| Abstract: | Information for businesses or organizations, is a very important asset. Problems accumulated data, companies are facing, not a lack of information, but too much information. For the large and growing amount of data if not effectively treated, it will result in the case of so-called "data dumping" is. After statistical or use data mining techniques to be addressed, it can be converted into information or knowledge can be utilized. Available to decision makers to make the right moves. Enterprises are facing a huge market, competitive pressures, customer's consumption habits continue to change, the enterprise-based revenue sources and new customers and old customers repeat purchase consumption. The advancement of technology, the increasingly common use of the database, so that large amounts of data to be stored and managed to save. By data mining technology, and explore the establishment of consumer purchase signature rule merchandise mix, master Consumer consumption situation, the establishment of market segments and dividing the target customer base, and to identify potential market relevant consumers to predict their consumption behavior. The results pointed out that the turnover in the other Saturday with high turnover association, turnover in the other Sunday with high turnover relevance, do not turnover in the third quarter with high turnover relevance turnover not the shop on Sundays and public high turnover association, turnover in love do not shop on Sunday with high turnover association, turnover in other Zonta store on Sunday with high turnover relevance, not in turnover Feng Chia shop Saturday with high turnover relevance of the week in other public stores and Sunday have relevance, do week in stores and on Sunday there are fraternity association, the other week and Sunday Zonta store has relevance, do not store in the week and Saturday Fengjia have relevance. |
