International Journal of Engineering and Techniques - Volume 4 Issue 1, Jan – Feb 2018 ISSN: 2395-1303 http://www.ijetjournal.org Page 186 Data Mining Systems to Determine Sales Trends and Quantity Forecast Using Association Rule and CRISP-DM Method Fifit Alfiah 1 , Bagus Wahyu Pandhito 2 , Ani Trio Sunarni 3 , Deni Muharam 4 , Pradiko Roliwinsyah Matusin 5 1,5(Teknik Informatika, STMIK Raharja, and Tangerang) 2,3,4 (Ilmu Komputer, Universitas Budi Luhur, and Jakarta) I. INTRODUCTION Data is something that has not been meaning, data that has been processed into an information, and that information that can be used. Knowledge in digging data is necessary in generating information so that it becomes something good that can be used for various purposes. Data mining describes the discovery of knowledge in the database. The use of mathematical techniques, data analysis, artificial intelligence, and machine learning to produce related information from large databases. Data mining deals with the discovery of something hidden and certain data patterns that are not previously known. In the management of data mining, the data set becomes a very important, coupled with the method of data mining will produce knowledge. This knowledge is in the form of patterns, formulas, rules, models emerging from the data. So it can be said that good knowledge will produce good information as wellunderline. At this time the research to determine the pattern of sales trend and quantity forecast done at PT. Pinus Merah Abadi which is a distributor of vegetable products spread across the dots all over Indonesia, turnover is an important part of a distributor company because a good turnover will be good for human resources as it will open a new point, so it will certainly require a new workforce and promotion for the able and responsible. Customer is the most important part of the sale, therefore need to be analyzed up to the level of customer, the data taken is a record of purchase transactions, The data must be considered include the product mix purchased in each transaction and RESEARCH ARTICLE OPEN ACCESS Abstract: Customer is the most important part of the business, the data taken is a record of the purchase transactions of products purchased in each transaction. With the existence of data mining expected hidden and unknown pattern can be utilized in customer purchasing pattern. Then apriori algorithm as the basis of which there are methods of association rules and CRISP-DM in this system can determine thats the most products of interest by customers by applying data mining system on each transaction data. Result of data mining processing to determine sales trend towards a sales product where with this sales trend management team can analyze by disclosing which product sales follow steep growth path and which stall or decrease. An example of data mining to determine the sales trend pattern based on a combination of 2 products. Where it has been determined is Threshold Support = 0.1 and Threshold Support x Confidence = 0.05 and for the quantity forecast of 23 products into the sample and who managed to enter into quantity forecast only 13 products. Where the successful product is determined quantity forecast only that has a support value above the threshold support value that has been determined by the authors in this paper. Result of the quantity forecast of the input specified such as Threshold Support: 0.2, Threshold SupportxConfidence: 0.1 and Percent Forcast: 0.15. Keywords — Data mining, sales trend, quantity forecast, support, confidence, apriori algorithm, CRIPS- DM.