DAYANANDA SAGAR ACADEMY OF TECHNOLOGY & MANAGEMENT

(Affiliated to Visvesvaraya Technological University, Belagavi & Approved by AICTE, New Delhi)
Accredited 3 years by NBA, New Delhi ( Validity : 26-07-2018 to 30-06-2021)
Udayapura, Opp. Art of living, Kanakapura road, Bangalore-082

DEPARTMENT OF INFORMATIONSCIENCE & ENGINEERING

SUBJECT: Artificial Intelligence SUB CODE: 15CS562
CLASS: V Sem ISE Date: 02-11-2018
COURSE OUTCOMES:
CO3: Analyze different Reasoning techniques in the context of AI.
CO4: Apply AI search techniques for solving various problems of Game Playing.
CO5: Differentiate between different forms of AI learning methods.

ASSIGNMENT-III
1. a. Write a note on semantic nets and represent Non-binary predicates using semantic nets.
(04 Marks, NS)(CO3)
Semantic Nets were originally designed as a way to represent the meaning of English words. The main
idea is that the meaning of a concept comes from the ways in which it is connected to other
concepts. In semantic nets, the information is stored by interconnecting nodes with labeled arcs.
Semantic nets initially were used to represent labeled connections between nodes. The Semantic
Nets can be represented in different ways by using relationships. Semantic nets have been used to
represent a variety of knowledge in a variety of different programs. An example of semantic net
is shown below:
These values can also be represented in logic as: isa(person, mammal), instance(Pee-Wee-Reese, person,
team(Pee-Wee-Reese, Brooklyn-Dodgers)
Additional relation could be derived using inheritance- has-part(Pee-Wee-Reese, Nose)

Representing Non-binary predicates using semantic nets: Semantic nets are natural way to represent
relationships that would appear as ground instances of binary predicates in predicate logic. Arcs
from the above figure could be represented in logic as
◦ isa(Person, Mammal)
◦ instance(Pee-Wee-Reese, Person)
◦ team(Pee-Wee-Reese, Brooklyn-Dodgers)
◦ uniform-color(Pee-Wee-Reese, Blue)
Many unary predicates in logic can be thought as binary predicates using isa and instance.
Example: man(Marcus) → instance(Marcus, man)

1
 Three or more place predicates can also be converted to a binary form by creating one new object
representing the entire predicate statement and then introducing binary predicates to describe the
relationship to this new object of each of the original arguments

Example: score(Cubs, Dodgers, 5-3) can be represented in semantic net by creating a node to
represent the specific game & then relating each of the three pieces of information to it.

b. Construct semantic net representations for the following: (a) Pompeian(Marcus),

Blacksmith(Marcus) (b) Every dog in town has bitten the constable. (04 Marks, NS)(CO3)
(a) Pompeian(Marcus), Blacksmith(Marcus)
● Pompeian(Marcus) → instance(Marcus, Pompeian)
● Blacksmith(Marcus) → instance(Marcus, Blacksmith)

Pompeian Blacksmith

instance instance

Marcus Marcus

(b) Every dog in town has bitten the constable: In this net, the node c representing the victim lies
outside the form of the general statements. Thus it is not viewed as an existentially quantified
variable whose value may depend on the value of d.

2. a. Explain Frame w.r.t slot & filler data structures of AI. Explain a simplified Frame system with
an example of league basketball team. (04 Marks, NS)(CO4)

● A frame is a collection of attributes (called slots) and associated values that describe some
entity in the world.

2
 ● Frames on their own are not particularly helpful but frame systems are a powerful way of
encoding information to support reasoning.
● Frame is a type of schema used in many AI applications including vision and natural
language processing.
● Frames provide a convenient structure for representing objects that are typical to stereotypical
situations. The situations to represent may be visual scenes, structure of complex physical
objects, etc.
● Frames are also useful for representing commonsense knowledge.

● The isa relationship is in fact the subset relation.

◦ The set of adult males is a subset of the set of people.
◦ The set of major league baseball players is a subset of adult males and so forth
● The instance relation corresponds to the relation element-of.
◦ Pee-Wee-Reese is the element of Fielder.

b. List all the properties used to represent slot as full-fledged objects. (04 Marks, NS)(CO4)
Some of the properties used to represent slot as full-fledged objects include:
● The class to which the attribute can attach.

3
 ● Constraints on either the type or the value of the attribute.
● A default value for the attribute. Rules for inheriting values for the attribute.
● To be able to represent these attributes of attributes, we need to describe attributes (slots)
as frames.
● These frames will organize into a isa hierarchy, just as any other frames, and that
hierarchy can then be used to support inheritance of values for attributes of slots.
● Define the slot. A slot here is a relation.
● It maps from alements of its domain (the classes for which it makes sense) to elements of
its range (its possible values).
● A relation is a set of ordered pairs. Thus it makes sense that relation R1 is a subset of
another relation R2.
● In that case, R1 is a specialization of R2. Since a slot is a set, the set of all slots, which we
will call SLOT, is a meta-class.
● Its instance are slots, which may have sub-slots.

3. a. Give an example of Baseball league for representing relationships among classes.

(04 Marks, NS)(CO4)

4
 The above diagram shows two orthogonal ways of decomposing the class of Major League
Baseball players. Everyone is a pitcher, a catcher, or a fielder (and no one is more than one of
these). In addition, everyone plays in either the National League or the American League, but
not both.
b. Write a note on tangled hierarchies giving an example. (04 Marks, NS)(CO4)
Hierarchies that are not simple trees are called Tangled hierarchies. These allow another type of
inheritance conflict. For example:

We want to decide whether Oliver can fly. The correct answer is no. Although birds in general can fly,
the subset of birds, ostriches do not. Although the class Pet-Bird provides a path from Oliver to
Bird and thus to the answer that Oliver can fly, it provides no information that conflicts with the
class Ostrich. To handle this case correctly, we need an algorithm for traversing the algorithm
which guarantees that specific knowledge will always dominate over general knowledge.

4. a. What is Conceptual Dependency? Explain how the sentence “I gave the man a book” be
represented using CD. (04 Marks, NS)(CO4)
Conceptual Dependency (CD) is the theory of how to represent the kind of knowledge about events that
is usually contained in natural language sentences. The goal is to represent the knowledge in a
way that
● Facilitates drawing inferences from the sentences.
● Is independent of the language in which the sentences are originally stated.
CD provides a structure into which nodes representing information can be placed a specific set of
primitives at a given level of granularity.
Representation of Conceptual Dependency: I gave the man a book

5
In CD representation of actions are built from a set of primitive Acts.

b. Explain how the sentence “Bill threatened John with a broken nose” be represented using CD.
(04 Marks, NS)(CO4)

6
The CD representation says that Bill informed John that he (Bill) will do something to break John’s
nose.

5. a. What is a Script? Describe the two ways of activating a Script. Illustrate the important
components of a script. List the two important uses of script in interpreting a particular
situation. (04 Marks, NS)(CO5)
A script is a structure that prescribes a set of circumstances which could be expected to follow on from
one another. It is similar to a thought sequence or a chain of situations which could be
anticipated. It could be considered to consist of a number of slots or frames but with more
specialized roles.

Scripts provide an ability for default reasoning when information is not available that directly states that
an action occurred. So we may assume, unless otherwise stated, that a diner at a restaurant was
served food, that the diner paid for the food, and that the dinner was served by a waiter/waitress.

The two ways of activating a Script: i) For fleeting scripts (ones that are mentioned briefly and may be
referred to again but are not central to the situation), it is sufficient to store a pointer to the script
so that it can be accessed later if necessary).
ii) For non-fleeting scripts it is appropriate to activate the script fully and fill the slots with particular
objects and people involved in the situation.
The important components of a script are:

● Entry Conditions -these must be satisfied before events in the script can occur.

● Results Conditions that will be true after events in script occur.

● Props Slots representing objects involved in events.

● Roles Persons involved in the events.

● Track Variations on the script. Different tracks may share components of the same
script.

● Scenes The sequence of events that occur. Events are represented in

conceptual dependency form.

Important uses of script: Scripts are useful in describing certain situations such as robbing a bank. This
might involve:
● Getting a gun.
● Hold up a bank.
● Escape with the money.

7
b. What is CYC? List the motivations and methods for acquiring and building large knowledge
bases. (04 Marks, NS)(CO5)
CYC is a very large knowledge base project aimed at capturing human commonsense knowledge. The
goal of CYC is to encode the large body of knowledge that is so obvious that it is easy to forget
to state it explicitly. Such a knowledge base could then be combined with specialized knowledge
bases to produce systems that are less brittle than most of the ones available today.
Motivations for building large knowledge bases:

● Brittleness-- Specialized knowledge bases are brittle. Hard to encode new situations and
non-graceful degradation in performance. Commonsense based knowledge bases should have
a firmer foundation.

● Form and Content-- Knowledge representation may not be suitable for AI. Commonsense
strategies could point out where difficulties in content may affect the form.
● Shared Knowledge-- Should allow greater communication among systems with common
bases and assumptions.

Building an immense knowledge base is a staggering task. There are two possibilities for acquiring
this knowledge automatically:

8
 1. Machine Learning: In order for a system to learn a great deal, it must already know a great
deal. In particular, systems with a lot of knowledge will be able to employ powerful
analogical reasoning.

2. Natural Language Understanding: Humans extend their own knowledge by reading

books and talking with other humans.

6. Explain the Minimax search algorithm with an example. (08 Marks, NS)(CO5)
Minimax is a kind of backtracking algorithm that is used in decision making and game theory to find the
optimal move for a player, assuming that your opponent also plays optimally. It is widely used in
two player turn-based games such as Tic-Tac-Toe, Backgammon, Mancala, Chess, etc.

In Minimax the two players are called maximizer and minimizer. The maximizer tries to get the highest
score possible while the minimizer tries to do the opposite and get the lowest score possible.

Every board state has a value associated with it. In a given state if the maximizer has upper hand then,
the score of the board will tend to be some positive value. If the minimizer has the upper hand in
that board state then it will tend to be some negative value. The values of the board are calculated
by some heuristics which are unique for every type of game.

Consider a game which has 4 final states and paths to reach final state are from root to 4 leaves of a
perfect binary tree as shown below. Assume you are the maximizing player and you get the first
chance to move, i.e., you are at the root and your opponent at next level. Which move you would
make as a maximizing player considering that your opponent also plays optimally?

Since this is a backtracking based algorithm, it tries all possible moves, then backtracks and makes a
decision.
● Maximizer goes LEFT: It is now the minimizers turn. The minimizer now has a choice
between 3 and 5. Being the minimizer it will definitely choose the least among both, that
is 3
● Maximizer goes RIGHT: It is now the minimizers turn. The minimizer now has a choice
between 2 and 9. He will choose 2 as it is the least among the two values.
Being the maximizer you would choose the larger value that is 3. Hence the optimal move for the
maximizer is to go LEFT and the optimal value is 3.
9
Now the game tree looks like below :

The above tree shows two possible scores when maximizer makes left and right moves.
Note: Even though there is a value of 9 on the right subtree, the minimizer will never pick that. We must
always assume that our opponent plays optimally.
Here, we assume that static evaluation function returns larger values to indicate good situation for us. So
our goal is to maximize the value of the static evaluation function of the next board position. The
opponent’s goal is to minimize the value of the static evaluation function.
The alternation of maximizing and minimizing at alternate ply corresponding to the opposing strategies
of the two players is called MINIMAX. It is a recursive procedure that depends on two
procedures.
● MOVEGEN(position, player) – The plausible-move generator, which returns a list of
nodes representing the moves made by the Player in position.
● STATIC(position, player) – static evaluation function, which returns a number
representing the goodness of Position from the standpoint of Player.
With any recursive program, we need to decide when recursive procedure should stop.
● DEEP-ENOUGH function is used to return TRUE if the search should be stopped at the
current level and FALSE otherwise.
● It takes two parameters, position and depth. It will ignore its position parameter and
simply return TRUE if its depth parameter exceeds a constant cut off value.
● One problem that rises in defining MINIMAX as a recursive procedure is that it needs to
return not one but two results.
● The backed-up value of the path it chooses.
● The path itself.
● We assume that MINIMAX returns a structure containing both results and two functions
VALUE and PATH that extracts the separate components.
● Initially, it takes three parameters, a board position, the current depth of the search and
the player to move,

10
 o MINIMAX(current, 0, player-one) If player-one is to move
o MINIMAX(current, 0, player-two) If player-two is to move

7. Explain the four additional refinements used to improve the performance of Minimax procedure.
(08 Marks, NS)(CO5)
The four additional refinements used to improve the performance of Minimax procedure are:
(i) Waiting for Quiescence
(ii) Secondary search
(iii) Using Book moves and
(iv) Alternatives to Minimax
Waiting for Quiescence: One of the factors that should be considered is determining when to
stop going deeper in the search tree. Consider the tree shown below (Fig 12.7). Suppose the node
B is expanded one more level, the result will be as shown in Fig. 12.8, i.e the estimation of the
worth of B changed drastically. If we stop exploring the tree at this level, we assign the value -4
to B and therefore decide that B is not a good move.
To make sure that such short-term measures do not influence our choice of move, we should
continue the search until no such drastic change occurs from one level to the next. This is called
waiting for Quiescence. If we do that, we may get the situation shown in Fig 12.9, in which the
move to B again looks like a reasonable move for us to make since the other half of the piece
exchange has occurred.

11
 Secondary search: After choosing a move, search a few more levels beneath it to be sure it still
looks good. Suppose we explore a game tree to an average depth of six ply and on the basis of
search choose a particular move. Although it is expensive to search the entire tree to a depth of
eight ply, it will not be very expensive to search an additional two levels from the single chosen
branch. This technique is called secondary search.

Using Book Moves: For some parts of the complicated game (especially initial and end moves),
keep a catalog of best moves to make. For example, in chess, both opening and endgame
sequences are highly stylized. In these situations, the performance of a program can be
considerably enhanced if it is provided with a list of moves (called Book moves) that should be
made. The use of book moves in the opening and endgame sequences, combined with the use of
minimax search procedure for the midgame provides a good example of the way that knowledge
and search can be combined in a single program to produce more effective results.

Alternatives to Minimax: Even with the refinements mentioned above, minimax still has some
problematic aspects. For instance, it relies heavily on the assumption that the opponent will
always choose the optimal move. This assumption is acceptable in winning situations, but in a
losing situation it is better to take the risk that the opponent will make a mistake. Suppose we
must choose between two moves, both of which, if the opponent plays perfectly, lead to
situations that are very bad for us, but one is slightly better than the other. Further, let us consider
that the less promising move could lead to a very good situation for us. To make these decisions
well, we must have access to the models of the individual opponents’ playing style, so that the
likelihood of various mistakes can be estimated. This is very hard to provide in real world.

8. Explain the various steps involved in Natural Language Processing. (08 Marks, NS)(CO5)
12
Natural Language Processing is:
(i) Processing written text, using lexical, syntactic and semantic knowledge of the language
as well as the required real-world information.
(ii) Processing spoken language, using all the information needed above plus additional
knowledge about phonology as well as enough added information to handle the further
ambiguities that arise in speech.

The various steps involved in in Natural Language Processing are:

(i) Morphological Analysis: Individual words are analyzed into their components and non-word
tokens such as punctuation are separated from the words.
Suppose we have an English interface to an operating system and the following sentence typed:
I want to print Bill’s .init file.
The morphological analysis must do the following things:
• Pull apart the word “Bill’s” into proper noun “Bill” and the possessive suffix “’s”
• Recognize the sequence “.init” as a file extension that is functioning as an adjective in the
sentence.
• This process will usually assign syntactic categories to all the words in the sentence.

(ii) Syntactic Analysis: Linear sequences of words are transformed into structures that show how the
words relate to each other. Some word sequences may be rejected if they violate the language’s
rule for how words may be combined.

• A syntactic analysis must exploit the results of the morphological analysis to build a structural
description of the sentence.
• The goal of this process, called parsing, is to convert the flat list of words that form the sentence
into a structure that defines the units represented by the flat list.
• The important thing here is that a flat sentence has been converted into a hierarchical structure.
The structure corresponds to meaning units when a semantic analysis is performed.
• Reference markers (set of entities) are shown in parenthesis in the parse tree (refer to the
diagram).
• Each reference marker corresponds to some entity that has been mentioned in the sentence.
• These reference markers are useful later since they provide a place to accumulate information
about the entities as we get it.

13
(iii) Semantic Analysis: The structures created by the syntactic analyzer are assigned meanings. A
mapping is made between the syntactic structures and objects in the task domain. Structures for
which no such mapping is possible may be rejected.
The semantic analysis must do two important things:
1. It must map individual words into appropriate objects in the knowledge base or database.
2. It must create the correct structures to correspond to the way the meanings of the individual
words combine with each other.

(iv) Discourse Integration: The meaning of an individual sentence may depend on the sentences that
precede it. And also, may influence the meanings of the sentences that follow it.
• Specifically, we do not know whom the pronoun “I” or the proper noun “Bill” refers to.
• To pin down these references requires an appeal to a model of the current discourse context, from
which we can learn that the current user is USER068 and that the only person named “Bill” about
whom we could be talking is USER073.
• Once the correct referent for Bill is known, we can also determine exactly which file is referred
to.

(v) Pragmatic Analysis: The structure representing what was said is reinterpreted to determine what
was actually meant.
• The final step toward effective understanding is to decide what to do as a result.
• One possible thing to do is to record what was said as a fact and done with it.
• For some sentences, the intended effect is clearly declarative, that is the precisely correct thing to
do.
• But for other sentences, including this one, the intended effect is different.
• We can discover this intended effect by applying a set of rules that characterize cooperative
dialogues.
• The final step in pragmatic processing is to translate, from the knowledge-based representation to
a command to be executed by the system.

Results of each of the main processes combine to form a natural language system. All of the processes
are important in a complete natural language understanding system. Not all programs are written
with exactly these components. Sometimes two or more of them collapsed.

9. Write a note on Augmented Transition Networks (ATN). (08 Marks, NS)(CO5)

● An augmented transition network is a top-down parsing procedure that allows various kinds
of knowledge to be incorporated into the parsing system so it can operate efficiently.
● ATNs build on the idea of using finite state machines (Markov model) to parse sentences.

● Instead of building an automaton for a particular sentence, a collection of transition graphs

are built.

● A grammatically correct sentence is parsed by reaching a final state in any state graph.

● Transitions between these graphs simply use subroutine calls from one state to any initial
state on any graph in the network.

14
 ● A sentence is determined to be grammatically correct if a final state is reached by the last
word in the sentence.

● The ATN is similar to a finite state machine in which the class of labels that can attach to the
arcs that define the transition between states has augmented.
Arcs may label with:

◦ Specific words such as “in’.

◦ Word categories such as noun.
◦ Procedures that build structures that will form part of the final parse.

◦ Procedures that perform arbitrary tests on current input and sentence components that
have identified.

10. a. Explain learning by taking advice and Winston’s learning program. (04 Marks, NS)(CO5)
Learning By Taking Advice: This is a simple form of learning. Suppose a programmer writes a set of
instructions to instruct the computer what to do, the programmer is a teacher and the computer is
a student. Once learned (i.e. programmed), the system will be in a position to do new things.

The advice may come from many sources: human experts, internet to name a few. This type of learning
requires more inference than rote learning. The knowledge must be transformed into an
operational form before stored in the knowledge base. Moreover the reliability of the source of
knowledge should be considered.

The system should ensure that the new knowledge is conflicting with the existing knowledge. FOO (First
Operational Operationalizer), for example, is a learning system which is used to learn the game
of Hearts. It converts the advice which is in the form of principles, problems, and methods into
effective executable (LISP) procedures (or knowledge). Now this knowledge is ready to use.

Winston’s learning program: Winston (1975) described a Blocks World Learning Program. This
program operated in a simple blocks domain. The goal is to construct representation of the
definition of concepts in the blocks domain.
Example: Concepts such as “house”
● Start with input, a line drawing of a blocks world structure.
It learned concepts House, Tent, Arch as:
brick ( rectangular block) with a wedge (triangular block) suitably placed on top of it,
tent – as 2 wedges touching side by side,
arch – as 2 non-touching bricks supporting a third wedge or brick.
● The program for each concept is learned through near miss. A near miss is an object that
is not an instance of the concept but a very similar to such instances.
● The program uses procedures to analyze the drawing and construct a semantic net
representation.
● An example of such a structural for the house is shown below.

15
 ● Node A represents entire structure, which is composed of two parts: node B, a wedge and
node C, a brick.
Links in network include supported-by, has-part and isa. Winston’s program followed 3 basic
steps in concept formulation:
1. Select one known instance of the concept. Call this the concept definition.
2. Examine definitions of other known instance of the concept. Generalize the definition
to include them.
3. Examine descriptions of near misses. Restrict the definition to exclude these.
● Both steps 2 and 3 of this procedure rely heavily on comparison process by which
similarities and differences between structures can be detected.
● Winston’s program can be similarly applied to learn other concepts such as “ARCH”.

b. Explain the various techniques of learning in problem solving. (04 Marks, NS)(CO5)
Classification or taxonomy of learning types serves as a guide in studying or comparing the differences
among them. One can develop learning taxonomies based on the type of knowledge
representation used (predicate calculus, rules, frames), the type of knowledge learned (concepts,
game playing, problem solving), or by the area of application (medical diagnosis, scheduling,
prediction and so on).

The classification is intuitively more appealing and is one which has become popular among machine
learning researchers. It is independent of the knowledge domain and the representation scheme
used. It is based on the type of inference strategy employed or the methods used in the learning
process. The five different learning methods under this taxonomy are:
Memorization (rote learning)
Direct instruction (by being told)
Analogy
Induction
Deduction

Learning by memorization is the simplest form of learning. It requires the least 5 amount of inference
and is accomplished by simply copying the knowledge in the same form that it will be used
directly into the knowledge base. We use this type of learning when we memorize multiplication
tables, for example.

A slightly more complex form of learning is by direct instruction. This type of learning requires more
understanding and inference than rote learning since the knowledge must be transformed into an
operational form before being integrated into the knowledge base. We use this type of learning
when a teacher presents a number of facts directly to us in a well-organized manner.

The third type listed, analogical learning, is the process of learning a concept or solution through the use
of similar known concepts or solutions. We use this type of learning while solving problems in
an examination where previously learned examples serve as a guide or when we learn to drive a
16
 truck using our knowledge of car driving. We make use of analogical learning. This form of
learning requires more inferring than either of the previous forms, since difficult transformations
must be made between the known and unknown situations. This is a kind of application of
knowledge in a new situation.

The fourth type of learning is also one that is used frequently by humans. It is a powerful form of
learning which, like analogical learning, also requires more inferring than the first two methods.
This form of learning requires the use of inductive inference, a form of invalid but useful
inference. We use inductive learning when we formulate a general concept after seeing a number
of instance or examples of the concept. For example, we learn the concepts of color, sweet taste,
after experiencing the sensation associated with several examples of colored objects or sweet
foods.

The final type of acquisition is deductive learning. It is accomplished through a sequence of deductive
inference steps using known facts. From the known facts, new facts or relationships are logically
derived. Deductive learning usually requires more inference than the other methods. The
inference method used is, of course, a deductive type, which is a valid form of inference.

In addition to the above classification, we will sometimes refer to learning methods as weak methods or
knowledge-rich methods. Weak methods are general purpose methods in which little or no initial
knowledge is available. These methods are more mechanical than the classical AI knowledge –
rich methods. They often rely on a form of heuristics search in the learning process.

*****************************

The alpha-beta procedure can speed up a depth-first minimax search.

• Alpha: a lower bound on the value that a max node may ultimately be assigned v >= α
• Beta: an upper bound on the value that a minimizing node may ultimately be assigned v <= β

17