INTELLIGENT SYSTEMS IN BUSINESS

1
CHAPTER 12

• INTELLIGENT SYSTEMSIN BUSINESS

2
Artificial Intelligence (AI)

• involves studying the thought processes of humans
• deals with representing those processes via
  machines
• objectives
• to make machines smarter
• to understand what intelligence is
• to make machines more useful

3
Intelligent Behavior

• Understanding experience
• Learning or inferring from experience
• Making sense of contradictory messages
• Fast successful response to a new situation
• Using reasoning to solve problems and direct
  actions effectively
• Range the elements according to their importance
  in a situation

4
Comparing Artificial and Natural Intelligence

• AI Advantages
• more permanent
• less expensive
• consistent and thorough
• can be documented
• easy for duplication and dissemination

• Natural intelligence
• Advantages
• creative
• enables people to use directly sensory
  experience
• enables people to recognize relationships,
  feelings
• human reasoning is more comprehensive

5
Conventional vs. AI Computing
6
Commercial AI Systems

• Expert systems (ESs)
• Natural language technology
• Speech (voice) understanding
• Handwriting recognizers
• Robotics and sensory systems
• Computer vision and scene recognition
• Machine learning

7
Expert Systems
Expertise is transferred from an expert to a
computer and it is stored there
8
Expertise Transformation

• Knowledge acquisition
• Knowledge representation
• Facts
• Rules
• Knowledge inference
• Knowledge distribution

9
Knowledge Inference

• Rules in implication form
• Two logical approaches
• Deductive If a hypothesis is true, then a
  conclusion is made
• Heuristic Assuming that a rule is correct, if a
  conclusion is true, then hypothesis is also true
  
• A complex rule is transformed into a set of
  simple rules

10
Expert Systems Components

• Knowledge base
• Inference Engine
• User Interface
• Explanation subsystem
• Workplace

11
Knowledge base

• Facts true statements
• Procedural knowledge rules to deal with facts
• Control Knowledge a set of possible strategies
• Metaknowledge

12
Inference Engine

• Production system most popular methodology
• Interpretation pair (fact, production rule)
• If pair (fact, production rule) is found, the
  conclusion is made based on the production rule

13
Other components of ESs

• User Interface
• Question answer form
• Explanation subsystem
• Keep track of facts and rules applied
• Metaknowledge
• Workplace (working area)

14
ES Development

• Knowledge acquisition
• Knowledge engineering
• Development tools
• Prolog logical programming language
• Knowledge refinement

15
The Process of ES Operation
Consultation Environment
Development Environment
User
Facts about the specific incident
Knowledge base Facts Rules
User interface
Knowledge engineer
Explanation facility
Knowledge acquisition
Inference engine draws conclusions
Expert and documented knowledge
Recommended action
Knowledge refinement
Workplace
16
Benefits of Expert Systems

• Increased output and productivity
• Increased quality and reliability
• Capture of scarce expertise
• Ability to operate in hazardous environment
• Improved customer service
• Human-like intelligence
• Fault tolerance
• Complex problem solving and decision making
• Training capabilities

17
Limitations of Expert Systems

• Limited expertise
• No single correct solution
• Natural cognitive limits
• Narrowly defined subject areas
• Limited vocabulary
• Cost
• Lack of trust by end users
• Liability issues

18
Natural Language Processing (NLP)

• Communicating with a computer in natural language
  via keyboard or voice
• Voice Technology
• Voice (speech) recognition
• Command
• Discrete
• Continuous
• Speech understanding
• Voice synthesis
• the technology by which computer speaks

19
Blackboard Method

• Knowledge sources
• Facts, hypotheses and conclusions are placed from
  bottom to top on the working area blackboard
• Hypotheses hierarchy phonemes, words, parts of a
  sentence, sentences
• Two main strategy upward and downward

20
Neural Computing

• An ANN emulates a biological neural network
• Parallel processors (artificial neurons) are
  interconnected with each other
• Each AN receives information from other neurons
  or from external sources
• Each input has assigned weight
• Weights can be adjusted after each case is
  processed (learning capability)
• Transforms the information, and passes it on to
  other neurons or as external outputs

21
ANNs Characteristics

• Information retrieval even if some nodes fail
• Fast modification of stored data
• Discover relationships and trends in large
  databases
• Solve complex problems with lack of data

22
ANNs Benefits

• Fault tolerance
• Generalization
• Incomplete inputs can lead to a reasonable
  response
• Learning capability
• Forecasting capabilities

23
Application of ANNs

• Pattern recognition and learning capability
• Financial application
• New product analysis
• Evaluation of personnel
• Resources allocation
• Handwriting text recognition
• Financial fraud detection

24
Case-Based Reasoning (CBR)

• Adapt solutions that were used to solve old
  problems and use them to solve new problems
• Input problem is described by assigning
  appropriate features
• Retrieval cases with the same features from
  memory
• Choosing the most relevant case
• An extremely effective approach in complex cases
• Applications tactical planning, political
  analysis, situation assessment, diagnostics

25
Fuzzy Logic

• Deals with uncertainties
• Uncertainty in facts and in rules
• Uncertainty is expressed in probability
• Procedures how to evaluate situation are
  introduced
• Allows to make reasonable decisions when there
  are no reliable data

26
Intelligent Agents (IA)

• Characteristics of Intelligent Agents
• autonomy -capability to work on their own,
  goal-oriented behavior
• mobility - transportable over networks
• dedication to a single repetitive task
• ability to interact with humans, systems, and
  other agents
• inclusion of knowledge base
• ability to learn

27
Applications of IA
28
Virtual Reality (VR)

• Most common definitions interactive,
  computer-generated, three-dimensional graphics,
  delivered to the user through a head-mounted
  display
• Technical definitions environment and/or
  technology that provides artificially generated
  sensory cues sufficient to engender in the user
  some willing suspension of disbelief

29
Applications of VR
30
Issues Related to AI

• Ethical
• Social
• Legal
• Global aspects