在一個資料庫裡，異常是指特性相異於其他資料的資料點。偵測出這些異常在許多應用中是很重要的。大部分傳統異常探勘的資料屬性都著重在數值屬性，但是現實生活中有很多類別屬性的資料，所以傳統的異常探勘方法就不適合使用。FindFPOF (frequent pattern outlier factor)演算法是利用頻繁項目集找出異常交易，但是FindFPOF會產生出許多頻繁樣式集。我們的目的是找出異常而不是在找出頻繁樣式，因此，本論文提出減少頻繁項目集(reduce frequent pattern set ; Rfp)來找出異常交易的方法，稱之為RfpFPOF。實驗結果顯示本論文所提出的方法在類別資料庫中發現異常交易的效能優於FindFPOF。An outlier in a database is an observation or a point that is considerably dissimilar to or inconsistent with the remainder of the data. Outlier detection is very import for many applications. Most existing methods are designed for numeric data. They will have problems with real-life applications that contain categorical data. In order to solve this problem, some scholars propose the FindFPOF (Frequent Pattern Outlier Factor) algorithm to detect outliers from discovered frequent patterns. However, FindFPOF generates too many frequent patterns. The purpose of this study is to identify outlier not to find frequent patterns. Therefore, in this study, we present an efficient method to detect outliers by reducing the number of frequent patterns, called the reduce frequent pattern FPOF (RfpFPOF) algorithm. The experimental results have shown that RfpFPOF outperformed the FindFPOF method on identifying outliers execution time.
