A proposal to Implement and test the ID-3 Algorithm on the Facebook Application Data set.
Project Documentation
By:
Anuj Patel
Akash Parekh
CPSC 531-03 (21103)

Professor: Dr. Chun-l Philip Chen

Department of Computer Science

California State University, Fullerton

Spring, 2013

Table of Contents Abstract 3 1.Introduction 5
1.1 Background 5
1.2 Problem Statement 5
1.3 Project Goals and benefits of study 5
1.4 Hypotheses 5
1.5Relevance and Significance………………………………...………………………….........6
1.6Assumption,Limitation and Delimitation………………………...…………………………6
1.7 Barriers and Issue…………...………………………………………………………………6
2. Literature Review…………………………...…………………………………………………..7 2.1 Brief Review of literature…………………….…………………………………………….7
3. Methodology..……...………………………. .…………………………………………………8 3.1 Overview……………………….……..…………………………………………….…...….8
3.2 Sample Dataset…………………………………………………………………………….10
3.3Selected Dataset 11
3.4Output Decision Tree 11
4. Results & discussion………………. .………………………………………....……………...18
4.1 Resulting Decision Tree……………………………………………………………………...18
4.2 Data Analysis and Findings……………………………………………………………….19
4.3 Summery of Result………………………………………………………………………..19
5. Conclusions, implications & recommendations ……………………………………………...19
References……………………………………………………………………………….……21

Table of Figures Figure 1: Small part of User's data from Datasets 10 Figure 2. Authentication page for the system (GUI).……………………………………………12 Figure 3. Dashboard of the System …………………………………………….…………………..13 Figure 4. Create a USER………………………………………………………………………………….14 Figure 5. Prediction using different attributes ……………………………………………………………15 Figure 6. Prediction using main attributes ……………………………………………………….16 Figure 7. System Generated Tree…………………………………………………………………………17 Figure 8: Decision Tree………………………………………………………………………………………………………………………..18

Abstract

Classification of data objects based on a predefined knowledge of the objects is Data Mining. The subject is very vast and it generates curiosity when it comes to the classification of data which revels unexpected patterns and relationships amongst the data. The patterns and relationships can be used as a base for future decision making for new product development and genera selections. In this project, we will discuss a classification algorithm in details and implement the project based on the algorithm in order to gain the desired outcomes which can help in future development.

1. INTRODUCTION

1.1 Background
Ever since Facebook released their Application Programming Interface, developers all around the world developed thousands of application for almost one billion Facebook users. They have applications for all most each and every kind of users. The users are kept increasing and to reach the requirements of the users application developers tries to build the application that is in demand by the users. We can do a survey on the previous usage of the applications and then can figure out what kind of application will be good to implement in the present situation. We have selected ID3 algorithm to demonstrate data mining. ID3 algorithm was invented by J. Ross Quinlan at the University of Sydney. ID3 is based on the Concept Learning System (CLS) algorithm. ID3 is very simple decision learning algorithm. It will construct decision tree by employing a top-down, greedy search through the given set of training data to each attribute at every node in the tree. ID3 uses nominal attributes for classification with no missing values. The application will implement the ID3 algorithm on the Facebook application generated dataset using open source technologies JAVA on front end and MySQL on back end.
1.2 Problem Statement
In the world of online social networking, deciding an application type which will receive high acceptance by the user is tough and this problem will appear each and every time when it comes to decision making. The sole reason