Clustering Application for Customer Segmentation in the JUSTA Database

• Autor(a) principal: Rocha, Allana Lais dos Santos
• Data de Publicação: 2022
• Outros Autores: de Macêdo, Ester Deschamps, de Oliveira, Letícia Castro Portela, Silva, Vinícius Ferreira
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Revista de Engenharia e Pesquisa Aplicada
• Texto Completo: http://revistas.poli.br/index.php/repa/article/view/2458

Financial technology companies, also known as fintechs, are innovative technology companies with the potential to transform the financial sector. For them to apply a personalised treatment of clients, extensive data analysis is required. Therefore, employing data mining techniques can offer advantages in classifying and visualising costumers. Justa, the company explored in this work, is a fintech that provides products and services through digital bank accounts, and it sought to improve its understanding of its client base. Using anonymised datasets provided by Justa, each client was represented by features they considered relevant. To arrive at the final dataset, the integration, reduction, cleansing, and transformation of the original data was performed. The algorithms tested for grouping customers were K-Means, fuzzy C-Means and K-Medoids, where K-medoids presented better results in the delineation of the profiles. The results indicated that there are different profiles of clients, but that these are barely perceptible and are concentrated in a few behavioral characteristics.