You are on page 1of 4

ASSIGNMENT # 4

SUBMITTED TO:- SUBMITTED BY:-

Ms. Simarpreet Kaur Ramnancy Saini

Roll no: - 41

Section-G27M1

Part-A

Q-1 what do you mean by language? What does natural language processing means and
explain its use in AI using suitable example?

Ans:- Language is basically a medium for communication. With the use of language humans
express themselves .By language we can explore the world with new theories. Natural language
is something which we use in our daily life which has not defined boundaries , set of rules
.expressing yourself in natural language as compare to any more structured language is relatively
easy .Natural language is a very powerful tool for communicating about the world.

Natural language processing: it means making a device which we can understand the natural
language and interprets its correct meaning. Basically ways in which natural language can be
process is processing written task and processing spoken language.

Example: There are many artificial devices that work on natural language processing. One of
such devices is MiPad which means multimodal interactive model. It is a wireless mobile PDA
prototype it is based on processing of spoken language that we call speech recognition devices.
MiPad not only recognizes but understands the command which is given to it via speech. In this
user speaks at a distance and picks up the voice but eliminates the unnecessary noise in
surroundings. So it is processing the natural language.

Q-2 Show conceptual dependency representation of the sentence

John begged Mary for pencil

How does this representation make it possible to answer?

Did John talk to Mary?


Ans:-

John

John MTRANS O Pencil

Mary

Yes with the help of this representation it is easy possible to give the answer. Because in the
above representation john take pencil from Mary and it is also mention that he is begging so it is
obvious that some information from john is passing to Mary and passing with the use of
language that means he has talked to Mary if the word begged has not mention then it could be
thought that without any communication (use of language) he has taken pencil.

Q-3 How Learning makes machine more efficient and intelligent in taking decision?
Explain all its issues step by step.

Ans:- Learning is acquiring the knowledge through taking advice from others, communicating
and sharing information, with experience, by self-educating,. Learning is like intelligence.
Learning makes Machine efficient because it is concerned with the design and development of
algorithms that allow computers to evolve behaviors based on empirical data. Learning make
possible to solve complex problems and make intelligent decisions by defining problem and
collecting the data.
• Learning is used in recognition i.e. machines are used that is based on intelligence. These
system are used in face recognition and verifying the signatures that is used for security
purposes.
• Making decision tree we can reach the target value. In this case, graph or maps are
defined and making decision how to get the target.
• Learning is concerned with how an agent ought to take actions in an environment so as
to maximize some notion of long-term reward.
• In robots or in intelligent system there is also use learning that how to do work. For
example in turing test machine decide that who is lying.
• Learning is used as genetic learning that is used to find the optimal task that a user can
do. It is used to optimize a population of computer programs according to a fitness
landscape determined by a program's ability to perform a given computational task.
Part-B

Q-4 Explain why inductive learning should require more inference than learning by being
told.

Ans:- The inductive learning is basically defining the classes. In case of inductive learning
whatever thing we added we need to classify a class for them. It could be either added in an
existing class or we can also develop a new class for it. Inductive learning can also be called as
learning from examples .in inductive learning there is a teacher involved .In case of inductive
learning the system induces a rule itself from a collection of observed instances (objects).In case
of learning by told it is simply giving the information to the system . The programmer acts as a
teacher whereas the system acts a student. But it does not understand the information that is
being fed into it. So we say that inductive learning require more inference because if the system
is based on the inductive learning it will try to infer the meaning of information being fed into it
by the classes that are already defined. So we can say that it requires more additional information
and it is much more certain than the learning by told.

Q-5 How does semantic Networks function as an alternative to Predicate Logic?

Ans:- Semantic network are just the graphical representation of the given sentence. , we can say
that semantic network is just alternative to predicate logic because both logic and semantic
network are the way of representing the knowledge. The difference is just that semantic
network uses the arc and nodes whereas logic uses the Boolean functions like AND, OR, NOT.

In case of predicate logic the representation is quite simple but somehow difficult to understand.
It is difficult to know relation between entities, classes.

For e.g. a statement like Marcus is a Pompeian will represented as follows:

In predicate logic:

Pompeian (Marcus)
Pompeian

Instance

Marcus

In the representation of semantic network to find the relation is easy i.e. it simply tells that
Pompeian is representing a class and Marcus is an entity of that class.
Q-6 How Expert system helps in image processing? Make a case study on it and show how
its different component works in it.

Ans:-Expert system is intelligent computer programs that are capable of offering solution related
to specific problem. Many expert systems are used to diagnose disorders from observation
symptoms. Image processing system is used to digitize and process serial sections of CT/MRI
scan and reconstruct three-dimensional images of brain structures and brain lesions. A
knowledge-based system employing rule-based programming has been built to help identifying
brain lesions and to help planning trajectory for operations. The volumes of the lesions are also
automatically determined. Such system is very useful for medical skills archival, tumor size
monitoring, survival and outcome forecasting, and consistent neurosurgical planning. Expert
system can reach a correct and accurate diagnosis through extracting symptoms from images.

For example:- to find out defects in a leaf we use image processing in expert system. Expert
system can reach a correct diagnosis through images. Expert systems analyze the difference from
other images and find correct disease for leaf. Expert system apply reasoning process while
taking into extracted symptoms.

You might also like