Akademik Veri Yönetim Sistemi
3rd International Congress on Scientific Advances (ICONSAD), Balıkesir, Türkiye, 20 - 23 Aralık 2023, ss.249-255
According to the accounting transactions, companies can keep their economic information and take their next steps based on this information. Today, through the contributions of digitalization and developing technology, these processes can be done digitally in the electronic environment. These digital systems such as e-invoice and e-ledger are offered to users as a service by different companies. However, the customers do not use all of the digital accounting transaction products provided by the accounting companies. The main idea of this paper emerges at that point. With the customer information we first cluster these customers by using k-means algorithms, and then using the similarity of the clusters, we identify applications that customers may need. According to the results, with the available data sets, there are three clusters. In the first cluster e-invoice, e-delivery, and e-archive should be used by the customers, which they do not use currently. In the second cluster, on the other hand, e-invoice and e-ledger should be suggested for the non-user customers. Unfortunately, for the third cluster, there is not enough data to get an inference. Thus, with the machine learning algorithms, the requirements of the customers can be defined and the missing products can be offered in advance.