Toward supporting real-time mining for data residing on enterprise systems

As data mining techniques are explored extensively, incorporating discovered knowledge into business strategies gives superior competitive advantage to corporations. Most techniques in mining association rules nowadays are designed to solve problems based on transaction files transformed to horizontal or vertical format. Namely, the transaction-normalized tables should be transformed before such methods could be applied, and some previous works have pointed out that such tasks of performing data transformation usually consume a lot of resources. As a result, traditionally, data mining technique has seldom being applied in real-time. However, in many cases, the decisions have to be made in a short time, such as the decisions of promoting fresh agriculture goods in retailing stores should be made daily and in the limit of one or two hours. This study therefore proposes a new method which incorporates mining algorithms with enterprise transaction databases directly to perform real-time mining. In addition, the proposed method has following advantages to support real-time mining performed in enterprise systems:
• raw data of enterprise systems are used directly,
• when the threshold is tuned, only newly qualified data are read and the data structure built for original data is kept intact,
• product assortments centered on particular product can be effectively performed,
• the performance of the mining algorithm is better than that of popular mining algorithms.