Date of this Version
Nguyen, Quan. 2021. Ameritas: Fraud Identification and Notification. Undergraduate Honors Thesis. University of Nebraska-Lincoln.
The goal of this project was to create a fraud identification and notification solution for Ameritas. The solution we created runs weblog data through a series of rules that check for potentially fraudulent activity. Weblog interactions that are identified as potentially fraudulent are flagged and made available to Ameritas agents on our dashboard. Our team proposes a dashboard and a business rules engine to help Ameritas solve this problem. The business rules engine will be responsible for evaluating each record of users and marking whether such activity is suspicious or not. The rules engine consists of a set of established rules that are easy to change and update so that Ameritas can add, change, and/or remove rules as they will. On the other hand, the dashboard will be the primary interface that Ameritas will interact with the data. The dashboard will display which users have suspicious activities and which rules users break. The dashboard will also provide additional information so that the Ameritas team can carry an investigation on these activities. It also has feedback features so that the Ameritas team can mark either a flagged record as correctly or incorrectly flagged. Such information will be useful later on when building a machine learning model. In the end, the team was able to successfully deliver both the business rules engine and the dashboard to Ameritas.