Innovative techniques to improve frequent pattern mining in data streams

Abstract

iii newlineABSTRACT newline Informational Society is the most proper and fitting term to describe newlinethe present day world. An astounding and ever expanding amount of information is newlinepouring in every day which is used for organizing the economy and the society itself. newlineThis has paved the way for the collection of such information with the aid of newlinesophisticated technologies and tools such as computers, satellites, remote sensors, newlineand many more. At the outset, much more importance was given to the collection newlineand storage of information as it is an established fact that information leads to power newlineand power in turn leads to success. Unfortunately, these massive collections of data newlinefrom heterogeneous sources, stored on disparate structures very rapidly became newlineoverwhelming. Nowadays even simple transactions such as using a credit card, a newlinephone or browsing the web have become ubiquitous and mammoth. In such newlinescenarios, analysis of the collected information becomes more essential than newlinecollection and storage. Thus automatically summarizing the data, extracting the core newlineinformation and discovering interesting patterns from the vast and ever growing raw newlinedata streams has become the need of the hour. newlineThe extraction of hidden predictive information from such large streams newlineof data is commonly referred to as data stream mining. It is a new kind of technology newlinethat blends traditional data analysis methods with sophisticated algorithms for newlineprocessing large volumes of dynamic data. Traditional data analysis tools and newlinetechniques cannot be used in such cases because of the massive size of data streams. newlineAlso the transactions arrive on the data streams continuously and so they cannot be newlinestored permanently even though memory devices are cheap and readily available. newlineHence there is an urgent need to develop new schemes for storing and mining these newlinedata streams. newlineThe objective of this thesis is to apply data mining techniques to discover newlinefrequent patterns from data streams in order to understand and serve better the needs newlineof the society. In this thesis, more eff