Extended predictive model markup language

Pao Yue-kong Library Electronic Theses Database

Extended predictive model markup language


Author: Lau, Tze-wah
Title: Extended predictive model markup language
Degree: M.Sc.
Year: 2001
Subject: Data mining
XML (Document markup language)
Hong Kong Polytechnic University -- Dissertations
Department: Multi-disciplinary Studies
Dept. of Computing
Pages: 113 leaves : ill. ; 30 cm
Language: English
InnoPac Record: http://library.polyu.edu.hk/record=b1599608
URI: http://theses.lib.polyu.edu.hk/handle/200/593
Abstract: Nowadays, Data mining application system is usually based on 2-Tiered architecture and 3-Tiered transaction based architecture. 2-Tiered architecture has been proved very efficient in environment with small number of servers and client. And 3-Tiered architecture is the logical extension to its predecessor when number of machines involved in the IT environment growing above 50 and its very efficient in usually OLTP environment. However, efficiency of these architectures is not profound in the OLAP environment. Unlike OLTP environment, OLAP is characterised by small amount but very long transactions. And these transactions involve large volume of data manipulation in multiple servers. These complex data manipulation procedures and logic have to code into the application logic in 2-Tiered and transactional 3-Tiered architecture. The portability and scalability of data mining applications based on these architectures is seriously affected. In addition, data mining applications are tightly bind to the architecture. Applications based on middleware from different vendors usually cannot communicate directly without using a process gateway. The interoperability of application is also affected. Message Passing 3-Tiered architecture is suggested in this dissertation to address shortcomings in 2-Tiered and transactional 3-Tiered architecture. Processes talks among themselves based on messages, which contains most of the intelligence required carrying out data mining processes. ePMML hence is developed as the standard message format used for inter-processes communications. It is a XML dialect detail out the overall data mining application. ePMML help to include information such as server locations, fields extracted from database, field mapping, data transformation and cleansing. Storing this information in messages avoids hard coding the data flow logic in middleware processes. By doing so, the portability, scalability and interoperability of the data mining application are enhanced. In the dissertation, a reference implementation is suggested to test the practicality and usability of Message Passing 3-tiered architecture in Data mining application. Reference implementation is developed on Linux using Java and using Web as front end.

Files in this item

Files Size Format
b15996086.pdf 3.445Mb PDF
Copyright Undertaking
As a bona fide Library user, I declare that:
  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.


Quick Search


More Information