In garment industry, the production process is composed of several
steps, starting with product development, production until it
reaches the customer. All these processes generate a large volume
of data, about products, customers, sales, among others. The lack
of knowledge about this data generated in the companies, can
generate problems such as product returns and consequently the
loss of customers. The application of data mining (DM) tasks can
extract hidden patterns from this data. In this context, DM
techniques following Knowledge Discovery in Database (KDD)
steps were used to identify patterns that may influence product
sales performance in an apparel manufacturing industry. The
dataset considered were composed of 31 attributes, about
customers, products, sales, from the years 2018 through 2020. The
main results of this analysis identified the attributes that most
influence product sales, which are: “product line”, “quota item”,
“customer code”, and “tax coupon number”.