Parallel Methods for Mining Frequent Sequential patterns

Quoc Bao Huynh

Doctoral thesis

Abstract:

The explosive growth of data and the rapid progress of technology have led to a huge amount of data that is collected every day. In that data volume contains much valuable information. Data mining is the emerging field of applying statistical and artificial intelligence techniques to the problem of finding novel, useful and non-trivial patterns from large databases. It is the task of discovering interesting patterns from large amounts of data. This is achieved by determining both implicit and explicit unidentified patterns in data that can direct the process of decision making. There are many data mining tasks, such as classification, clustering, association rule mining and sequential pattern mining. In that, sequential pattern mining is an important problem in data mining. It provides an effective way to analyze the sequence data. The goal of sequential pattern mining is to discover interesting, unexpected and useful patterns from sequence databases. This task is used in many wide applications such as financial data analysis of banks, retail industry, customer shopping history, goods transportation, consumption and services, telecommunication industry, biological data analysis, scientific applications, network intrusion detection, scientific research, etc. Different types of sequential pattern mining can be performed, they are sequential patterns, maximal sequential patterns, closed sequences, constraint based and time interval based sequential patterns. Sequential pattern mining refers to the identification of frequent subsequences in sequence databases as patterns. In the last two decades, researchers have proposed many techniques and algorithms for extracting the frequent sequential patterns, in which the downward closure property plays a fundamental role. Sequential pattern is a sequence of itemsets that frequently occur in a specific order, where all items in the same itemsets are supposed to have the same transaction time value. One of the challenges for sequential pattern mining is the computational costs beside that is the potentially huge number of extracted patterns. In this thesis, we present an overview of the work done for sequential pattern mining and develop parallel methods for mining frequent sequential patterns in sequence databases that can tackle emerging data processing workloads while coping with larger and larger scales.

Abstract:

The explosive growth of data and the rapid progress of technology have led to a huge amount of data that is collected every day. In that data volume contains much valuable information. Data mining is the emerging field of applying statistical and artificial intelligence techniques to the problem of finding novel, useful and non-trivial patterns from large databases. It is the task of discovering interesting patterns from large amounts of data. This is achieved by determining both implicit and explicit unidentified patterns in data that can direct the process of decision making. There are many data mining tasks, such as classification, clustering, association rule mining and sequential pattern mining. In that, sequential pattern mining is an important problem in data mining. It provides an effective way to analyze the sequence data. The goal of sequential pattern mining is to discover interesting, unexpected and useful patterns from sequence databases. This task is used in many wide applications such as financial data analysis of banks, retail industry, customer shopping history, goods transportation, consumption and services, telecommunication industry, biological data analysis, scientific applications, network intrusion detection, scientific research, etc. Different types of sequential pattern mining can be performed, they are sequential patterns, maximal sequential patterns, closed sequences, constraint based and time interval based sequential patterns. Sequential pattern mining refers to the identification of frequent subsequences in sequence databases as patterns. In the last two decades, researchers have proposed many techniques and algorithms for extracting the frequent sequential patterns, in which the downward closure property plays a fundamental role. Sequential pattern is a sequence of itemsets that frequently occur in a specific order, where all items in the same itemsets are supposed to have the same transaction time value. One of the challenges for sequential pattern mining is the computational costs beside that is the potentially huge number of extracted patterns. In this thesis, we present an overview of the work done for sequential pattern mining and develop parallel methods for mining frequent sequential patterns in sequence databases that can tackle emerging data processing workloads while coping with larger and larger scales.

Keywords

Data Mining, Frequent Pattern, Sequential Pattern, Parallel Mining, Dynamic Load Balancing, Multi-core processors, Sequence Database, Large Sequence Database

Keywords

Data Mining, Frequent Pattern, Sequential Pattern, Parallel Mining, Dynamic Load Balancing, Multi-core processors, Sequence Database, Large Sequence Database

Language used: English

Date on which the thesis was submitted / produced: 1. 11. 2017

Thesis defence

• Date of defence: 14. 11. 2017
• Supervisor: Václav Snášel
• Reader: Jan Platoš, Roman Šenkeřík, Jaroslav Zendulka

VŠB - Technical University of Ostrava

Fakulta elektrotechniky a informatiky

Doctoral programme / field:
Informatika, komunikační technologie a aplikovaná matematika / Informatika