Data mining
Tina Malekolkalami; Khadijah Khodabakhshi Parijani; Maliheh Alifari
Abstract
The purpose of this research is to use data mining to detect accounting fraud in the database of stock exchange member companies. The combination of discrete and continuous data has increased the necessity of using data mining and machine learning methods in the field of fraud detection. This research ...
Read More
The purpose of this research is to use data mining to detect accounting fraud in the database of stock exchange member companies. The combination of discrete and continuous data has increased the necessity of using data mining and machine learning methods in the field of fraud detection. This research is applied in terms of purpose and descriptive in terms of method. The document review method was used to collect information in the field of literature and research background. The prepared questionnaire includes 7 main indicators consisting of 48 questions for each of the variables. This questionnaire was made available by the researcher to 400 accountants of companies admitted to the Tehran Stock Exchange by sampling method. In order to fit the model, the structural equation method was used in SMARTPLS software. In the data mining section, all IB1, IBK, LWL, KSTAR, and KNN algorithms were used to simulate the proposed model in Rapidminer software. Effectiveness of internal control, compensation system, asymmetry of information, compliance with accounting rules, management ethics, and ethical principles are effective and meaningful on accounting fraud. In evaluating parameters and according to the graphs, the K-STAR algorithm has better performance than other algorithms. The proposed data mining model for financial fraud detection showed that since the amount of data creation in financial companies is increasing day by day with the development of technology, it is possible to provide early detection of fraud by reviewing and analyzing the data.
