Upload
marjorie-mcdaniel
View
215
Download
1
Embed Size (px)
Citation preview
Intelligent systemsLection 2 – Applied AI Systems
Part 1
Kinds of Applied AI Systems• Expert systems (ES) • Intelligent Robots• Programs with Natural Language Processing (NLP)• Systems for Intelligent Data Analyzing (IDA)• Artificial Life• AI in games• Identification in security systems• Systems for smart home, in particular, caregiver for
elders• Embedded AI in different useful systems (OS, office
programs, programming tools, control systems)
Experts systems.Tasks solved by ES and areas of using
Tasks:
•Diagnostics•Monitoring•Control•Identification•Planning•Forecasting•Simulation•Training•Projecting (configuring)
Areas:
•Business•Medicine •Manufacturing•Computers •Communications •Science (mathematics, physics, chemistry)•Transport•Space•Defense •Education
Examples of ES• MYCIN – (1970) for diagnostics of enteric infection• DENDRAL (1969) – for determination of structure
of molecules from spectral analyzing• PROSPECTOR – for searching of minerals• XCON – for configuring of computer systems• YES/MVS – for supervising of OS/370 IBM• POMME – for care for apple trees• AIRPLANE – for help to pilot to land on aircraft-
carrier• MODIS – for diagnostics of hypertension• NetWizard – for configuring of local networks• MIDAS – for diagnostics and repair of power system
Structure of ES
Knowledge base (KB)
Engine forsolving of task
Subsystem formodification of KB
Subsystem forexplanation
Subsystem forinteraction
Working memory (base of facts and hypothesis)
User
Programs with Natural Language Processing (NLP)
– Translators (ex. - PROMT from ABBYY)
– OCR-Systems (ex. – Finereader of ABBYY)
– Recognition of speech (ex. - Dragon from IBM)
– Dialog systems (ex. - ALICE-like programs)
– QA-Systems (Question-Answer)
– Virtual Persons (ex. – Ramona http://www.kurzweilai.net/index.html)
– Automatic reader (with synthesis of speech)
– Search of documents by query by NL
Systems for Intelligent Data Analyzing (IDA)
• Statistica
• PolyAnalyst from Megaputer
• TextAnalyst from Megaputer
• MatLab
• MathCad
Artificial Life• Models of evolution• Models of animals (fish, ants)• Models of cooperation in population
Goal:• Investigations of animals and life (test of our
performances)• Entertainment• Tutoring in school
AI in games (AIG)• AI of personage in Actions or Real Time Strategies• AI for support of player• AI for simulation of world (for example, in strategies)
Difference between AI in games and AI of scientists:• AIG is more simple• AIG is oriented on intelligent-like behavior but not
deep simulation of mind (on commercial result)• Often the main requirement is speed of action
Identification in security systems
• Recognition of face
• Recognition of emotions and intention of people
• Identification of person by prints of fingers
• Identification of behavior with credit card
• Identification of hacker’s attacks
• Computer vision for control of highways
Systems for smart home, in particular, caregiver for elders
7
System Function
Executing a daily medical checking plan
Generating efficient discourse plansaccording to the user’s symptoms
Dealing with the user’s interruptions which can cause a spontaneous topic shifting
Providing a usable interface for elderly people using audio and visual interactions
Functions of caregiving system developed in Fukuoka University
Embedded AI in different useful systems (OS, office programs, programming
tools, control systems and so on)• Devices for checking of natural language in
MS Word• Fuzzy control programs in consumer
electronics• Recognition of speech commands in OS/2
and OS MS Windows• Synthesis of speech in auto electronics• Speech recognition in cellular phones