Detecting and Measuring Risk with Predictive Models Using Content Mining

• US 20090234683A1
• Filed: 10/04/2007
• Published: 09/17/2009
• Est. Priority Date: 06/30/2000
• Status: Active Grant

• Current Assignee: Fair Isaac Corporation
• Original Assignee: Fair Isaac Corporation
• Inventors: Anderson, Russell, Dulany, Paul C., Baker, James C., Shao, Xuhui, Milana, Joseph P., Peranich, Larry S., Dungca, Ricardo M., Hassibi, Khosrow M.
• Granted Patent: US 8,032,448 B2
• Time in Patent Office: Days
• US Class Current: 705/7
• CPC Class Codes: G06Q 20/10 specially adapted for electronic funds transfer [EFT] systems; specially adapted for home banking systems
• CPC Class Codes: G06Q 40/03 Credit; Loans; Processing thereof
• CPC Class Codes: G06Q 40/08 Insurance
• CPC Class Codes: G06Q 40/12 Accounting

Abstract

Computer implemented methods and systems of processing transactions to determine the risk of transaction convert high categorical information, such as text data, to low categorical information, such as category or cluster IDs. The text data may be merchant names or other textual content of the transactions, or data related to a consumer, or any other type of entity which engages in the transaction. Content mining techniques are used to provide the conversion from high to low categorical information. In operation, the resulting low categorical information is input, along with other data, into a statistical model. The statistical model provides an output of the level of risk in the transaction. Methods of converting the high categorical information to low categorical clusters, of using such information, and other aspects of the use of such clusters are disclosed.

Claims

1. A system for detecting risk in a transaction, comprising the steps of:

a database of unique merchant names, each merchant name associated with a merchant cluster and each merchant name is textual data or other high categorical data;

a transaction processing component that receives a transaction between a consumer and a entity, that derives transaction data from the transaction, and determine a unique entity identity for the entity from the database; and

a statistical model that receives the data derived from the transaction and the unique entity identity, and outputs a score indicative of the level of risk in a transaction.

2. A method of determining a level of risk in a transaction, the method comprising:

receiving a transaction between a first entity and a second entity;

deriving high categorical information elements from at least one of transaction, the first entity or the second entity, wherein the high categorical information elements are text data;

determining a low categorical information cluster closest to the high categorical information elements; and

applying the low categorical information cluster and data derived from the transaction to a predictive model, and outputting the level of risk in the transaction to detect if the transaction is fraudulent.

3. The method of claim 2, wherein the second entity a entity of the transaction, and the high categorical information elements are entity notes associated with the transaction.

4. The method of claim 2, further comprising:

selecting a plurality of high categorical information elements;

associating each high categorical information element with a context vector in

a vector space, such that high categorical information elements that frequently proximally co-occur in the transactions have context vectors that are similarly oriented in the vector space;

clustering the context vectors of the high categorical information elements into a number of clusters less than the number of high categorical information elements, each cluster being a low categorical information cluster;

wherein determining a low categorical information cluster closest to the high categorical information elements further comprises determining the low categorical information cluster closest in the vector space to a context vector derived from the context vectors of the high categorical information elements.

5. A computer implemented method for determining a level of risk of a transaction between a consumer and an entity, comprising the steps of:

storing a plurality of entity clusters, the entity clusters determined from statistical co-occurrences of the entity identified in a plurality of transactions, wherein said entity identified are entity identifiers or other high categorical data;

receiving data from said transaction between said consumer and said entity;

determining one of the plurality of entity clusters associated with the entity of the transaction based on the entity'"'"'s identity, wherein said entity'"'"'s identity is an entity identifier or other high categorical data; and

applying the entity cluster in conjunction with data derived from the transaction to a predictive model, and outputting a level of risk of the transaction to detect if the transaction is fraudulent.

6. The method of claim 5, further comprising the step of:

estimating a likelihood that the transaction is fraudulent.

7. The method of claim 5, further comprising the step of:

responsive to the level of risk of the transaction, determining whether to approve the transaction, decline the transaction, or obtain additional information regarding the transaction or the consumer.

8. The method of claim 5, wherein the step of determining one of the plurality of entity clusters associated with the entity further comprises the steps of:

storing a lookup table associating each entity cluster, with at least one entity identity, wherein the entity identities are each unique; and

applying the entity'"'"'s identity to the lookup table to determine the associated entity cluster.

9. The method of claim 8, wherein the unique entity identities are derived from a plurality of raw entity identities in transaction data by stemming and equivalencing the raw entity identities.

10. The method of claim 5, further comprising the steps of:

storing for each entity cluster a risk factor indicative of the likelihood that transactions at entities within the entity cluster are fraudulent; and

applying the risk factor of the entity cluster to the predictive model.

11. The method of claim 10, wherein the risk factor is an estimate of the percentage of transactions at entities within the entity cluster that are fraudulent.

12. The method of claim 5, further comprising the steps of:

storing a plurality of consumer clusters;

storing for each combination of a consumer cluster and an entity cluster a risk factor indicative of the likelihood that transactions by consumers in the consumer cluster at entities within the entity cluster are fraudulent;

determining a current consumer cluster associated with the consumer; and

applying the risk factor of the combination of the current consumer cluster and the entity cluster to the predictive model.

13. The method of claim 12, wherein the risk factor is an estimate of the percentage of transactions at entities within the entity cluster by consumers in the consumer cluster that are fraudulent.

14. The method of claim 5, further comprising the steps of:

storing for each entity cluster at least one summarized transaction statistic, descriptive of transactions occurring at entities in the entity cluster; and

applying the at least one summarized transaction statistic of the entity cluster to the predictive model.

15. The method of claim 14, wherein the at least one summarized transaction statistic are selected from a group consisting of:

average transaction amount; and

average transaction volume.

16. The method of claim 5, further comprising the steps of:

storing for each of a plurality of consumer clusters, at least one summarized transaction statistic, descriptive of transactions by consumers in the consumer cluster; and

applying the at least one summarized transaction statistic of the consumer cluster to the predictive model.

17. A computer implemented method for determining a level of risk of a transaction between a consumer and an entity, comprising the steps of:

storing a plurality of entity clusters, the entity clusters determined from statistical co-occurrences of the entity identities in a plurality of transactions, wherein said entity identities are entity identifiers or other high categorical data;

receiving data from said transaction between said consumer and said entity;

determining one of the plurality of entity clusters associated with the entity of the transaction based on the entity identity, wherein said entity'"'"'s identity is an entity identifier or other high categorical data;

determining an affinity measure of an affinity of a consumer to the entity cluster; and

applying the affinity measure in conjunction with data derived from the transaction to a predictive model, and outputting the level of risk of the transaction to detect if the transaction is fraudulent.

18. The method of claim 17, wherein the step of determining the affinity measure of an affinity of the consumer to the entity cluster further comprises the step of:

determining an affinity vector of the affinity of the consumer to each of a plurality of entity clusters, including the entity cluster of the entity of the transaction.

19. The method of claim 17, wherein the step of determining the affinity measure of an affinity of the consumer to the entity cluster further comprises the steps of:

determining a consumer cluster associated with the consumer; and

determining an affinity measure of the affinity of the consumer cluster to the entity cluster.

20. The method of claim 17, wherein the step of determining the affinity measure of an affinity of the consumer to the entity cluster further comprises the steps of:

determining a consumer cluster associated with the consumer; and

determining an affinity vector of the affinity of the consumer cluster to each of a plurality of entity clusters, including the entity cluster of the entity of the transaction.

21. A method of determining the level of risk in a transaction by a consumer, comprising the steps of:

storing a plurality of entity clusters, the entity clusters determined from statistical co-occurrences of the entity identities in a plurality of transactions, wherein said entity identities are entity identifiers or other high categorical data;

receiving data of a current transaction between a consumer and entity;

determining a predicted entity cluster in which the consumer is predicted to have a future transaction based on transactions of the consumer prior to the current transaction;

determining an actual entity cluster associated with the entity of the transaction based on the entity identity, wherein said entity identity is an entity identifier or other high categorical data;

determining a difference measure between the predicted entity cluster and the actual entity cluster; and

applying the difference measure in conjunction with data derived from the transaction to a predictive model, and outputting the level of risk of the transaction to detect if the transaction is fraudulent.