(Lukasiewicz Research Network - Institute of Biopolymers and Chemical Fibres, Poland, 2017) Istrat, Višnja; Lalić, Nenad
Sales process disfunctions in the textile industry are problems that cause loss of customers, incomplete market supply, etc. The objective of the research is to analyse transactions from the textile industry database in order to find patterns in buyers’ behavior and improve the model of decision-making. Association rules, one of the most noticeable data mining techniques, is used as methodology to learn rules and market patterns that occur in sales in the textile industry, which will enhance the decision-making process, by making it more effective and efficient. The Apriori algorithm was applied and open source software Orange was used. It has been shown using a real-life dataset containing 2000 transactions from the textile industry of the South East Europe region that the approach proposed is useful in discovering effective knowledge in data associated with sales. The study reports new interesting rules and the dependence of the following parameters: support, confidence, lift and leverage on making more customized offers in the textile industry.
