Artificial intelligence trending system

• US 20070094181A1
• Filed: 09/18/2006
• Published: 04/26/2007
• Est. Priority Date: 02/07/2001
• Status: Active Grant

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: MCI Incorporated (Verizon Communications Inc.)
• Inventors: Van Arkel, John, Tayebnejad, Mohammad, Dallas, Charles, Van Camp, Karl
• Granted Patent: US 7,403,931 B2
• Time in Patent Office: Days
• US Class Current: 706/21
• CPC Class Codes: G06Q 40/08 Insurance

Abstract

A data processing system program to develop, train, and implement a neural network for identifying customers who represent a bad debt risk is disclosed. A feature vector is applied to a neural network to generate outputs that approximate the relative likelihood that customers who are the subjects of the records used to generate the feature vector will be a bad debt risk. Statistical values relating categorical attributes of the customers to the likelihood of their becoming a bad debt risk are substituted for the categorical attributes, and the attributes are normalized before the feature vector is applied to the network. In one embodiment the customers are customers of a long distance service provider.

Claims

1. (canceled)

2. A computer-implemented method comprising:

determining a first risk probability according to an artificial intelligence (AI) score, wherein the AI score is generated based on information corresponding to an account;

determining a second risk probability according to a balance value of the account; and

outputting a prioritization value based on the first risk probability and the second risk probability, wherein the prioritization value specifies priority for reviewing the information.

3. A computer-implemented method according to claim 2, further comprising:

receiving the AI score, wherein the AI score is generated based on attributes of the information, the attributes being derived from traffic data.

4. A computer-implemented method according to claim 3, wherein the attributes include customer data.

5. A computer-implemented method according to claim 2, further comprising:

transmitting the AI score and the prioritization value to a credit risk management system.

6. A computer-implemented method according to claim 2, wherein the first risk probability relates to risk of non-payment for a service.

7. A computer-implemented method according to claim 2, wherein the account is managed by a service provider of telecommunication services, and the information includes deactivation information.

8. A computer-implemented method according to claim 2, wherein the prioritization value is determined according to:

Priority=[Prob(AI)+Prob(balance)]/[MaxProb(AI)+MaxProb(balance)], wherein Prob(AI) represents the first risk probability, MaxProb(AI) represents the maximum value of the Prob(AI), Prob(balance) represents the second risk probability, and MaxProb(balance) represents the maximum value of the Prob(balance).

9. A system comprising:

a processor configured to determine a first risk probability according to an artificial intelligence (AI) score that is generated based on information corresponding to an account,

wherein the processor is further configured to determine a second risk probability according to a balance value of the account, and to output a prioritization value based on the first risk probability and the second risk probability, the prioritization value specifying priority for reviewing the information.

10. A system according to claim 9, wherein the AI score is generated based on attributes of the information, the attributes being derived from traffic data.

11. A system according to claim 10, wherein the attributes include customer data.

12. A system according to claim 9, wherein the AI score and the prioritization value are transmitted to a credit risk management system.

13. A system according to claim 9, wherein the first risk probability relates to risk of non-payment for a service.

14. A system according to claim 9, wherein the account is managed by a service provider of telecommunication services, and the information includes deactivation information.

15. A system according to claim 9, wherein the prioritization value is determined according to:

Priority [Prob(AI)+Prob(balance)]/[MaxProb(AI)+MaxProb(balance)], wherein Prob(AI) represents the first risk probability, MaxProb(AI) represents the maximum value of the Prob(AI), Prob(balance) represents the second risk probability, and MaxProb(balance) represents the maximum value of the Prob(balance).

16. A computer-implemented method comprising:

selecting a group of records as training records for financial trending;

selecting a current topology and learning algorithm to configure a neural network;

applying attributes from the training records and characteristic values for subjects of the training records to the network, wherein the attributes are selected based on scores generated by a relevance analysis;

selecting a group of the records as evaluation records;

applying attributes from the evaluation records to the network to generate outputs for the evaluation records;

ordering the evaluation records in rank order in accordance with the outputs for the evaluation records;

evaluating the rank order of the evaluation records in accordance with predetermined criteria; and

modifying the current topology or the learning algorithm or both to configure the network;

generating a plurality of neural networks; and

selecting one of the plurality of neural networks according to a criteria.

17. A computer-implemented method according to claim 16, further comprising:

performing the relevance analysis according to different evaluation methods.

18. A computer-implemented method according to claim 17, wherein the evaluation methods include a chi square test, an uncertainty test, an information gain test, and a gini test.

19. A computer-implemented method comprising:

estimating a statistic relating values of a categorical attribute to a characteristic of subject among a plurality of subjects;

for each of the subjects, processing a plurality of attributes, including the categorical attribute to generate an input vector about each subject, by substituting a value of the statistic for corresponding values of the categorical attribute, wherein the attributes are selected on the basis of a level of significance as determined by a relevance analysis employing a plurality of different evaluation methods; and

for each of the subjects, generating an output value as a function of the input vector.

20. A computer-implemented method according to claim 19, wherein the function is defined by a trained neural network and an associated set of weights.

21. Sew) A computer-implemented method according to claim 19, wherein the evaluation methods include a chi square test, an uncertainty test, an information gain test, and a gini test.