Predictive Modeling for Identity Theft Detection: A Design Science Approach Using Machine Learning and Historical Data

Aim/Purpose The proliferation of online banking has led to a significant increase in identity theft and cybercrimes, creating an urgent need to develop more effective predictive models to detect and prevent such fraudulent activities using advanced technological approaches. Background This study addresses the critical gap in the existing literature by exploring how historical identity theft victim data can be leveraged through supervised ma-chine learning to create a comprehensive prediction model for identifying and preventing future identity theft incidents. Methodology A design science quantitative-focused research approach was employed, analyzing public records of identity theft victims in the United States from 2016. The study utilized a nonexperimental research design to compare and group data sets by number of victims and types of identity theft, applying an onto-logical research philosophy to examine the potential of supervised machine learning in prediction. Contribution The research contributes to the body of knowledge by demonstrating a novel approach to identity theft prevention through data-driven predictive modeling, bridging the gap between historical victim data and machine learning technologies. Findings The study confirmed that supervised machine learning can be effectively used to create an identity theft prediction model using historical victim data, providing insights into potential detection and prevention strategies for cybercrime. Recommendations for Practitioners (1) Implement advanced machine learning models for early identity theft detection. (2) Develop comprehensive data collection strategies for historical victim information. (3) Integrate predictive analytics into existing fraud prevention systems. Recommendations for Researchers (1) Expand the dataset to include more recent and diverse identity theft incidents. (2) Develop more sophisticated machine learning algorithms. (3) Investigate cross-border identity theft patterns. (4) Explore the integration of multiple data sources for improved prediction accuracy. Impact on Society The research offers a potential breakthrough in combating identity theft, potentially reducing financial losses, protecting individual privacy, and enhancing overall cybersecurity for online banking and financial transactions. Future Research (1) Developing real-time prediction models. (2) Exploring artificial intelligence techniques for more dynamic fraud detection. (3) Investigating the psychological and social factors contributing to identity theft.