Artificial Intelligence In Natural Language Processing Computer Science Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

When the first computers were invented, the only way that you could give them instructions was to connect and reconnect a series of wires, somewhat reminiscent of an old-fashioned telephone switchboard. Although today's computers have become much easier to use, communicating with a computer is not yet as simple as the most "natural" kind of communication: communicating with another person.We don't have to use any kind of specialized, technical language to communicate with other people-we use NATURAL lanugaes, such as English. If computers could understand natural language, you could tell them what you wanted them to do in ordinary, everyday English. If computers could generate natural language, they could ask you questions and give you information in language that would be easy to understand. For that we introduces the NATURAL LANGUAGE PROCESSING.The phrase Natural Language Processing generally refers to language that is typed, printed, or displayed, rather than being spoken. Getting computers to understand spoken language is the focus of related AI technologies called speech recognition and speech understanding.


The definition of artificial intelligence (AI) which is the branch of computer science, it is "the study and design of "Intelligent Agents" where an intelligent agent is a system that perceives its environment and takes actions which maximizes its chances of success.

John McCarthy, who coined the term in 1956 at the Massachusetts Institute of Technology(MIT).

According to him, AI is the science and engineering of making intelligent machines, especially intelligenat computer programs.

It is related the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable.


We can define Intelligence as the ability to acquire, retrieve, and use knowledge in a meaningful way. It includes both raw and refined knowledge and the ability to memorise, recall facts, and express emotions. Reasearch in AI has focussed chiefly on the following components of intelligence:






LEARNING: It is the process of acquiring knowledge, skills, experience or values by study, experience or training.

REASONING: It refers to the ability of drawing conclusions that are appropriate to the situation in hand.

UNDERSTANDING: It refers to the identification of the significance, interpretation, or explanation for certain data or information. Simply put, it is the ability to employ knowledge.

CREATIVITY: It is the ability to generate new ideasor to conveive new perspectives on existing ideas. The creativity process involves producing ideas, which are original and potentially useful.

INTUITION: It is the inner knowledge, without rational processes and without being aware of how we know. Essentially, intuition is an uncanny sixth sense that tells people that whether they are right or not.


"The automation of activites that we associate with human thinking, activities such asdecision-making, problem-solving, learning".

-BELLMAN (1978).

"The study of how to make computers do things at which, at the moment, people are better".


"The study of mental faculties through the use of computational modes".


"The branch of computer science that is concerned with the automation of intelligent behaviour".


According to above definitions, AI falls into four categories.

Systems that think like humans

Systems that act like humans

Systems that think rationally

Systems that act rationally








Each of these categories represents a unique field in its own right but at the same time all four categories share a common ground, with each group of pople providing valuable insights.


In 1950, Alan Mathison Turing proposed a test called TURING TEST to prove the intelligtence of a machine. This test provides the basis of what came out to be known as Artificial Intelligence.

Described by Alan Turing in the 1950 paper "Computing machinery and intelligence," it proceeds as follows:

A human judge engages in a natural language conversation with one human and one machine, each of which try to appear human if the judge cannot reliably tell which is which, then the machine is said to pass the test. In order to keep the test setting simple and universal, the conversation is usually limited to a text-only channel such as a teletype machine as Turing suggested.

This test actually consists of three participants:

Two humans

A computer

The idea was that a human and a machine, both outside the view, would give

answers to a human interrogator, preferably in the form of printed response. Based on the answers receive, the interrogator would decide which respondent was the machine. The task of the machine is to deceive the interrogator into believing that he or she is human. If the interrogator cannot reliably distinguish the human from the machine then the machine does possess(artificial) intelligence.


Game playing

Speech Recognition

Understanding Natural Language

Computer Vision

Expert System

Heuristic Classification

Neural Networks




In building expert systems, scientists are attempting to have machines perform intelligent activities at a higher level than most people; after all, few humans are experts.

On the other hand, in trying to create programs that allow computers to understand Natural Language, scientists are trying to teach computers to emulate a skill that nearly all of us perform without any trouble. Oddly enough, while expert systems already have been built that perform at the level of human eperts, computers still cannot understand natural language as well as a typical four-year-old child.

Compared to people, computers required a great deal of precision in communication. For example, if you want some one to bring you something you drink, you might get the same result by saying any of the following:

Bring me a drink.

Got anything to drink?

I'm thirsty-can you get me something?

Unfortunately, a computer does not have that kind of linguistic flexibility. For example, a common "operating system" called MS-DOS requires the following command to instruct a computer to copy all files from one diskette to another:

COPY A: *.*. B:

If you type "Please copy all the files on diskette A to diskette B" instead, the computers does not what you are trying to tell it to do. The computer only accepts an instruction entered precisily in a form that it has been programmed to understand. If you misspell a word, displace a colon, or omit an asterik, the computer cannot execute your instruction properly.


Natural language processing (NLP) is a subfield of artificial intelligence and computational linguistics. It studies the problems of automated generation and understanding of natural human languages.

It is the method of human-computer interaction, natural language processing enables computers to extract meaning from the words and phrases that people use - and respond in kind when presenting information back to them.

Natural language processing(NLP) involves the interpretation, manipulation, or generation of human language by computer so that one can communicate with a computer as one communicates with another person. It studies the problems inherent in the processing and manipulation of natural language.

The result: people can interact more naturally with computer-based information - using normal, familiar expressions - instead of using carefully constructed computer jargon.


The natural language with which we normally communicate is informal and can be extremely


Multiple word meanings.

Syntactic ambiguity.

Unclear antecedents.




even when it is written correctly.


Many of the things we say can be interpreted in more than one way. This ambiguity sometimes results in miscommunication between people and is one of the primary problems in programming computers to understand natural language. Some of the factors that contribute to the ambiguity of natural language are as follows.


It is not uncommon for a single word to have more than one meaning, as in the following sentences.

The pitcher is angry.

The pitcher is empty.

Of course, the word "pitcher" in the first sentence refers to a baseball player, while the "pitcher" in the second sentence is a container designed to hold and pour liquids. Without knowing something about the characteristics of both kinds of pitchers, a computer could not determine which kind of pitcher is meant in each sentence.


Some of the ambiguity in English is caused by peculiarities in its syntax. Consider the following sentence.

I hit the man with the hammer.

How do you interpret the sentence? Did I pick up a hammer and hit a man, or did I hit a man who was holding a hammer? Unless it is able to understand the context in which the sentence appears, a computer may be unable to determine the intended meaning.


We frequently use pronouns in place of previously used nouns. This can create occasional ambiguity, as in the following sentence.

John hit Bill because he sympathized with Mary.

Is John or Bill the antecedent of "he"? In other words, who sympathized with Mary? As in the case of syntactic ambiguity, you cannot determine the antecedent of "he" without establishing a context for the sentence.


People often express concepts with vague and inexact terminology. For example, how long is "a long time"? Consider the following sentences.

I've been waiting in the doctor's office for a long time.

The crops died because it hadn't rained in a long time.

The dinosaurs ruled the earth a long time ago.

If you read a story that included these sentences and then someone asked you about the length of the wait in the doctor's office, you might respond that it was no longer than few hours because you are familiar with the concepts discussed in the sentences. Without that conceptual familiarity, a computer would not be able to differentiate between the three different lengths of time represented by the same phrase.


We do not always say all of what we mean. Because we share common experiences, we usually can omit many details without fear of being misunderstood; we assume that our listeners can "read between the lines".

John went out to a restaurant last night. He ordered steak. When he paid for it, he noticed that he was running out of money.

Did John eat the steak? Although it is not stated explicitly in the story, you probably assumed that he did; after all, why else would he have paid for it? Your expectations of likely events in that particular situation allowed you to understand information that was not included in the text. To be able to comprehend incomplete information, a computer must possess the same kind of situational expectations.


There are two schemes that currently enjoy the greatest popularity for natural language understanding are frames and scripts.


Some psychologists feel that when we mentally recall the image of a particular object, we recall a group of typical attributes of that object at the same time. The form of AI knowledge representation that attempts to emulate this cohesive grouping of attributes is called a frame and was first proposed by Marvin Minsky in the early 1970's. For example, if a furniture salesperson were to say to you, "Follow me, I have a chair that I want you to see, "the word chair would "trigger" a series of expectations in your mind. You would probably expect to see an object with four legs, a seat, a back, and possibly(but not necessarily) two arms. Youl would expect that it would be capable of serving as a place for you to sit. You might not have preconception of any particular color, but you would probably have a general expectation of size. All of your expectations about the attibutes of a chair contribute to your ability to understand what the salesperson means by the word "chair". In an AI program, a CHAIR frame might include knowledge organized as shown in below figure.

Parts : seat, back, legs, arms

Number of legs : 4

Number of arms : 0 or 2

Default : 0

Color : any

Size (in feet):

Height : 2.5-5

Width : 1-3

Depth : 1-3

Styles : dinette, rocking, reclining, office …

Function : a place to sit

Each attribute of the object descibed in the frame in Figure is stored in a separate slot.For example,the Number of arms slot the CHAIR frame is filled with either 0 or 2. Notice that the number of arms slot has a default,an attribute value that is assumed to be in the slot unless there is evidence for the country.


Another knowledge representation system that is especially useful in the area of natural language understanding is a system called scripts, proposed by Roge Schank at Yale University. Scripts are composed of a series of slots that describe, in sequence, the events that we expect to take place in familiar situatins. Just as the concept of frames is based on the assumptions that we have a set of expectations about objects, the use of scripts assumes that we also expect certain sequences of events to occur in particular times and places.So far, the restaurant could be represented as a frame, but here Schank adds a critical concept: Scenes. The script representing your expeience in a restaurant is composd of a series of scenes that represent, in sequence, events that you expect to encounter.Your restaurant script might contain, for example, an entering scene, an ordering scene, and eating scene, and an exiting scene.Each scene has its own script. The script that Schank suggests for entering, for example, appears in the figure.

SCRIPT: Restaurant

SCENE: Entering.

Go into restaurant.

Look at the tables.

Decide where to sit.

Go to table.



Much of the research that has been conducted on natural language understanding has been concerned with the development of NATURAL LANGUAGE INTERFACES(NLI's), programs that allow you to "interface"(communicate) with a computer in everyday English. Also known as NATURAL LANGUAGE FRONT ENDS, NLI's usually include both understanding and generation capabilities so that they can both understand what you type and display text that is easy for you to understand.

In effect, an NLI "stands" between you and the computer as illustrated in the below diagram.

User sees the info on screen.

Computer generates info

User types info.

Computer receives info


NLI translates

it into a ordinary English.

NLI translates it into a form the computer understands



Notice that you actually do not communicate directly with the computer, both you and the computer communicate with the NLI, which translates and forwards the communicated information.

One of the primary uses of NLI's has been to retrieve information from databases. Typically, to request information from a structure your request in a precise format. For example, to request a information service, you might type the following request.


If any word is misspelled, or if the words are not entered in the correct order, or if any parentheses are misplaced, the program does not retrieve the desired information However, if the service were equipped with a natural language interface, you might be able to request the same information in any of the followings ways.

Show me a list of articles about automobile insurance.

What do you have about car insurance?

Anythng about car insurance?

Notice that the last example contains two misspellings and is not a complete sentence. However, you were able to figure out what it meant, so a natural language interface theoretically should be able to do the same.

Currenly, an NLI does not exist that can interpret correctly every request entered in every format. However, natural language technology is improving rapidly, and programs do exist that allow wide flexibility in requesting information from databases.


Some of the natural language interface are








SHRDLU was an early natural language understanding computer program, developed by Terry Winograd at MIT from 1968-1970. It was written in the Micro Planner and Lisp programming language on the DEC PDP-6 computer and a DEC graphics terminal. Later additions were made at the computer graphics labs at the University of Utah, adding a full 3D rendering of SHRDLU's "world".

The name SHRDLU was derived from ETAOIN SHRDLU, the arrangement of the alpha keys on a Linotype machine, arranged in descending order of usage frequency in English. Winograd later distanced himself from SHRDLU and the field of AI, believing SHRDLU a research dead end.


SHRDLU allowed user interaction using English terms. The user instructed SHRDLU to move various objects around in a small "blocks world" containing various basic objects: blocks, cones, balls, etc. What made SHRDLU unique was the combination of four simple ideas that added up to make the simulation of "understanding" far more convincing.

One was that SHRDLU's world was so simple that the entire set of objects and locations could be described by including as few as perhaps 50 words, nouns like "block" and "cone", verbs like "place on" and "move to", and adjectives like "big" and "blue". The possible combinations of these basic language building blocks were quite simple, and the program was fairly adept at figuring out what the user meant.

SHRDLU also included a basic memory to supply context. One could ask SHRDLU to "put the green cone on the red block" and then "take the cone off"; "the cone" would be taken to mean the cone one had just talked about. SHRDLU could search back further to find the proper context in most cases when additional adjectives were supplied. One could also ask questions about the history, for instance one could ask "did you pick up anything before the cone?"

A side effect of this memory, and the original rules SHRDLU was supplied with, is that the program could answer questions about what was possible in the world and what was not. For instance, SHRDLU would deduce that blocks could be stacked by looking for examples, but would realize that triangles couldn't be stacked, after having tried it. The "world" contained basic physics to make blocks fall over, independent of the language parser.

Finally, SHRDLU could also remember names given to objects, or arrangements of them. For instance one could say "a steeple is a small triangle on top of a tall rectangle"; SHRDLU could then answer questions about steeples in the blocks world, and build new ones.



COMPUTER: OK. (Does it)





COMPUTER: OK. (does it)


Natural Language Processing is an important criterion to achieve in the field of Artificial Intelligence. Although today's computers have become much easier to use, communicating with a computer is not yet as simple as the most "natural" kind of communication: communicating with another person. By acheiveing this we can able to communicate with the computer as what we communicate to a normal person.The goal of natural language understanding is not to have computers understand everything we say; after all, even people misunderstand each other occasionally. The goal of natural language understanding is to allow computers to understand people as well as people understand peopleThere are many problems to achieve Natural language processing. If the researchers overcome this problem, we can communicate with computers in the future.