期刊名称:International Journal of Computer Science and Network Security
印刷版ISSN:1738-7906
出版年度:2018
卷号:18
期号:7
页码:6-16
出版社:International Journal of Computer Science and Network Security
摘要:The goal is to survey big data techniques applications to financial crime prevention and detection in more than two decades. Also, to determine the industrial sector of financial crime that has gathered the most interest and which fields still lack research. The study describes the most common methods for detecting financial crime, indicating some of the current research problems, trends, and issues in big data application. Meanwhile, the survey focuses on financial crime detection techniques applied in several sectors such as banks, computer networks, insurance, securities, exchange commodities, stock markets and money laundering. The methods considered in the survey include big data analytics foundational technologies and big data analytics emerging research. In most of the data-collection strategies and data analysis, there was a shift from traditional data collection method to computer-based methods. In the aspect of data analysis, there was an increase in the use of descriptive, inferential statistical analysis. We made an evaluation of the detecting techniques based on data analysis factors such as processing speed, latency, volume, performance, fault tolerance, scalability, and accuracy. Then we propose that anomaly, data-mining, clustering, hybrid-technologies, neural networks, rough clustering, k-Means clustering, neuro-fuzzy, genetic algorithms and fuzzy support vector machine models are performing better than other methods currently in practice. However, more research is required in big data infrastructures like hardware equipment, software licensing, and maintenance which are still very expensive. Likewise, further research is needed in the human analysis of big data approaches to financial crime detection because of the challenges of sorting out information.
关键词:;;;; ;;;;;; Financial-crime; big data techniques; detecting; prevention; methods.