透過您的圖書館登入
IP:3.131.110.169
  • 期刊

An Efficient Boolean Method for Discovering the Maximum Frequent Set

使用一種高效率方法找尋最大頻繁項目集

摘要


在交易資料庫裡找出項目集間的關係是資料探勘領域裡非常重要的議題,其中關聯規則的探勘是主要的問題。而在資訊管理方面,關聯規則亦能應用於相當多的層面,例如知識管理、客戶關係管理和決策支援等等。在本篇論文中,提出一個以布林為基礎尋找最大頻繁項目集演算法,我們稱之為Boolean Method for Discovering Maximum Frequent Set (BMFS)。BMFS結合了Boolean和Pincer-Search兩個演算法的優點,利用非頻繁項目集來尋找最大頻繁項目集並且在搜尋的過程中可以有效的減少資料庫掃描次數。在效能評估方面我們以IBM synthetic benchmark所產生的資料庫為測試的對象。實驗結果顯示,在所有測試資料庫中BMFS的效能皆優於Pincer-Search演算法。

並列摘要


A very important issue in the data mining field is to find out a useful related rule among the itemsets from a large trade database. It is useful for information management or decision making, such as knowledge management, customer relationship management and decision making etc. In this paper, we propose an efficient Boolean base method for Discovering Maximum Frequent Set (BMFS) in a large trade database, called Boolean Method. The BMFS method combines the advantages of a Boolean and Pincer-Search algorithms to reduce the number of database scans, and uses infrequent itemsets to find the maximum frequent itemsets. We evaluate the performance of the algorithm using a well-known synthetic benchmark database. The experimental results showed that the BMFS method has better performance than the Pincer-Search algorithm for most of the test cases.

延伸閱讀