the strategic management of information technology chapter 10 complex decisions and artificial...
TRANSCRIPT
The StrategicManagement of
InformationTechnology
Chapter 10Chapter 10Complex Decisions andComplex Decisions and
Artificial IntelligenceArtificial Intelligence
Transaction Processing Transaction Processing SystemSystem
Input OutputProcess
Information
Communication
Systems Development
OverviewOverview
Business Problems– Complex, less structured
Data– Non-numerical, messy, complex relationship
Artificial Intelligence– Goal is to make computers “think” like humans
Building Expert SystemsBuilding Expert Systems
Knowledge Base Knowledge Engineers Case-Based Reasoning Limitations of Expert Systems
Expert SystemExpert System
Expert Symbolic and/or Numeric Knowledge Knowledge Base Expert Decisions made by non-experts
Decision Support System Decision Support System Compared to Expert SystemCompared to Expert System
DSS ESS
Goal Help User MakeDecision
Provide ExpertAdvice
Method DataModelPresentation
Asks QuestionsApplies rules andExplains
Type ofProblems
General, limited byuser
Narrow Domain
Building Expert SystemsBuilding Expert Systems
Shell = Tool to Build Expert System Knowledge Engineer Builds Cooperative Expert Key Components:
– Knowledge Base– Information Engineer applies rules to new data
for each conclusion Custom Program, Shell, or Pre-packaged
Additional Issues to ConsiderAdditional Issues to Consider
Pattern Recognition/Neural Nets Voice and Speech Recognition Language Comprehension Massively Parallel Computers Robotics and Motion Statistics, Uncertainty, Fuzzy Logic
Expert SystemsExpert Systems
Goal: Make same decision an expert would make with the same data
Capture and program expert’s knowledge Advantage of speed and consistency
Expert Systems Problem TypeExpert Systems Problem Type
Narrow, well-defined domain Solutions require an expert Complex logical processing Handle missing, ill-structured data Need a cooperative expert
Limitations of Expert SystemsLimitations of Expert Systems
Fragile Systems– Small environment changes can force revision of
all of the rules Mistakes
– Who is responsible? Expert Multiple Expert Knowledge Engineer Company that uses it
Limitations of Expert SystemsLimitations of Expert Systems
Vague Rules– Rules can be hard to define
Conflicting Experts– With multiple opinions, who is right?– Can diverse methods be combined?
Limitations of Expert SystemsLimitations of Expert Systems
Unforeseen events– Events outside of domain can lead to nonsense
decisions– Human experts adapt– Will human novice recognize a nonsense
result?
AI Research AreasAI Research Areas Computer Science
– Parallel Processing
– Symbolic Processing
– Neural Networks Robotics Applications
– Visual Perception
– Tactility
– Dexterity
– Locomotion and Navigation
AI Research AreasAI Research Areas
Natural Language– Speech Recognition– Language Translation– Language Comprehension
Cognitive Science– Expert Systems– Learning Systems– Knowledge-Based Systems
Neural NetworksNeural Networks
Based on brain design Hardware and software Recognize patterns
– Design specifications– Spiegel Catalogs– Pick stocks
Machine VisionMachine Vision
Advantages of Machine Vision– Broader spectrum of light– Will not suffer fatigue– Damage less easy
Literal– Problems less detection than processing
Speech RecognitionSpeech Recognition
Voice: primarily ID Speech
– Transcripts– Hands-free operations
Limitations– Need to train– Accents and colds– Synonyms, punctuation, context
AI QuestionsAI Questions
What is intelligence? Can machines ever think like humans? How do humans think? Do we really want computers to think like
us?
Other AI ApplicationsOther AI Applications
Massively Parallel Processing– only if task can be split into independent pieces– math computation and database searches
Robotics and Motion– welding and painting
Statistics, Unclear, and Fuzzy Logic– use subjective and incomplete description
The FutureThe Future
Intelligent Agents– Learn what you want from what you ask for
and go get it for you– Automated personal assistant– Network traffic can be a problem– Agents are independent of one another
ProductChange
Process Change
Dynamic
Stable
Stable Dynamic
Mass customization Invention
Mass production Continuous improvement
Product-Process Change Matrix
Product Change
Process Change
Dynamic
Stable
Stable Dynamic
Mass ProductionChange conditions Periodic/forecastable changes in product
market demand and process technology
Strategy Production
Key organizational tool Standardized, dedicated production process
Workflows Serial, linear flow of work, executed to plan
Employee roles Separate doers and thinkers
Control system Centralized, hierarchical command system
I/T alignment challenge Automation of manual processes to achieve costjustified efficiency enhancement
Critical synergy Reliance on invention form to supply new product designs and new process tech.; linked with invention forms in single corporate entity
Product-process change matrix
Product change
Process change
Dynamic
Stable
Stable Dynamic
InventionChange conditions Constant/unforecastable changes in product
market demand and process technology
Strategy Production of unique or novel product or process
Key organization tool Specialization of creative or high craft skills
Workflows Independent work
Employee roles Professionals and craftspeople
Control system System decentralized to specialized individuals and groups
I/T alignment Development and distribution of customized systems
Critical synergy Mass production form supplied with new processes; operates in market niches too dynamic or small for mass production; sometimes incorporated into single corporate entity with multiproduct mass-production forms
Figure 3 Product-process change matrix
Product change
Process change
Dynamic
Stable
Stable Dynamic
Mass CustomizationChange conditions Constant/unforecastable changes in market
demand; periodic/forcastable changes in process technology
Strategy Low cost process differentiation within new markets
Key organization tool Loosely coupled networks of modular,flexible processing units
Workflows Customer/product unique value chains
Employee roles Network coordinator and on-demand processors
Control system Hub and web system; centralized network coordination, independent processing control
I/T alignment Integration of constantly changing network info processing/communication requirements; interoperability, data communication, and coprocessing critical to network efficiency
Critical synergy Reliance on continuous improvement form for increasing process flexibility within processing units
Figure 5 Product-process change matrix
Product change
Process change
Dynamic
Stable
Stable Dynamic
Continuous ImprovementChange conditions Constant/unforecastable changes in process
technology, periodic/forecastable changes in market demand
Strategy Low cost process differentiation within mature markets
Key organization tool Self-managing/cross-functional teams
Workflows Intensive and reciprocal workflow within teams
Employee roles Dual, combined doers and thinkers
Control system Microtransformations; rapid and frequent switching between decentralized team decision making and team-managed command systems
I/T alignment Design of cross-functional info and communication systems that support micro-transformations
Critical synergy Mass-customization form supplied with flexible new processes; sometimes functions as transition form in re-engineering to mass customization
Figure 6 Product-process change matrix