差異性顯示對關聯法則使用之評估

The Evaluation of Using Differential Display for Association Rules

沈明賢(Ming Hsien Shen)

指導教授 ： 林志麟

元智大學/資訊學院/資訊管理學系/碩士(2001年)

摘要

關聯法則(Association Rules)在資料探勘(Data Mining)的領域中，是非常重要的一個探勘技術。由於此技術具有統計理論基礎的支持且實用性極高，因此，與關聯法則相關的研究也就非常的多。然而大部分的研究大多集中於關聯法則演算法速度的改良，對於如何管理、如何充分應用這些關聯法則的研究反而少之又少。如何針對這些關聯法則做最有效的應用，以反應出關聯法則所潛藏的資訊，乃本研究的動機所在。
對於企業經營者而言，與其提供大量最新的關聯法則來描述目前市場的最新狀況，倒不如告訴企業經營者目前市場和過去相比，出現了那些變化。一般而言，描述市場的變化情形所需的資料量遠比描述市場整個情況的資料量少。故本研究的第一個目的便是有效的將關聯法則的變化趨勢提供給企業經營者。此外，本研究的第二個目的在於如何表示關聯法則之間的相關性。針對上述目的，本研究將開發一套可供下鑽的系統，使用者透過此系統，可有效的分析關聯法則間的相關性，以助於使用者對關聯法則的瞭解。

關鍵字

關聯法則 ； 資料探勘

並列摘要

Mining association rules is one of the most important problems in the area of data mining. Due to its practicality, much work has been done for the past few years. Most literatures focus on how to discover the association rules efficiently, however, very little has been done regarding how to help business users utilizing the association rules more effectively, and this is the motivation of this research.
Providing business users with the changing behavior of association rules is usually more important than providing them with all of the association rules. The amount of information used to describe the changing behavior of association rules is usually much less than the amount of information used to describe all of the association rules. The first objective of this research is to provide users with a way to uncover the changing behavior of association rules. Besides, there are some implications among association rules that can be used to discover how each item in an association rule affects the rule. The second objective of this research is to design a technique to discover these implications and display them in a user-friendly fashion. Overall speaking, this research will provide business users with a system to better utilizing the association rules.

並列關鍵字

Data Mining ； Association Rule

參考文獻

[1]Alex Berson and Stephen J. Smith, Data Warehousing, Data Mining, & OLAP, McGraw-Hill, 1997.

[3]A. Savasere, E. Omiecinski and S. Navathe. "Mining for Strong Negative Associations in a Large Database of Customer Transactions," Proceedings of the 14th International Conference on Data Engineering, pages 494-502, 1998.

[5]Bayardo and R. J. Jr. "Efficiently Mining Long Patterns from Databases," Proceeding of the 1998 ACM-SIGMOD Int’l Conf. On Management of Data, 1998, Page(s):85-93.

[6]Bing Liu, Wynne Hsu, Shu Chen and Yiming Ma. "Analyzing the subjective interestingness of association rules," IEEE Intelligent Systems [see also IEEE Expert] , Volume: 15 Issue: 5, Sept.-Oct. 2000 ,Page(s): 47 —55.

[7]B. Lent, A. Swami and J. Widom. "Clustering Association Rules," Proceedings of the 13th International Conference on Data Engineering. pages 220-231, April 1997.