Mining Data in Identification of Consumer Patterns of Agricultural Machine Sales Using Fp-Growth Algorithm

The sales transaction data for agricultural machinery at the Mandiri Jaya Teknik Solok store is a large data set making it difficult to identify consumer purchasing patterns. Large data sets can be processed into useful information. Sales transaction data available at the Mandiri Jaya Teknik Solok store can be processed into useful information to increase sales. This study aims to identify consumer purchasing patterns in order to know which items are often sold and to find out which items need to be stocked more and to increase sales. The data that is processed in this study uses the sales transaction data obtained from the sales invoice of Toko Mandiri Jaya Teknik Solok. Data is in the form of sales data for 13 weeks of 20 items with a minimum support value of 15% and a confidence value of 60%. The method uses one of the data mining techniques associated with the FP-Growth algorithm, where the Fp-Growth algorithm uses the concept of tree development in searching for the types of items that are often purchased (frequency item sets). The tools used are Rapidminer 9.8 so that the purchase patterns of goods are obtained which are used as information to predict the level of frequently sold items. The result of the sales data processing process is the association rule. Association Rule is obtained in the form of a relationship between goods sold together with other goods in a transaction. From this pattern, it can be recommended to the shop owner as information for preparing stock of goods to increase sales results. This research is very suitable to be applied to determine the patterns of consumer spending such as the relationship of each item purchased by consumers, so this research is appropriate for use by stores