Exploring and Expanding the World of Artificial Intelligence

Abstract

This work is an effort to explore the available applications of Artificial Intelligence, in conjugation with other computing applications and put them to interesting and meaningful use through simple modeling. Tic-Tac-Toe model is considered as a suitable candidate for this experiment due to its simple yet versatile facets.

Full text

International Journal of Computer Applications (0975 – 8887)
Volume 134 – No.2, January 2016
1
Exploring and Expanding the World of Artificial
Intelligence
Ridip De
B.Tech, 4th Year,
Computer Science and
Engineering,
SRM University, India
R.P. Mahapatra, PhD
Professor and HOD,
Computer Science and
Engineering,
SRM University, India
Partha Sarathi Chakraborty
Assistant Professor, Member
IEEE,
Computer Science and
Engineering,
SRM University, India
ABSTRACT
This work is an effort to explore the available applications of
Artificial Intelligence, in conjugation with other computing
applications and put them to interesting and meaningful use
through simple modeling. Tic-Tac-Toe model is considered as
a suitable candidate for this experiment due to its simple yet
versatile facets.
General Terms
Artificial Intelligence, Speech Recognition, Text-to-Speech
(TTS), Tic-Tac-Toe Algorithms and Modes.
Keywords
AI, Speech Recognition, Text-to-Speech, Algorithm, Tic-Tac-
Toe modes, Smart Tic-Tac-Toe.
1. INTRODUCTION
Artificial intelligence can be well-defined as the capability of
a digital computer to execute tasks commonly associated with
intelligence of the human beings. Artificially Intelligent
systems are endowed with the intellectual process traits of
humans. Such systems have the ability to think, reason out
their action, learn from past experience, ascertain meaning or
generalize. Humans have always been aiming to develop
systems with human-level intelligence.
McCarthy, the father of AI, coined the term in 1955. He
defined Artificial intelligence as "the science and engineering
of making intelligent machines".
Artificial Intelligence can be used in various aspects of day to
day life be it stock trading, remote sensing, maintaining law,
scientific discovery, medical diagnosis etc. One of the major
uses of AI can be easing how humans interact with the
computer which not only does make the human life easier but
would also help the differently-abled persons to interact with
today‟s high end machines.
To demonstrate the above mentioned major use, Tic-Tac-Toe
model can be considered as a suitable candidate for this
experiment due to its simple yet versatile facets. There are a
couple of ways of developing the algorithm to simulate the
algorithm for “Tic-Tac-Toe”. Our approach has adopted to
aim at the following:
i. Explore the available options under Artificial
Intelligence which can be put to interesting and
meaningful use.
ii. Designing a user friendly user interface.
iii. To help differently abled interact with computers.
2. STRATEGIES
Contrary to a popular belief, tic-tac-toe isn't purely a
game of chance. There are some strategies that can enhance
the chances of victory. The first most move will decides the
destiny of the game. The following points describe in brief
about the various strategies that one must keep in mind before
making the first move:
1. The first move should be in one of the four corners.
Doing so will reduce the predictive-ness of the game by
the opponent.
2. Avoid the edges as a first move (edges are the four
boxes that are neither the center nor the corner)
3. Placing the first move at the center will restrict victory
from other positions
3. EXISTING MODELS
Several models and algorithms exist to develop Tic-Tac-Toe.
The common one being is MinMax algorithm, which was
initially developed for two-player zero-sum game theory.
MinMax algorithm is a decision theory for reducing the
probable loss. It is widely used in decision theory, game
theory, statistics and philosophy.
The main drawback of this algorithm, while developing a
game like Tic-Tac-Toe is that the moves from the computer is
quite predictable and monotonous thus making the gaming
experience unrealistic.
To maintain spontaneity of the game we decided not to adapt
existing algorithm.
Fig 1: Possible Moves

International Journal of Computer Applications (0975 – 8887)
Volume 134 – No.2, January 2016
2
The game tree starts with the initial position and consecutive
nodes contain the possible unique ways the game can be
played from each position. Overall a complete game tree for
3x3 Tic-Tac-Toe has 255168 leaf nodes. The partial game tree
for Tic-Tac-Toe is shown in fig. 2.
Fig 2: Game Tree for Tic-Tac-Toe
4. UPGRADED MODEL
We have developed an algorithm for our Smart Tic-Tac-Toe
that makes the game experience more realistic and give a feel
of playing against another human. The new algorithm is
designed using a combination of recursion and backtracking
algorithm techniques. A method CheckResult() was created to
concurrently check for the result of the game, after each
move. The first move being the crucial one, has to be
meticulously made as discussed in section 2. Our algorithm
follows the first move strategy.
This new model depicts an environment that makes use of
Artificial Intelligence which can be used by the differently
abled people to interact with the computers. The model makes
use of Touch interface, Speech Recognition System and Text
To Speech (TTS) technologies which empowers a system to
communicate and interact with the user.
Our algorithm for Tic-Tac-Toe has been architectured using
two different modes. How the computer reacts to the
opponent‟s move, depends on the mode selected:
1) Defensive Mode: This is the easy play mode for
beginners. Playing in this mode will allow the
computer to only defend the opponent‟s moves. Its
goal will not be to win but to prevent the opponent
from winning. The algorithm would make its move
so that the opponent cannot win even if it had a
chance of winning the game.
Example: if there exists such a scenario as in fig
3(a), the computer in DEFENDING mode would
respond by playing „O‟ at position 6 instead of
position 4 to defend the opponent from winning fig
3(b).
(a)
(b)
Fig 3: Defensive Mode Algorithm
2) Attacking Mode: This is the harder mode. The main
intention of the computer shifts to the other side of
just defending. Its motive will be to win the game.
The algorithm would never miss an opportunity to
win a game.
Example: if there exists such a scenario as in figure fig 4a),
the computer in ATTACKING mode would respond by
playing „O‟ at position 4 to win the game fig. 4(b).
(a)

International Journal of Computer Applications (0975 – 8887)
Volume 134 – No.2, January 2016
3
(b)
Fig 4: Attacking Mode Algorithm
4.1 Integrating Speech Recognition System
Vocal Speech is the natural means of humans to
communicate. Computation will go to the next level if one
will be able to communicate with the computers completely
with his or her voice.
Fig. 5 depicts the general working principle of the Speech
Recognition System. When a user speaks a word, it needs to
be converted it into a digital form so that it can be recognized
by the computer. The speech recognition engine based on the
set of grammar identifies the user‟s words which can later be
used for computation and further processing.
Fig 5: Speech Recognition Working
There are various API‟s available in the market. One of such
is Microsoft‟s Speech Engine would help me to achieve the
same.
First associate Microsoft‟s Speech Engine with your project.
Doing so would enable us to use the Bing.Speech namespace.
Now, instantiate the SpeechRecognizer class as follows:
SpeechRecognizer SR = new SpeechRecognizer("en-US",
credentials);
The SR is the object of the SpeechRecognizer class and the
credentials are the access tokens that I was provided by
Microsoft before associating the Bing.Speech package with
my project. The SpeechRecognizer property binds the
SpeechControl control to an object of the
SpeechRecognizer class. This step is required for the
SpeechControl control to function.
SpeechControl.SpeechRecognizer = SR;
The various properties and methods associated with the
SpeechRecognizer class are mentioned below:
Dispose() : Disposes the SpeechRecognizer object.
RequestCancelOperation() : intrudes speech recognition. The
control now returns to the caller
RecognizeSpeechToTextAsync() : to initiate speech
recognition session, which captures and interprets user
speech, and then returns the results as a
SpeechRecognitionResult object.
With the help of an asynchronous function which keeps
running in passive mode, SR.RecognizeSpeechToTextAsync()
method converts the Speech from the user to Text which is
then used to respond to the users actions.
4.2 The voice of Computer
The voice of the computer has been achieved by using
“Text to speech” technology by using
Windows.Media.SpeechSynthesis;
After instantiating the SpeechSynthesizer class, activate the
voice for the computer using:
SpeechSynthesizer synthesizer = new SpeechSynthesizer();
The SpeechSynthesisStream class is instantiated to enable
reading and writing audio data generated by the speech
synthesis engine (voice) to/from a random access stream:
SpeechSynthesisStreamsynthesisStream = await
synthesizer.SynthesizeTextToStreamAsync(text_to_speak);
The method SynthesizeTextToStreamAsync asynchronously
generates voice output from a string which is passed as a
parameter the method SynthesizeTextToStreamAsync(). Here
the text_to_speak is the string that has to be spoken out by the
computer.
5. POSSIBLE FURTHER
UPGRADATION
1. Increase the grid size:
Moving from 3x3 grid size to say 5x5 grid would not only
make the gaming entirely different but will also increase the
difficulty level of the game and the complexity of developing
it would make it more enjoyable. This would also require the
beginning and consequent moves to be strategized out wisely.
2. AI learning:
If successfully introduced, defeating the computer would be
next to impossible. Victory and future will always belong to
the computer.
3. Play over internet:
Taking the gaming online, would allow people to find players
and compete against each other. There should also be a
dedicated scoring system to store the details of the players.
4. Pattern Recognition:
Introducing Pattern Recognition System would enhance the
chances of victory of computer against the user.
6. CONCLUSION
The Smart Tic-Tac-Toe developed by me may appear similar
to the familiar game. But the following unique features have
been introduced:
 A completely new algorithm, which makes the
moves unpredictable and gives a sigh of playing
against human.
 Speech recognition and Text to speech technology
have been exploited and integrated in this model, to
enable the visually impaired people to interact with
computers.

International Journal of Computer Applications (0975 – 8887)
Volume 134 – No.2, January 2016
4
The game which was accorded a gold rating from the
Microsoft‟s Windows AppLab team and can be downloaded
for free from the Windows App Store under the name “Smart
Tic-Tac-Toe”.
An enormous scope still exists, how Artificial Intelligence can
enhance and improve the Human Machine interaction.
7. ACKNOWLEDGMENTS
The logic, programs and experiments were developed at
Department of Computer Science and Engineering, SRM
University. The author acknowledges the guidance from co-
authors Dr. R. P. Mahapatra (Professor and HOD) and Mr.
Partha Sarathi Chakraborty (Asst. Professor).
8. DOWNLOAD LINK
The application developed by us is suitable for laptops and
tablets running windows 8.1 or above. It can be downloaded
for free from the Windows App Store. The link to which is:
http://apps.microsoft.com/windows/en-us/app/68556ec4-
de47-487a-a55f-e3571fde2175
or
http://goo.gl/d3JGq6
9. REFERENCES
[1] Lalitha Saraoja Thota, Manal Elsayeed, Suresh Babu,
“Implementation of Tic-Tac Toe Game in LabView”,
International Journal of Computer Trends and
Technology Volume 12, June 2 2014
[2] Preeti Saini, Pameet Kaur, “Automatic Speech
Recognition: A Review”, International Journal of
Engineering Trends and Technology, Volume 4, Issue 4,
2013.
[3] Santosh k. Gaikwad, Bharti W. Gawali, Pravin
Yannavar, “A review on Speech recognition
Techniques”, Internation Journal of Computer
Applications Volume 10, No-3, Nov, 2010
[4] Van Cranenburgh, Andreas, Ricus Samid, Maarten van
Someran, “Tic-Tac-Toe” , 2007
[5] Allen, Jonathan, M. Sharon Hunnicutt and Denis Kaltt
1987, From text to speech: the MITalk system,
Cambridge University Press New York, NY, USA 1987
[6] MicrosoftTechSupportForum,https://social.microsoft.co
m/forums
[7] MicrosoftDevelopmentNetwork,http://msdn.microsoft.co
m/
[8] Herbert Schildt, “C# 3.0 – The complete reference to
C#”, McGraw Hill Professional
IJCATM : www.ijcaonline.org