Document Sample
					                  KONGU ENGINEERING COLLEGE
                           ERODE-638 052.




                                     PRESENTED BY

                                                M.Saravana kumar,
                                                B.E(EEE) (III-yr),

                                    E-MAIL ID:

Artificial intelligence is a branch of computer science concerned with creation of
computer systems that exhibits some form of intelligence. Artificial intelligence is
concerned with two basic ideas 1) Thought processes of humans 2) Representing those
processes via machines symbolic processing, heuristics and pattern matching are the key
elements of artificial intelligence. The pivotal software tool for Artificial intelligence is
lisp (List processing programming language). The hard ware tool is "lisp machine".

       Newell and Simon proposed a model of human problem solving that makes use of
the analogy between computer processing and human information. Artificial intelligence
is widely used in various fields. Perception and knowledge representation in computer
become theme of many Artificial intelligence researches.

       Expert systems are computerized programs that imitate humans. Natural language
processing is a field in Artificial intelligence that processes natural language. Another
important application of Artificial intelligence is the field of robotics. Robot is
mechanical Artificial intelligence. The application of Artificial intelligence relates to
almost many fields. Artificial intelligence has civilian uses, professional uses, and
applications in music, military and gaming. Artificial intelligence theories imply much
futuristic applications which are expected to progress as fast as computer technology.

       Artificial intelligence is the study of mental faculties through the use of
computational models. Intelligent computer systems that exhibit the traits are associate
with human intelligence as called as artificial intelligence.

       1.0 Concepts of Artificial intelligence

       2.0 Elements of Artificial intelligence

       3.0 Hardware and software tools of Artificial intelligence

       4.0 Artificial Intelligence Vs Human Intelligence

       5.0 The Newell Simon model of human information processing

       6.0 Artificial intelligence applications

       7.0 Future Outlook

       Artificial intelligence is a branch of science and technology concerned with study
and creation of computer systems that exhibits some form of intelligence. Systems that
learn new systems and tasks, systems that can understand a natural language or perceive
and comprehend a visual scene and systems that perform other types of feats that require
human types of intelligence.

       It is concerned with 2 basic ideas.
       1. It involves the thought processes of human.
       2. It deals with representing those processes via machines.
1.1. The Turing test
       According to this test, a computer could be considered thinking only when a
human interviewing conversing with both an unseen human being and unseen computer
could not determine which is which.

2.1. Symbolic processing
       Symbolic processing is an essential characteristic of Artificial intelligence.
Artificial intelligence deals with Artificial intelligence methods of problem solving and
human reasoning processes symbolic, non-algorithmic methods of problem are symbolic
and non-algorithmic solving. Many human reasoning processes are symbolic and non-
algorithmic. So much Artificial intelligence research continues to be devoted to symbolic
and non-algorithmic processing techniques to emulate human reasoning process by a

2.2. Heuristics and Artificial intelligence
       Heuristics is a key element of Artificial intelligence that deals with ways of
representing knowledge using symbols. People use heuristics to make decisions Artificial
intelligence uses heuristic methods for processing information. Researchers use the term
heuristic to describe the judge-mental or common sense part of problem solving.
Heuristics deals with problem definitions, the selection of reasonable strategies to be
followed and formulation of hypothesis and hunches. Though humans are superior in
heuristics, it is possible to develop new systems with improved heuristic abilities.

2.3. Pattern recognition
       Artificial intelligence works with pattern matching methods which attempts to
describe objects or events or processes in terms of their qualitative features and logical
and computational relationships. If computers are to be intelligent, they must be able to
make same kinds of associations among qualities of objects, events and processes that
come too naturally to people.
       The pivotal software tool for Artificial intelligence (1950's) is lisp (list processing
programming language) developed by John McCarthy in 1957. Lisp deals with complex
objects such as rules, sentences, names not just numbers. Therefore lisp lends itself to
development of flexible systems that can accommodate ambiguities and complex
interrelationships among data. David Marr developed PROLOG language which was
convenient language for Artificial intelligence researches in 1970. The pivotal hardware
tool for users of lisp is "LISP machine" or symbolic processor. A computer system that
has a logical architecture specifically designed to efficiently and effectively support
Artificial intelligence program development and execution. These LISP machines were
replaced by standard micro computers with Artificial intelligence software’s in C
language in 1987.

       Artificial Intelligence is permanent but natural intelligence is perishable. Artificial
intelligence offers case of duplication and so knowledge from one system can be copied
from one computer to other computer across the globe. Artificial intelligence is consistent
and thorough but human intelligence is erratic and has no consistency Artificial
intelligence can be documented. Decisions can be documented by tracing activities of
system. Human Intelligence is creative, enables people to benefit by and use sensory
experience directly and human reasoning has a wide context of experience.

       In applying Artificial intelligence, we consider a different approach to problem
solving and decision making. Of special interest to Artificial intelligence is the Newell –
Simon model. Allen Newell and Herbert A. Simon proposed a model of human problem
solving that makes use of the computer processing and human information. This model
helps us to know how Artificial intelligence works and what its limitations are. The
human information processing systems consist of following sub-systems.
       A perceptual sub-system, a cognitive sub-system, motor sub-system and an
external memory.

5.1. The perceptual sub system
       External stimuli are the input for human information processing system. These
stimuli enter through our sensors like our eyes and ears. The perceptual sub system
consists of these sensors along with buffer memories that briefly store incoming
information while it awaits processing by the cognitive sub-system.

5.2. The cognitive sub system
       The senses are constantly placing a huge amount of information in buffer
memories. Like a CPU in a computer, the cognitive processor selects appropriate
information from sensory buffers and transfers it to short term memory. The processor
works in cycles and during each cycle it obtains information and transfer it to another
memory, which is analogous to "fetch-execute" cycles of computer.

       The processor has three parts: The elementary processor, the short term memory
and the interpreter which interprets part or all of program of instructions for problem
solving. The program used by an individual will depend on no. of variables such as task
and intelligence of problem solver. In the simplest tasks, cognitive system merely serves
as a point for transferring information from sensory inputs to motor inputs. More
complex tasks involve more information which calls for elaborate processing. To
accomplish these tasks, the cognitive processor will be drawn on a second memory
system, long term memory.

5.3. Long term and external memories
       Long term memory consists of a large no. of stored symbols with a complex
indexing system. There are competing hypothesis how these symbols are arranged. In the
simplest memory model related symbols are associated with one another but in a complex
model they are arranged as temporal scripts. Another view is that memory consists of
clusters of symbols called chunks (or) memory is a vast network of chunks, each chunk
being a unit of stored information and can be a digit, a symbol or a word associated with
a set or pattern of stimuli. It requires only few 100 millisec to read but write time is fairly
long (5 N to 10 N sec for N symbols). The processing, retrieval and storage of data by
computers is faster than that of humans. The long term memory has unlimited capacity.
The short term memory is quite small, but 2 chunks can be retained while another task is
being performed. This suggests that part of short term memory is used for input and
output processing. This is one major limitation of human as compared with a computer.
The limits of short term memory can be expanded through analogies and graphics. The
human operates according to this model in a serial fashion, rather than in parallel, but a
computer can operates in series and parallel designs.

       Human beings support decision making with external memory (i.e.) a paper pad
or chalkboard. Humans are limited in their ability to generate, integrate and interpret
probabilistic data.

5.4. Motor output
       After scanning and searching memories, the processor sends information to motor
sub-system. Motor processors initiate actions of muscles and other internal human
systems. This in turn results in observable human activity such as talking.

6.1. Expert systems
       Expert systems are computerized advisory programs that attempt to initiate or
substitute the reasoning processes and knowledge of experts in solving specific types of
problems. An expert system is a computer program that attempts to behave like a human
expert in some useful ways. Today's state of art is that expert systems can solve simple
problems and provide limited explanations. Computer Aided Instruction (CAI) refers to
machines that tutor humans. To a certain extent, it can be viewed as an expert system.
XCON was the first expert system which symbolized the budding of real Artificial
intelligence application in 1980's.
6.2. Natural language processing
        Natural language processing (NLP) is an attempt to allow computers to interpret
normal statements expressed in native human language. Speech recognition is a process
that allows one to communicate with a computer. Combining speech recognition and
NLP will be required to realize the capability of the computer to converse in a manner
normal to humans. There are 3 elements in NLP parser, figures out how a sentence is put
together with nouns semantic analyzer. Interprets meaning of words in sentence code
generator, translates user sentences into machine language codes.

Some of the Natural Language Processors are ELIZA, IQATS
        ELIZA: ELIZA was a psychiatrist, particularly one that posed analytical questions
for every answer the user gave it. Though sometimes they may have seemed ambiguous,
people actually felt ELIZA could take care of their needs just as well as any other

        IQATS stands for Intelligent Question and Answer Test Summarizer. A program
that creates questions for test making purposes. IQATS also learns to ask new questions
if given a sample question and answer.

6.3. Robotics
        An area that is closely related to Artificial intelligence deals with sensory systems
such as visual system, tactile system and signal processing system. Such systems
combined with Artificial intelligence define a broad category of systems referred to as
robotics. The robot senses the environment and modifies the behavior as a result of
information gained. The intelligent robot is thought to have human like capabilities and
attributes. A robot is a mechanical Artificial intelligence. A robot is a reprogrammable,
multi-functional manipulation designed to move material, parts, and tools on specialized
devices through various programmed motions for the performance of variety of tasks.
6.4. Applications in gaming
       Games like Quake 2, unreal are showing some interesting advances in Artificial
intelligence, most simulators (both flight simulators like US Navy fighters and
Mechwarrior use goal based Artificial intelligence techniques. Additionally games such
as flight simulators require Artificial intelligence for planes, boats vehicles and troops.
The victory of deep blue over chess champion kasparov in 1996 led to a new summit of
Artificial intelligence gaming.

6.5. Applications in Music
       Artificial intelligence can be used in music in many different ways, it can be used
to compose (create music) and transpose (create written music from listening to pieces).
Computers that compose of ten require human input to determine whether music sounds
good or not. For a computer to listen to a 4 member rock band playing and to create 4
written pieces of music to correspond perfectly with the guitar, bass, drums and singer
would be an incredible task.

6.6. Applications in military
       Military sees that mass potential in Artificial intelligence on the battle field.
Artificial intelligence can be used to protect the safety of the crew. For e.g. the radar of
Apache long bow uses advanced algorithms that the threat determined is infact a threat.
Other uses of Artificial intelligence are advanced autopilot features that take away
tedious tasks of flying. Automated landing systems (or) computer aided landing systems
can greatly decrease peacetime and war time accidents.

6.7. Civilian uses
       1. An artificial intelligent vacuum eleven goes around the house, remembers the
           layout of the house after a few trial runs and memorizes the areas that tend to
           be dirtier.
       2. An artificial intelligent program corrects the spellings in an essay, analyzes
           the document for grammar errors and alerts you of such errors just like a
           human corrector would do.
       3. An Artificial intelligence powered search engine finds web pages given your
           natural language search requirement.
       4. Your security system is linked up to year computer, running an image
           recognition package.

6.8. Professional uses

    1. Aeronautical science: Weather radars are designed to cleverly detect wind shears,
        turbulence and other such hazards.
    2. Medical science:        Doctors could refer to a knowledge base that stores all the
        patients’ problems over the course of practice. Artificial intelligence is used in
        gaming, music and has military applications as well.

       Artificial intelligence is a result of merge of philosophy, mathematics,
psychology, neurology, linguistics, computer hard ware and software, hydraulics topics.
Further the application of Artificial intelligence relates to many fields which give
Artificial intelligence an endless potential. The intersection between psychology and
Artificial intelligence centers on areas known as cognition and psycholinguistics.
Artificial intelligence theories are ingenuous and subtle and imply futuristic applications.
Artificial intelligence incorporated with computer technology is expected to progress as
fast as the computer technology, conclusionally; Artificial intelligence is a young,
energetic and attractive science.

Books referred
1. Decision support and expert systems – Efraim Turban.
2. Computer applications in business – S.P. Rajagopalan
3. Computer aided analysis and design – Ghoshal
4. Computer aided design – Software and analytical tools
       C.S. Krishnamoorthy, S. Rajeev

Shared By: