System and method of detecting mortgage related fraud
View Patent ↗Embodiments include systems and methods of detecting fraud. In particular, one embodiment includes a system and method of detecting fraud in mortgage applications. For example, one embodiment includes a computerized method of detecting fraud that includes receiving mortgage data associated with an applicant and at least one entity related to processing of the mortgage data, determining a first score for the mortgage data based at least partly on a first model that is based on data from a plurality of historical mortgage transactions associated with the entity, and generating data indicative of fraud based at least partly on the first score. Other embodiments include systems and method of generating models for use in fraud detection systems.
1. A computerized method of detecting fraud, the method comprising:
determining, by a computing system, based on mortgage data associated with an applicant, a first score for the mortgage data based at least partly on a first model that is based at least in part on historical data from a plurality of historical mortgage transactions that have each been identified as fraudulent or non-fraudulent, the first score indicative of a likelihood of the mortgage data being associated with a fraudulent transaction, the first model being generated based on one or more input data points within the historical data that are each correlated with a probability of fraud;
generating, by computing system, at least one risk indicator for the mortgage data based at least partly on said first model, the risk indicator identifying a portion of the mortgage data that is indicative of a potential source of fraud; and
generating a report comprising data indicative of whether the mortgage data is associated with a fraudulent transaction, said data comprising data reflective of at least the first score and the at least one risk indicator.
2. The method of claim 1 , wherein the mortgage data comprises at least one of mortgage application data, funded mortgage data, or bid tapes.
3. The method of claim 1 , wherein the plurality of historical mortgage transactions comprise a plurality of historical mortgage applications.
4. The method of claim 1 , further comprising determining, by the computing system, a second score for the mortgage data based at least partly on a second model that is based on data from a plurality of historical mortgage transactions associated with a plurality of entities related to processing of the transactions, wherein said report further comprises data reflective of the second score.
5. The method of claim 4 , wherein the plurality of entities include at least one of an account executive, a broker, a loan officer, or an appraiser.
6. The method of claim 1 , wherein the first model is a computer data model that comprises at least one of neural network, logistic regression, linear regression, decision trees, a classification and regression tree (CART) model, and an expert system.
7. The method of claim 6 , wherein said first model is retrieved by the computing system from an electronic database.
8. The method of claim 1 , wherein generating said at least one risk indicator of fraud further comprises:
identifying within the mortgage data, presence of at least one of said one or more input data points that is each correlated with a probability of fraud in said first model.
9. The method of claim 1 , wherein the first model includes at least one cluster of at least a portion of the data of the plurality of historical mortgage transactions.
10. The method of claim 9 , wherein determining the first score comprises comparing data of the mortgage data with the at least one cluster.
11. The method of claim 1 , wherein the mortgage data comprises at least one of: data indicative of at least one person associated with a mortgage application, data indicative of income of a mortgage applicant, and data indicative of a property subject to a mortgage transaction.
12. The method of claim 1 , further comprising selecting an application associated with the mortgage data for further review based on the data indicative of whether the mortgage data is associated with a fraudulent transaction, where said selecting can be performed manually or by the computing system.
13. The method of claim 1 , wherein generating the at least one risk indicator is performed substantially in real time.
14. The method of claim 1 , wherein generating the at least one risk indicator is performed in a batch mode.
15. The method of claim 1 further comprising:
determining, by the computing system, a second score for the mortgage data based at least partly on one or more risk rule models comprising rules that identify data patterns indicative of fraud; and
generating, by the computing system, a weighted score based at least in part on the first and second scores,
wherein said data indicative of whether the mortgage data is associated with a fraudulent transaction comprises data indicative of at least the weighted score.
16. The method of claim 1 , further comprising: generating, by the computing system, said first model by using at least one of a neural network, logistic regression, linear regression, decision trees, a classification and regression tree (CART) model, and an expert system.
17. The method of claim 1 , further comprising: retrieving, by the computing system, said mortgage data from an electronic database.
18. The method of claim 1 , further comprising: retrieving, by the computing system, said historical data from an electronic database.
19. The method of claim 1 , further comprising repeating said determining and generating for data associated with a statistically significant number of mortgage applicants.
20. A system for detecting fraud, the system comprising:
a storage system configured to store mortgage data associated with an applicant; and
a computing system, comprising one or more computing devices, configured to implement:
a score integrator component configured to:
determine a first score for the mortgage data based at least partly on a first model that is based at least in part on historical data from a plurality of historical mortgage transactions that have each been identified as fraudulent or non-fraudulent, the first score indicative of a likelihood of the mortgage data being associated with a fraudulent transaction, the first model being generated based on one or more input data points within the historical data that are each correlated with a probability of fraud; and
generate at least one risk indicator for the mortgage data based at least partly on said first model, the risk indicator identifying a portion of the mortgage data that is indicative of a potential source of fraud; and
a report component configured to generate a report comprising data indicative of whether the mortgage data is associated with a fraudulent transaction, said data comprising data reflective of at least the first score and the at least one risk indicator.
21. The system of claim 20 , wherein said score integrator component is further configured to determine a second score for the mortgage data based at least partly on a second model that is based on data from a plurality of historical mortgage transactions associated with a plurality of entities related to processing of the transactions, wherein said report comprises data reflective of the second score.
22. The system of claim 21 , wherein the score integrator component is further configured to generate at least one risk indicator of fraud based on said second model.
23. The system of claim 20 , wherein the first model includes at least one cluster of at least a portion of the data of the plurality of historical mortgage transactions and wherein the score integrator component determines the first score at least in part by comparing data of the mortgage data with the at least one cluster.
24. The system of claim 20 , wherein the computing system is further configured to select an application associated with the mortgage data for further review based at least partly on the data indicative of whether the mortgage data is associated with a fraudulent transaction.
25. The system of claim 20 , wherein the computing system is further configured to implement:
a model generator component to:
generate said first model based at least in part on said historical data; and
generate a second model based on data from a plurality of historical mortgage transactions associated with at least one entity related to processing of the mortgage transactions;
wherein the computing system is further configured to:
store the first model in a first computer readable medium; and
store the second model in a second computer readable medium.
26. The system of claim 20 , wherein the first model is a computer data model generated by a model generator component implemented in the computing system.
27. The system of claim 26 , wherein the first model is generated based on at least one of a neural network, logistic regression, linear regression, decision trees, a classification and regression tree (CART) model, and an expert system.
28. The system of claim 20 , wherein the storage system comprises one or more electronic databases.