Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi
Yazarlar: Kevser ÖZDEM, M. Ali AKCAYOL
Konular:Bilgisayar Bilimleri, Bilgi Sistemleri
DOI:10.28948/ngumuh.766503
Anahtar Kelimeler:Big data analysis,Data mining,Association analysis,Bank marketing campaign
Özet: Campaign based sales continue to increase at a very rapid rate in recent years. Today, sales in many sectors are based on the campaign. Therefore, campaign management has become a very important issue. There is no detailed study on campaign planning and management in the literature. In this article, a model is developed for campaign management in the banking sector and prediction of prospective behaviors of customers towards the campaign. Using the association analysis structure developed specifically for the banking sector, frequent itemsets and association rules were created from the campaign data of a Portuguese bank. The prospective behavior of the customers participating in the campaign was estimated with the rules obtained. In addition, attributes that affect the behavior of customers have been identified. Experimental results have shown that marital status and credit status affect customer behavior the most. Using the developed model, a prediction is made on whether the customers will participate in the campaign or not. 87% success was achieved in the prediction of customers' participation in the campaign.