使用模糊分割自概念層級架構中找出關聯規則

Mining Association Rules at a Concept Hierarchy Using Fuzzy Partition

胡宜中(Yi-Chung Hu)

《資訊管理學報》 13卷3期 (2006/07) Pp. 63-80

https://doi.org/10.6382/JIM.200607.0063

摘要

關聯規則可輔助管理者進行行銷策略與商場架位之規畫，然而大部份的探勘方法均僅針對實體項目或商品進行分析。有別於僅在單一層級中找出關聯規則，以及考量由使用者感認與主觀判斷所產生的認知不確定性，本研究的目的在於使用模糊分割，以提出一個可自概念層級架構中找出模糊關聯規則的方法。所提出之方法主要是由兩階段所組成：在第一階段中依據層級架構將資料項目做適當的抽象化，並找出高頻的模糊格；而在第二階段中係由高頻模糊格進一步產生多層級模糊關聯規則。其特色在於使用表格結構以存放高頻模糊格，且高頻模糊格與多層級模糊關聯規則均為使用布林運算所產生。本文亦探討所提出方法在不同的資料庫大小與相關參數設定下，對執行時問與關聯規則之產生所造成的影響。實驗結果顯示所提出方法可有效提升執行效率。

關鍵字

模糊集合 ； 關聯規則 ； 資料探勘 ； 概念層級架構 ； 決策支援

並列摘要

Association rules can help managers to plan marketing or design store layouts. However, many methods are developed by analyzing the relationships among data items at a single level. Since cognitive uncertainty arising from human perception, cognition and subject judgment should be taken into account, in this paper, a new method is proposed to mine multiple-level fuzzy association rules among data items at a concept hierarchy, using fuzzy partition by a simple fuzzy grid. The proposed method primarily consists of two phases: one to find frequent fuzzy grids at each level, and the other to generate multiple-level fuzzy association rules from those frequent patterns. The main feature of the proposed method is to employ the table structure to store frequent fuzzy grids. In particular, both frequent fuzzy grids and multiple-level fuzzy association rules can be efficiently generated by applying the Boolean operations on the table structure. To understand the impact of the proposed method on the execution time and the number of generated association rules, the experiments are performed by using different sizes of databases and thresholds. The experimental results demonstrate the proposed method is efficient.

並列關鍵字

Fuzzy sets ； Association rules ； Data mining ； Concept hierarchy ； Decision support

參考文獻

Agrawal, R.,Gehrke, J.,Gunopulos, D.,Raghavan, P.(1998).Automatic Subspace Clustering of High Dimensional Data for Data Mining Applications.Proceedings of the ACM SIGMOD International Conference on Management of Data.(Proceedings of the ACM SIGMOD International Conference on Management of Data).

Agrawal, R.,Mannila, H.,Srikant, R.,Toivonen, H.,Verkamo, A. I.,U. M. Fayyad (Eds.),G. Piatetsky-Shapiro, P. (Eds.),Smyth, R. Uthurusamy, (Eds.)(1996).Advances in Knowledge Discovery and Data Mining.Menlo Park:AAAI Press.

Agrawal, R.,Shafer, J. C.(1996).Parallel Mining of Association Rules.IEEE Transactions on Knowledge and Data Engineering.8(6),962-969.

Berry, M.,Linoff, G.(1997).Data Mining Techniques: for Marketing, Sales, and Customer Support.NY:John Wiley & Sons.

Bezdek, J. C.(1981).Pattern Recognition with Fuzzy Objective Function Algorithms.NY:Plenum.