

S ystemsAnalysis LaboratoryHelsinki University of Technology

WINPRE Software

Page 39: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

39

S ystemsAnalysis LaboratoryHelsinki University of Technology

PRIME Decisions Software

Page 40: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

40

S ystemsAnalysis LaboratoryHelsinki University of Technology

Literature – Methodology

Salo, A. and Hämäläinen, R.P.: Preference assessment by imprecise ratio statements, Operations Research, Vol. 40, No. 6, 1992, pp. 1053-1061.

Salo, A. and Hämäläinen, R.P.: Preference programming through approximate ratio comparisons, European Journal of Operational Research, Vol. 82, No. 3, 1995, pp. 458-475.

Salo, A. and Hämäläinen, R.P.: Preference ratios in multiattribute evaluation (PRIME) – Elicitation and decision procedures under incomplete information, IEEE Transactions on Systems, Man and Cybernetics – Part A: Systems and Humans, Vol. 31, No. 6, 2001, pp. 533-545.

Salo, A. and Hämäläinen, R.P.: Preference Programming. (Manuscript) Downloadable at http://www.sal.hut.fi/Publications/pdf-files/msal03b.pdf

Mustajoki, J., Hämäläinen, R.P. and Salo, A.: Decision Support by Interval SMART/SWING - Incorporating Imprecision in the SMART and SWING Methods, Decision Sciences, Vol. 36, No.2, 2005, pp. 317-339.

Page 41: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

41

S ystemsAnalysis LaboratoryHelsinki University of Technology

Literature – Tools and applicationsGustafsson, J., Salo, A. and Gustafsson, T.: PRIME Decisions - An Interactive

Tool for Value Tree Analysis, Lecture Notes in Economics and Mathematical Systems, M. Köksalan and S. Zionts (eds.), 507, 2001, pp. 165-176.

Hämäläinen, R.P., Salo, A. and Pöysti, K.: Observations about consensus seeking in a multiple criteria environment, Proc. of the Twenty-Fifth Hawaii International Conference on Systems Sciences, Hawaii, Vol. IV, January 1992, pp. 190-198.

Hämäläinen, R.P. and Pöyhönen, M.: On-line group decision support by preference programming in traffic planning, Group Decision and Negotiation, Vol. 5, 1996, pp. 485-500.

Liesiö, J., Mild, P. and Salo, A.: Preference Programming for Robust Portfolio Modeling and Project Selection, European Journal of Operational Research (to appear)

Mustajoki, J., Hämäläinen, R.P. and Lindstedt, M.R.K.: Using intervals for Global Sensitivity and Worst Case Analyses in Multiattribute Value Trees, European Journal of Operational Research. (to appear)

Page 42: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

S ystemsAnalysis LaboratoryHelsinki University of Technology

RICH Decisions

www.rich.hut.fi

Design: Ahti Salo and Antti Punkka

Programming: Juuso Liesiö

Systems Analysis LaboratoryHelsinki University of Technology

http://www.sal.hut.fi

Page 43: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

43

S ystemsAnalysis LaboratoryHelsinki University of Technology

The RICH Method

Based on:

Incomplete ordinal information about the relative importance of attributes

• ”environmental aspects belongs to the three most important attributes” or

• ”either cost or environmental aspects is the most important attribute”

Page 44: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

44

S ystemsAnalysis LaboratoryHelsinki University of Technology

Score Elicitation

• Upper and lower bounds for the scores

• Type or use the scroll bar

Page 45: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

45

S ystemsAnalysis LaboratoryHelsinki University of Technology

The user specifies sets of attributes and corresponding sets of rankings.

Here attributes distance to harbour and distance to office are the two most important ones.

The table displays the possible rankings.

Weight Elicitation

Page 46: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

46

S ystemsAnalysis LaboratoryHelsinki University of Technology

Dominance Structure and Decision Rules

Page 47: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

47

S ystemsAnalysis LaboratoryHelsinki University of Technology

Literature

Salo, A. and Punkka, A.: Rank Inclusion in Criteria Hierarchies, European Journal of Operational Research, Vol. 163, No. 2, 2005, pp. 338-356.

Salo, A. and Hämäläinen, R.P.: Preference ratios in multiattribute evaluation (PRIME) – Elicitation and decision procedures under incomplete information, IEEE Transactions on Systems, Man and Cybernetics – Part A: Systems and Humans, Vol. 31, No. 6, 2001, pp. 533-545.

Salo A. and Hämäläinen, R.P.: Preference Programming. (manuscript)

Ojanen, O., Makkonen, S. and Salo, A.: A Multi-Criteria Framework for the Selection of Risk Analysis Methods at Energy Utilities. International Journal of Risk Assessment and Management, Vol. 5, No. 1, 2005, pp. 16-35.

Punkka, A. and Salo, A.: RICHER: Preference Programming with Incomplete Ordinal Information. (submitted manuscript)

Salo, A. and Liesiö, J.: A Case Study in Participatory Priority-Setting for a Scandinavian Research Program, International Journal of Information Technology & Decision Making. (to appear)

Page 48: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

S ystemsAnalysis LaboratoryHelsinki University of Technology

Smart-Swaps

Smart Choices with the Even Swaps Method

Design: Raimo P. Hämäläinen and Jyri MustajokiProgramming: Pauli Alanaatu

www.smart-swaps.hut.fi

Systems Analysis LaboratoryHelsinki University of Technology

http://www.sal.hut.fi

Page 49: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

49

S ystemsAnalysis LaboratoryHelsinki University of Technology

Smart Choices

• An iterative process to support multicriteria decision making

• Uses the even swaps method to make trade-offs

(Harvard Business School Press, Boston, MA, 1999)

Page 50: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

50

S ystemsAnalysis LaboratoryHelsinki University of Technology

Even Swaps

• Carry out even swaps that makeAlternatives dominated (attribute-wise)

• There is another alternative, which is equal or better than this in every attribute, and better at least in one attribute

Attributes irrelevant• Each alternative has the same value on this attribute

These can be eliminated

• Process continues until one alternative, i.e. the best one, remains

Page 51: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

51

S ystemsAnalysis LaboratoryHelsinki University of Technology

Supporting Even Swaps with Preference Programming

• Even Swaps process carried out as usual

• The DM’s preferences simultaneously modeled with Preference Programming– Intervals allow us to deal with incomplete

information – Trade-off information given in the even swaps can

be used to update the model

Suggestions for the Even Swaps process

Page 52: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

52

S ystemsAnalysis LaboratoryHelsinki University of Technology

Decision support

Problem initialization

Updating of

the model

Make an even swap

Even Swaps Preference Programming

Practical dominance candidates

Initial statements about the attributes

Eliminate irrelevant attributes

Eliminate dominated alternatives

Even swap suggestions

More than oneremaining alternative

Yes

The most preferred alternative is found

No

Trade-off information

Page 53: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

53

S ystemsAnalysis LaboratoryHelsinki University of Technology

• Identification of practical dominances

• Suggestions for the next even swap to be made

• Additional supportInformation about what can be achieved with each

swap

Notification of dominances

Rankings indicated by colours

Process history allows backtracking

Smart-Swaps

Page 54: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

54

S ystemsAnalysis LaboratoryHelsinki University of Technology

Example• Office selection problem (Hammond et al. 1999)

Dominatedby

Lombard

Practicallydominated

byMontana

(Slightly better in Monthly Cost, but equal or worse in all other attributes)

78

25

An even swap

Commute time removed as irrelevant

Page 55: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

55

S ystemsAnalysis LaboratoryHelsinki University of Technology

Problem definition

Page 56: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

56

S ystemsAnalysis LaboratoryHelsinki University of Technology

Entering trade-offs

Page 57: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

57

S ystemsAnalysis LaboratoryHelsinki University of Technology

Process history

Page 58: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

58

S ystemsAnalysis LaboratoryHelsinki University of Technology

Hammond, J.S., Keeney, R.L., Raiffa, H., 1998. Even swaps: A rational method for making trade-offs, Harvard Business Review, 76(2), 137-149.

Hammond, J.S., Keeney, R.L., Raiffa, H., 1999. Smart choices. A practical guide to making better decisions, Harvard Business School Press, Boston.

Mustajoki, J. Hämäläinen, R.P., 2005. A Preference Programming Approach to Make the Even Swaps Method Even Easier, Decision Analysis, 2(2), 110-123.

Salo, A., Hämäläinen, R.P., 1992. Preference assessment by imprecise ratio statements, Operations Research, 40(6), 1053-1061.

Applications of Even Swaps:

Gregory, R., Wellman, K., 2001. Bringing stakeholder values into environmental policy choices: a community-based estuary case study, Ecological Economics, 39, 37-52.

Kajanus, M., Ahola, J., Kurttila, M., Pesonen, M., 2001. Application of even swaps for strategy selection in a rural enterprise, Management Decision, 39(5), 394-402.

Literature

Page 59: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

S ystemsAnalysis LaboratoryHelsinki University of Technology

Joint-Gains

Negotiation Support in the Internet

Eero Kettunen, Raimo P. Hämäläinen

and Harri Ehtamo

www.jointgains.hut.fi

Systems Analysis LaboratoryHelsinki University of Technology

http://www.sal.hut.fi

Page 60: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

60

S ystemsAnalysis LaboratoryHelsinki University of Technology

Method of Improving DirectionsEhtamo, Kettunen, and

Hämäläinen (2002)

• Interactive method for reaching efficient alternatives

• Search of joint gains from a given initial alternative

• In the mediation process participants are given simple comparison tasks:

“Which one of these two alternatives do you prefer, alternative A or B?”

Efficient frontier

..

..

Utility of DM 1

Utilit

y of

DM

2

Page 61: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

61

S ystemsAnalysis LaboratoryHelsinki University of Technology

Mediation Process Tasks in Preference Identification

• Initial alternative considered as “current alternative”

• Task 1 for identifying participants’

most preferred directions

• Joint Gains calculates a jointly improving direction

• Task 2 for identifying participants’ most preferred

alternatives in the jointly improving direction

series of pairwise series of pairwise comparisonscomparisons

series of pairwise comparisonsseries of pairwise comparisons

Page 62: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

62

S ystemsAnalysis LaboratoryHelsinki University of Technology

Joint Gains Negotiation

• User can create his own case

• 2 to N participants (negotiating parties,

DM’s)

• 2 to M continuous decision variables

• Linear inequality constraints

• Participants distributed in the web

Page 63: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

63

S ystemsAnalysis LaboratoryHelsinki University of Technology

DM’s Utility Functions

• DM’s reply holistically

• No explicit assessment of utility functions

• Joint Gains only calls for local preference

information

• Post-settlement setting in the neighbourhood

of the current alternative

• Joint Gains allows learning and change of

preferences during the process

Page 64: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

64

S ystemsAnalysis LaboratoryHelsinki University of Technology

• Two participants

buyer and seller• Three decision variables

unit price ($): 10..50

amount (lb): 1..1000

delivery (days): 1..30• Delivery constraint (figure):

999*delivery - 29*amount 970• Initial agreement: 30 $, 100 lb, 25 days

amount (lb)1 1000

1

deliv

ery

(day

s)

30

Case example: Business

Page 65: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

65

S ystemsAnalysis LaboratoryHelsinki University of Technology

Creating a case: Criteria to provide optional decision aiding

Page 66: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

66

S ystemsAnalysis LaboratoryHelsinki University of Technology

Sessions

• Sessions produce efficient alternatives

• Case administrator can start new sessions on-line and define new initial starting points

• Sessions can be parallel• Each session has an independent

mediation process

Session 1

Session 2

Session 3

Joint Gains - Business

Session n

...

• Participants take part in sessions within the case

efficient point

efficient point

efficient point

efficient point

Page 67: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

67

S ystemsAnalysis LaboratoryHelsinki University of Technology

Preference identification task 2

Not started

Preference identification task 1

JOINT GAIN?

Stopped

New comparison task is given after all participants have completed the first one

Page 68: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

68

S ystemsAnalysis LaboratoryHelsinki University of Technology

Session view - joint gains after two steps

unit_price

10

20

30

1 2 3

amount

406080

100

1 2 3

delivery

10

20

30

1 2 3

Page 69: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

69

S ystemsAnalysis LaboratoryHelsinki University of Technology

Literature

Ehtamo, H., M. Verkama, and R.P. Hämäläinen (1999). How to select Fair Improving Directions in a negotiation Model over Continuous Issues, IEEE Trans. On Syst., Man, and Cybern. – Part C, Vol. 29, No. 1, pp. 26-33.

Ehtamo, H., E. Kettunen, and R. P. Hämäläinen (2001). Searching for Joint Gains in Multi-Party Negotiations, European Journal of Operational Research, Vol. 130, No. 1, pp. 54-69.

Hämäläinen, H., E. Kettunen, M. Marttunen, and H. Ehtamo (2001). Evaluating a Framework for Multi-Stakeholder Decision Support in Water Resources Management, Group Decision and Negotiation, Vol. 10, No. 4, pp. 331-353.

Ehtamo, H., R.P. Hämäläinen, and V. Koskinen (2004). An E-learning Module on Negotiation Analysis, Proc. of the Hawaii International Conference on System Sciences, IEEE Computer Society Press, Hawaii, January 5-8.

Page 70: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

S ystemsAnalysis LaboratoryHelsinki University of Technology

eLearning Decision Makingwww.mcda.hut.fi

eLearning sites on:Multiple Criteria Decision Analysis

Decision Making Under Uncertainty Negotiation Analysis

Prof. Raimo P. HämäläinenSystems Analysis Laboratory

Helsinki University of Technologyhttp://www.sal.hut.fi

Page 71: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

71

S ystemsAnalysis LaboratoryHelsinki University of Technology

eLearning sitesMaterial:• Theory sections, interactive computer assignments• Animations and video clips, online quizzes, theory assignments

Decisionarium software:• Web-HIPRE, PRIME Decisions, Opinions-Online.vote, and Joint Gains, video clips help the use

eLearning modules: • 4 - 6 hours study time• Instructors can create their own modules using the material and software• Academic non-profit use is free

Page 72: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

72

S ystemsAnalysis LaboratoryHelsinki University of Technology

Page 73: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

73

S ystemsAnalysis LaboratoryHelsinki University of Technology

Learning paths and modules

Learning path: guided route through the learning material Learning module: represents 2-4 h of traditional lectures and exercises

AssignmentsTheory VideosCases QuizzesLearningPaths Evaluation

Introduction to Value Tree AnalysisIntroduction to Value Tree Analysis

Module 3Module 3

Module 2Module 2

Page 74: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

74

S ystemsAnalysis LaboratoryHelsinki University of Technology

Learning modules

Theory• HTML pages

Theory• HTML pages

• motivation, detailed instructions, 2 to 4 hour sessions

Case• slide shows

• video clips

Case• slide shows

• video clips

Assignments• online quizzes

• software tasks

• report templates

Assignments• online quizzes

• software tasks

• report templates

Evaluation • Opinions Online

Evaluation • Opinions Online

Web software

• Web-HIPRE

• video clips

Web software

• Web-HIPRE

• video clips

AssignmentsTheory VideosCases Quizzes

LearningPaths Evaluation

Introduction to Value Tree Analysis

Module 3

Module 2Module 2

Page 75: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

75

S ystemsAnalysis LaboratoryHelsinki University of Technology

Evaluation

Cases

AssignmentsTheory

Intro

Theoreticalfoundations

Problemstructuring

Preferenceelicitation

Family selecting a car

Job selection case• basics of value tree analysis• how to use Web-HIPRE

Car selection case• imprecise preference statements, interval value trees• basics of Prime Decisions software

Family selecting a car• group decision-making with Web-HIPRE• weighted arithmetic mean method

Job selection case• basics of value tree analysis• how to use Web-HIPRE

Car selection case• imprecise preference statements, interval value trees• basics of Prime Decisions software

Family selecting a car• group decision-making with Web-HIPRE• weighted arithmetic mean method

Page 76: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

76

S ystemsAnalysis LaboratoryHelsinki University of Technology

Video clips

VideosWorking with Web-HIPREStructuring a value tree Entering consequences of ...Assessing the form of value...Direct rating SMARTSMARTSWINGAHPViewing the resultsSensitivity analysisGroup decision makingPRIME method

AssignmentsTheory Cases QuizzesLearningPaths

Videos

• Recorded software use with voice explanations (1-4 min)

• Screen capturing with Camtasia

• AVI format for video players– e.g. Windows Media Player,

RealPlayer

• GIF format for common browsers - no sound

Page 77: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

77

S ystemsAnalysis LaboratoryHelsinki University of Technology

Theory VideosCases QuizzesLearningPaths

Assignments

Report templates• detailed instructions in a word document• to be returned in printed format

testing the knowledge on the subject, learning by doing, individual and group reports

Software use• value tree analysis and group decisions with Web-HIPRE

Page 78: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

78

S ystemsAnalysis LaboratoryHelsinki University of Technology

Academic Test Use is Free !

Opinions-Online (www.opinions.hut.fi)

Commercial site and pricing: www.opinions-online.com

Web-HIPRE (www.hipre.hut.fi)

WINPRE and PRIME Decisions (Windows)

RICH Decisions (www.rich.hut.fi)

Joint Gains (www.jointgains.hut.fi)

Smart-Swaps (www.smart-swaps.hut.fi)

Please, let us know your experiences.

Page 79: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

79

S ystemsAnalysis LaboratoryHelsinki University of Technology

Contributions of colleagues andstudents at SAL

• HIPRE 3 +: Hannu Lauri• Web-HIPRE: Jyri Mustajoki, Ville Likitalo, Sami Nousiainen• Joint Gains: Eero Kettunen, Harri Jäälinoja, Tero Karttunen, Sampo

Vuorinen• Opinions-Online: Reijo Kalenius, Ville Koskinen Janne Pöllönen• Smart-Swaps: Pauli Alanaatu, Ville Karttunen, Arttu Arstila, Juuso

Nissinen• WINPRE: Jyri Helenius• PRIME Decisions: Janne Gustafsson, Tommi Gustafsson• RICH Decisions: Juuso Liesiö, Antti Punkka• e-learning MCDA: Ville Koskinen, Jaakko Dietrich, Markus Porthin

Thank you!

Page 80: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

80

S ystemsAnalysis LaboratoryHelsinki University of Technology

Public participation project sites

• PÄIJÄNNE - Lake Regulation(www.paijanne.hut.fi)

• PRIMEREG / Kallavesi - Lake Regulation(www.kallavesi.hut.fi, www.opinion.hut.fi/servlet/tulokset?foldername=syke)

• STUK / Milk Conference - Radiation Emergency(www.riihi.hut.fi/stuk)

Page 81: S ystems Analysis Laboratory Helsinki University of Technology 1 Raimo P. Hämäläinen Systems Analysis Laboratory Helsinki University of Technology

81

S ystemsAnalysis LaboratoryHelsinki University of Technology

SAL eLearning sites

• www.dm.hut.fiDecision making resources at Systems Analysis Laboratory

• www.mcda.hut.fieLearning in Multiple Criteria Decision Analysis

• www.negotiation.hut.fieLearning in Negotiation Analysis

• www.decisionarium.hut.fiDecision support tools and resources at Systems Analysis Laboratory

• www.or-world.comOR-World project site