Mining Association Rules in Distributed Systems with the Rough Set Theory

• Main Authors: SHIH, JHAO-LONG, 施兆隆
• Other Authors: Chao, Ching-Ming
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/ub4y26

碩士 === 東吳大學 === 資訊管理學系 === 106 === Mining Association rule is a technology that can find out the correlation between products from the transaction database, and useing the items which customer purchased, the enterprise can summarize the customer's habit, and then implement the appropriate marketing strategy. However, because of the increasing amount of data, in order to improve the speed of data mining, the technology of distributed database exploration is gradually valued by enterprises. This study proposes a method of distributed mining association rules call ApRoughSet algorithm, through the distributed databases to handle the big volume of data.Mining association rules at the local database, and extract the data which contain local association rules at the local database, using rough set theory to obtain global rules. The characteristics of rough sets in finding rules make it possible to derive rules which contain parchased and unpurchased products to increase the efficiency of finding association rules and unrevealed information.Experiment results show that the global association rules obtained by rough set theory are more streamliner ,and can also find the relationship between parchased and unpurchased products so that enterprises can have more information to make decisions.