SYST 542 Decision Support Systems Engineering

Department of Systems Engineering and Operations ResearchSYST 542Decision Support SystemsEngineeringInstructor: Kathryn Blackmond LaskeyFall Semester, 2006Unit 1: Decision Makingand Decision SupportSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 1 -

Department of Systems Engineering and Operations ResearchLearning Objectives Understand course objectives &requirements Define a DSS Describe the history of DSS Name and define the major functionalcomponents of a DSS Describe the DSS lifecycle Describe the role of DSS in decisionmaking and the kinds of decisions mostamenable to DSSSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 2 -

Department of Systems Engineering and Operations ResearchCourse Objectives Introduce decision support systems Provide sound basis for:– Designing DSS– Managing DSS lifecycle process– Evaluating DSS Provide systems view of DSS developmentand integration into organizationSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 3 -

Department of Systems Engineering and Operations ResearchCourse Requirements Weekly discussion question (30% of grade)––––Asynchronous discussionIn-class discussion with assigned facilitatorWritten summary50% participation, 50% content Project (50% of grade)––––Small groupsDesign and implement DSS for problem of your choiceWritten reportOral presentation Paper review (15% of grade)– Read a paper from the literature– Write report on paper– Give oral presentation Lead discussion session (5% of grade)SYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 4 -

Department of Systems Engineering and Operations ResearchA decision support system is acomputer-based system that supportsthe decision making process Assist decision makers in semi-structured tasksSupport not replace human judgmentHighly interactiveImprove effectiveness of human decision makers“A decision support system is a system under the control of oneor more decision makers that assists in the process of decisionmaking by providing an organized set of tools to impartstructure to portions of the decision-making situation andimprove the ultimate effectiveness of the decsion outcome”- MarakasSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 5 -

Department of Systems Engineering and Operations ResearchWhy DSS? Increasing complexity of decisions– Technology– Information:» “Data, data everywhere, and not the time to think!”– Number and complexity of options– Pace of change Increasing availability of computerized support– Inexpensive high-powered computing– Better software– More efficient software development process Increasing usability of computers– COTS tools– CustomizationComputer support for decision makingSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 6 -

Department of Systems Engineering and Operations ResearchRational Decision Making Rationality is the use of reason to make the bestchoice one can in the circumstances– What does “best” mean? Aspects of rationality (Kant)– Cognitive rationality: What to believe?– Practical rationality: What to do?– Evaluative rationality: What to value? GOOD-D mnemonic for rational decision making–––––Identify the goal to be achieved by the decisionIdentify the options available to the decision makerEvaluate the likely outcomes if each option is chosenDecide which option is best And then Do it! Decision makers need support with all GOOD-DelementsSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 7 -

Department of Systems Engineering and Operations ResearchDecision Making ProcessDefine andStructure theProblemGenerateOptionsGather, Collectand Fuse dOptionSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 8 -

Department of Systems Engineering and Operations ResearchTypes of Problems Structured– Repetitive– Standard solution methods exist– Complete automation may be feasible Unstructured––––One-timeNo standard solutionsRely on judgmentAutomation is usually infeasible Semi-structured– Some elements and/or phases of decision making processhave repetitive elementsDSS most useful for repetitive aspectsof semi-structured problemsSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 9 -

Department of Systems Engineering and Operations ResearchHistory of DSSOperations ResearchInformation Systems Management Science AI/Expert Systems1940’sOptimizationCognitive ScienceTransactionProcessing SystemsExpert SystemsSimulationMISKnowledgeRepresentation2000 Judgment &Decision MakingHuman/ComputerInteractionDecision Support SystemsGoal: Use best parts of IS, OR/MS, AI & cognitivescience to support more effective decision makingThanks to Andy LoerchSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 10 -

Department of Systems Engineering and Operations ResearchHistory: Business Computing World War II Era 1980-90’s– Introduction of computers - Military andscientific applications– Computers were for “numbercrunching” 1950’s– Business applications– Transaction processing systems:billing & payroll– Large mainframe computers 1960-70’s– Use of computers in management– Large volumes of data stored incomputers– Invention of relational databases andSQL– Management Information Systems born– Automation of paper-and-pencilprocesses for repeatable tasksSYST 542– Movement towardcustomization & flexibility– Movement toward new userinteraction metaphors– Increasing emphasis onintelligent systems 21st Century– Move from “stovepipes” tointeroperable systems– Distributed systems– Web servicesCopyright 2006, Kathryn Blackmond LaskeyUnit 1 - 11 -

Department of Systems Engineering and Operations ResearchHistory: Artificial Intelligence 1950’s 1980-90’s– Introduction of symboliccomputing– Newell and Simon: GeneralProblem Solver– Differentiation fromscientific computing» “AI is about symbols andnot numbers”Commercialization of AIExpert system shellsConnectionist movementMachine learningIncorporation of methodsfrom decision theory andoperations research 21st Century 1960-70’s– First expert systems» e.g., HEARSAY I (Speechrecognition); MYCIN(Medical diagnosis)– Knowledge representation e.g., frames, rules– Fuzzy logicSYST 542–––––– Agent-based systems– Distributed AI– Semantic Web & Intelligent“Web Bots”Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 12 -

Department of Systems Engineering and Operations ResearchHistory: OR/MS 1980-90’s World War II Era– Application of scientific method tooperational problems» e.g., efficient movement of troops &equipment 1950’s– OR/MS established as a discipline– Standard methods developed 1960-70’s––––Expansion of OR/MSBusiness applicationsGovernment: McNamara’s “whiz kids”Problems with appropriate use andacceptance» Limited computing power» Exclusion of factors not easilyquantified» Human factors issues in how OR/MSintegrated into organizationsSYST 542– Movement towardcustomization & flexibility– Attention to organizationaland human factors– Incorporation of methodsfrom artificial intelligence– Model bases and modelmanagement 21st Century– Embedded systems– Agile, reconfigurable supplychains– OR for everyone» Excel Solver» OR in middle school– OR in a Web Services worldCopyright 2006, Kathryn Blackmond LaskeyUnit 1 - 13 -

Department of Systems Engineering and Operations ResearchDecision Support Trends IT is increasingly pervasive Users are increasingly computer savvy Computer hardware is increasingly smallerand more powerful Systems are increasingly interconnected The Web is increasingly interwoven into allaspects of our lives Demand for usable, flexible, powerful decisionsupport will continue to grow Decision support will be embedded into a widevariety of consumer and business productsSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 14 -

Department of Systems Engineering and Operations ResearchDiscussion Give an example of some decision supportsystems you have encountered– What kind of decision was supported?– How did it work?– How helpful was it? What makes for successful decision support? What pitfalls should be avoided?SYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 15 -

Department of Systems Engineering and Operations ResearchSome Terminology DBMS - System for storing and retrieving data andprocessing queries Data warehouse - Consolidated database, usuallygathered from multiple primary sources, organized andoptimized for reporting and analysis MIS - System to provide managers with summaries ofdecision-relevant information Expert system - computerized system that exhibitsexpert-like behavior in a given problem domain Decision aid - automated support to help users conformto some normative ideal of rational decision making DSS - provide automated support for any or all aspects ofthe decision making process EIS (Executive information system) - A kind of DSSspecialized to the needs of top executivesSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 16 -

Department of Systems Engineering and Operations ResearchTraditional EmphasisDataDialogueModelDBMS/DWMISESDADSS/EISSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 17 -

Department of Systems Engineering and Operations ResearchHumans and Computers:Complementary Strengths Human decision makers––––––Good at seeing patternsCan work with incomplete problem representationsExercise subtle judgment we do not know how to automateOften unaware of how they perform tasksPoor at integrating large numbers of cuesUnreliable and slow at tedious bookkeeping tasks andcomplex calculations Computers– Still inferior to humans at pattern recognition, messyunstructured problems– Good at integrating large numbers of features– Good at tedious bookkeeping– Rapid and accurate at complex calculationsSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 18 -

Department of Systems Engineering and Operations ResearchThe Challenge:Find and Exploit Synergy Computers provide cognitive tools– You would not build a house without appropriate tools– Complex decision problems require cognitive tools to assist with:» Collecting and organizing relevant information» Weighing multiple factors relevant to choice» Integrating large numbers of factors and combining to form overallevaluation» Presenting results so rationale for choice is clear» Analyzing multiple “what-if” scenarios Goal of DSS:– Use strengths of computer to augment strengths of human– Improve overall effectiveness of decision making processSYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 19 -

Department of Systems Engineering and Operations ResearchDanger: The Worst of Both Worlds A poorly designed or improperly deployeddecision support system can be– Worse than leaving users unassisted– Worse than replacing the users with automated system Can you explain why? How do we keep this from happening?SYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 20 -

Department of Systems Engineering and Operations ResearchAchieving the Promise Understand the stakeholders– Involve stakeholders early and often– Listen to feedback (especially negative!) Understand the task– Objectives to be achieved by decision– Current decision making process– Human and organizational factors Understand the technology– What parts of current process can be automated– COTS versus custom development– Integration of components and non-automated functions Understand the DSS development process– Co-evolution of process, DSS, human skill sets– Why change is resisted– Importance of good systems engineeringIterate! Evaluate! Improve!SYST 542Copyright 2006, Kathryn Blackmond LaskeyUnit 1 - 21 -