New implementation of unsupervised ID3 algorithm (NIU-ID3) using Visual Basic.net

  • Authors:

  • Faraj A. El-Mouadib;Zakaria S. Zubi;Ahmed A. Alhouni

  • Affiliations:

  • Computer Science Department, Faculty of Information Technology, Garyounis University, Benghazi, Libya;Computer Science Department, Faculty of Science, Altahadi University, Sirte , Libya;Computer Science Department, Faculty of Science, Altahadi University, Sirte , Libya

  • Venue:
  • DNCOCO'09 Proceedings of the 8th WSEAS international conference on Data networks, communications, computers
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

The data volumes have increased noticeably in the few passed years, for this reason some researchers think that the volume of data will be duplicated every year. So data mining seems to be the most promising solution for the dilemma of dealing with too much data and very little knowledge. Database technology has dramatically evolved since 1970s and data mining became the area of attraction as it promises to turn those raw data into meaningful knowledge, which businesses can use to increase their profitability. The data mining systems are classified based on specific set of criteria such as classification according to kinds of databases mined, classification according to kinds of knowledge mined, classification according to kinds of techniques utilized and classification according to applications adapted. This classification can also be helpful to potential users to distinguish data mining systems and identify those that are best match their specific needs. The purpose of this paper is to implement one of the data mining techniques (classification) to deal with labeled data sets and merging it with another data mining technique (clustering) to deal with unlabeled data sets in a computer system using VB.net 2005. Our system (NIU- ID3), can deal with two types of data files namely; text data files and access database files. It can also preprocess unlabeled data (clustering of data objects) and process label data (classification). The NIU-ID3 can discover knowledge in two different forms, namely; decision trees and decision rules (classification rules), this approach is implemented in Visual Basic.net language with SQL. The system is tested with access database, text data (labeled datasets and unlabeled datasets) and presents the results in the form of decision trees, decision rules or simplified rules.