The article investigates the current challenges with financial fraud and how artificial intelligence can be used in fraud detection under the IoT environment. Financial fraud methods through machine learning and deep learning technology were examined between the year 2016 and 2018. The researchers then proposed an accurate fraud detection mechanism based on the advantages and limitations of their research. The conclusion arrived at from the researchers was that the detection of financial fraud would be possible through processing large amounts of data. There were relevant data that was used to support the article as various artificial intelligence techniques were examined between 2016 and 2018. The general method to combat fraud was accurate as it was based on the limitations and advantages of the study. The article was general relevant as it provided an accurate means which data could be selected, sampled, and supervised and unsupervised algorithms applied. The strength of the article was that it provided specific data regarding the accuracy of the model which was applied in Korea in the year 2015. The article is relevant to the current study of how the criminal justice can improve its ability to detect mobile fraud. A systematic approach through the use of data selection, sampling, and applying various AI technology was provided as a means to detect fraud. The aim of the article was to provide an analysis of the potential use of artificial intelligence to assist in digital crime investigation. The approach that was proposed in the paper was the use of AI in Incident response. The process proposed was that the investigator collects data from the crime scene, stores them in a database, and then groups then in order to identify relationships with any past crimes. The system should be able to learn by itself and adapts the corresponding crime and aids in the solving of future crime. The data used to carry out the investigation was relevant because it made use of past historical data to carry out the analysis. The authors investigated past crimes and identified how A.I techniques could be used by an investigator. One of the strengths of the proposed model is that it would be applicable to solve various crimes other than digital crime. The article was limited as there was no empirical data that could substantiate the results. The article was relevant to the current study of how law enforcement can benefit in fighting digital crime such as fraud and identity theft. It provided a practical model that should be followed by AI technology to fight the crime. The aim of the article was to help law enforcement officers to be able to have a clear approach regarding how fraud works and to be able to detect it. The author further observed that by understanding the process of fraud, AI technologies can be implemented to detect any of the sections of the fraud. The three steps to every fraud were identified as a fraud plan, having ‘red flags, and linking of the ‘fraud plan’ to the red flags. A particular fraud plan can thus be used to predict a new fraud through the use of AI technology. The model that can be used to assist with the fraud detection would a ‘fraud simulator’ that is based on red flags and the fraud plan. The authors make use of relevant data regarding various case studies and historical events relating to fraud. They accurately arrive at the conclusion of the steps of detecting fraud by providing a practical plan. The advantage of the article is that the model selected is easy to understand and law enforcement officers can use it without necessarily making use of AI technology. The drawback of the article is that their approach did not provide guidance on the development of the AI simulator. The given article is relevant to the thesis because it provides three steps that AI technology can be integrated with the knowledge of fraud to create a fraud detection system. This will make the work of law enforcement officers easy as they will quickly detect fraud and respond in time. The article examined some of the artificial intelligence hybrid models that have been developed that can be used to detect credit card fraud. The artificial models that were developed included the decision tree, Bayesian network, random forest, support vector machine, Naïve Bayes, and K*models. Three proposed models were created from these basic AI models and included optimistic, pessimistic, and weighted voting strategies. Real credit card transactions were evaluated through the models. From the analysis, it was established that optimistic voting strategy enabled detection of 31.59% of all fraudulent transactions, pessimistic voting strategy detected 93.92% of all fraudulent transactions while weighted voting strategy detected 64.02% of all fraudulent transaction. The analysis showed that voting mechanisms that were based AI models can be used for fraud detection. The data provided data was relevant and accurate in the analysis of the AI models. One of the strengths of the data was that it made use of practical real credit card data to ascertain the transactions. The article is thus sound as it provides a practical application of how artificial intelligence can be used to solve credit card fraud. The article supports the thesis by showing how current AI technology can be used to prevent credit card fraud. It provides a practical way with high detection percentages that law enforcement can be used to prevent the ever-increasing credit card frauds. The article provides an overview of the current artificial intelligence techniques that are being used to combat fraud. The first section of the article deals with how AI can be used to detect credit card fraud and insider trading. The article then observes the use of genetic algorithms, neural nets, and statistical analysis to detect fraud and insider trading. The model that has been applied was referred to as MonITARS (Monitor Insider Trading and Regulatory Surveillance). The article accurately described the current challenges that law enforcement officers face with regard to internet fraud. It then accurately identified some of the AI technologies that can be used to detect fraud. Additionally, the article had a strength in providing a further development of a practical approach that can be used to fight fraud through the MonITARS system. However, one of the disadvantage of the article was that it did not consider empirical evidence and provide actual data regarding the evidence of that was gathered. The article is applicable to the topic under study of how law enforcement officers can combat fraud through AI by providing a system known as MonITARS that can be used to detect credit card fraud and insider trading.
References
Choi, D., & Lee, K. (2018). An Artificial Intelligence Approach to Financial Fraud Detection under IoT Environment: A Survey and Implementation. Security and Communication Networks , 2018 .
Delegate your assignment to our experts and they will do the rest.
Hasan, R., Raghav, A., Mahmood, S., & Hasan, M. A. (2011, November). Artificial intelligence based model for incident response. In 2011 International Conference on Information Management, Innovation Management and Industrial Engineering (Vol. 3, pp. 91-93). IEEE.
Kingston, J. K. (2017, May). Representing, reasoning and predicting fraud using fraud plans. In 2017 11th International Conference on Research Challenges in Information Science (RCIS) (pp. 128-133). IEEE.
Kültür, Y., & Çağlayan, M. U. (2017). Hybrid approaches for detecting credit card fraud. Expert Systems , 34 (2), e12191.
Wong, E. (2009). Employing Artificial Intelligence To Minimise Internet Fraud. International Journal of Cyber Society and Education , 2 (1), 61-72.
