Title: Sentiment analysis and support vector machine one versus one for collectibility classification of bank's house ownership loan

Authors: Carmelia Nabila Permatasari; Adji Achmad Rinaldo Fernandes

Addresses: Faculty of Mathematics and Natural Science, University of Brawijaya, St. Veteran, Malang City, 65145, East Java, Indonesia ' Faculty of Mathematics and Natural Science, University of Brawijaya, St. Veteran, Malang City, 65145, East Java, Indonesia

Abstract: Not many applications of sentiment analysis have been developed for the Indonesian language. One classification method that can be applied to extracting information for large databases is the support vector machine (SVM). This study aims to compile research variables based on the results of sentiment analysis and examine SVM performance to solve multi-class cases using the one versus one method with linear kernel functions, quadratic polynomial kernel functions, and radial basis function (RBF) kernel functions in classifying the collectibility classification of bank's house ownership loan based on Pernyataan Standar Akuntansi Keuangan (PSAK) 71, among others: 1) performing loan; 2) under-performing loan; 3) non-performing loan. The results showed that SVM one versus one with the kernel RBF is the most appropriate method in classifying collectibility levels bank mortgage debtors based on the Big Five personality because the accuracy, sensitivity, and specificity values obtained on the perfect testing data are 100%.

Keywords: performance; debtor collectibility; bank: house ownership loan: one versus one: support vector machine; SVM.

DOI: 10.1504/IJADS.2024.138193

International Journal of Applied Decision Sciences, 2024 Vol.17 No.3, pp.293 - 312

Received: 15 Jun 2022
Accepted: 19 Oct 2022
Published online: 30 Apr 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article