Systems and methods for determining thin-file records and determining thin-file risk levels

• US 7,742,982 B2
• Filed: 10/12/2007
• Issued: 06/22/2010
• Est. Priority Date: 04/12/2007
• Status: Active Grant

• Current Assignee: Experian Marketing Solutions, LLC (Experian plc)
• Original Assignee: Experian Marketing Solutions Incorporated (Experian plc)
• Inventors: Hjermstad, Erik, Hadenfeldt, Nancy A., Chaudhuri, Anamitra
• Primary Examiner(s): Kazimi; Hani M
• Assistant Examiner(s): MADAMBA, CLIFFORD B
• Application Number:
  US11/871,572
• Publication Number:
  US 20080255975A1
• Time in Patent Office: 984 Days
• Field of Search: 705/38, 705/36, 705/99
• US Class Current: 705/38
• CPC Class Codes: G06Q 20/10 specially adapted for electronic funds transfer [EFT] systems; specially adapted for home banking systems
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes
• CPC Class Codes: G06Q 40/03 Credit; Loans; Processing thereof
• CPC Class Codes: G06Q 40/04 Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
• CPC Class Codes: G06Q 40/08 Insurance

Abstract

In some embodiments, systems and methods are disclosed for generating filters to determine whether a consumer is likely to have a scoreable credit record based on non-credit data, and to determine a potential risk level associated with an unscoreable credit record based on non-credit data. Existing scoreable and unscoreable records are compared to determine factors correlated with having an unscoreable record, and a multi-level filter is developed. Unscoreable records having at least one entry are compared to determine whether they are “good” or “bad” risks, factors correlated with either condition are determined, and a filter is developed. The filters can be applied to records comprising demographic data to determine consumers that are likely to have unscoreable records but represent good risks.

Claims

1. A computing system, comprising:

a computer-readable storage comprising a database and storing a plurality of records, the plurality of records comprising demographic data and credit data;

a processor configured to access the plurality of records in the computer-readable storage and configured to filter, from the plurality of records, records that include less than a minimum number of credit data entries;

a thin-file module configured to access the filtered records that include less than the minimum number of credit data entries and determine demographic data that is correlated with the filtered records by at least,

identifying demographic characteristics that appear within the demographic data of the filtered records; and

associating a weight with each of the demographic characteristics, the weight being based at least in part on the correlation of the demographic characteristic with a likelihood of a record being a thin-file record;

a data record filter configured to predict a likelihood that a record will include less than the minimum number of credit data entries by at least comparing the correlated demographic data to demographic data associated with the record, the demographic data comprising one or more of the following;

residence address data;

age data;

household data;

marital status data;

delinquent data for consumers in a geographic area;

data related to length of residency; and

,

a communications link configured to receive a customer record comprising a set of customer demographic data,

wherein the data record filter is further configured to generate a score corresponding to the likelihood of the customer record having less than the minimum number of credit data entries by at least,

locating, within the set of customer demographic data, demographic characteristics identified within the demographic data of the filtered records;

assigning a value to each located demographic characteristic in accordance with the weight associated with the demographic characteristic; and

,

combining the values assigned to the located demographic characteristics to generate the score.

2. The system of claim 1, wherein the minimum number of credit data entries is one.

3. A computer implemented method of developing a filter for thin-file records comprising:

accessing, by a computer processor, a database storing a plurality of records, each of the records comprising demographic data fields and credit data fields;

identifying, by a computer processor, a first subset of the plurality of records stored in the database, the first subset corresponding a first type comprising those records that include at least a minimum number of credit data entries;

identifying, by a computer processor, a second subset of the plurality of records stored in the database, the second subset of records corresponding to a second type comprising those records that do not include at least the minimum number of credit data entries;

determining, by a computer processor, demographic data fields that are correlated with the first type of records, the determining further comprising;

identifying, by a computer processor, demographic characteristics that appear within the demographic data fields of the first type of records; and

,

associating, by a computer processor, a weight with each of the demographic characteristics, the weight being based at least in part on the correlation of the demographic characteristic with a likelihood of a record being the first type of records;

developing, by a computer processor, a data record filter that predicts a likelihood of a record comprising demographic data fields being of the first type by at least comparing the demographic data fields of the record to the correlated demographic data fields;

receiving, by a computer processor, a set of customer data comprising customer demographic data; and

generating, by a computer processor, a score by at least,

locating, by a computer processor, within the customer demographic data, demographic characteristics identified within the demographic data fields of the first type of records;

assigning, by a computer processor, a value to each located demographic characteristic in accordance with the weight associated with the demographic characteristic; and

combining, by a computer processor, the values assigned to the located demographic characteristics to generate the score,

wherein the demographic data fields comprise one or more of the following;

residence address data;

age data;

household data;

marital status data;

delinquent data for consumers in a geographic area; and

data related to length of residency.

4. The computer implemented method of claim 3, wherein the minimum number of credit data entries is one.

5. The computer implemented method of claim 3, wherein the second subset is a thin-file collection.

6. A computing system, comprising:

a computer-readable storage comprising a database and storing a plurality of records, the plurality of records comprising demographic data and credit data, the records comprising thin-file records;

a processor configured to access the plurality of thin-file records in the computer-readable storage and configured to filter the thin-file records according to whether the credit data in the thin-file records corresponds to a positive credit activity;

a thin-file module configured to access the filtered thin-file records having credit data corresponding to a positive credit activity and determine demographic data that is correlated with the filtered thin-file records by at least,

identifying demographic characteristics that appear within the demographic data of the filtered thin-file records; and

associating a weight with each of the demographic characteristics, the weight being based at least in part on the correlation of the demographic characteristic with a likelihood of a record having credit data corresponding to a positive credit activity;

a data record filter configured to predict a likelihood of thin-file records having credit data corresponding to a positive credit activity based at least in part on the correlated demographic data; and

a communications link configured to receive a record of a thin-file customer comprising a set of customer demographic data associated with the thin-file customer,

wherein the data record filter is further configured to generate a score corresponding to the likelihood of the record of the thin-file customer having credit data corresponding to a positive credit activity by at least,

locating, within the set of customer demographic data, demographic characteristics identified within the demographic data of the filtered thin-file records;

assigning a value to each located demographic characteristic in accordance with the weight associated with the demographic characteristic; and

combining the values assigned to the located demographic characteristics to generate the score, and

wherein the demographic data comprises one or more of the following;

residence address data;

age data;

household data;

marital status data;

delinquent data for consumers in a geographic area; and

data related to length of residency.

7. A computer implemented method of generating a filter for determining a risk level associated with a thin-file record, comprising:

accessing, by a computer processor, a database storing a plurality of thin-file records, the thin-file records comprising demographic data and credit data;

monitoring, by a computer processor, the plurality of thin-file records in order to determine a status of each of the records based on the credit data of the record, the status indicating a credit risk;

determining, by a computer processor, demographic data that is correlated with the status of each of the records, the determining further comprising;

identifying, by a computer processor, demographic characteristics that appear within the demographic data of the thin-file records having credit data reflective of a credit risk; and

associating, by a computer processor, a weight with each of the demographic characteristics, the weight being based at least in part on the correlation of the demographic characteristic with a likelihood of a record having credit data reflective of a credit risk;

developing, by a computer processor, a data record filter that predicts the probable credit risk of a record based on the correlated demographic data;

receiving, by a computer processor, a set of customer data associated with a thin-file customer, the customer data comprising customer demographic data corresponding to at least a portion of the correlated demographic data; and

generating, by a computer processor, a score, the generating further comprising;

locating, by a computer processor, within the customer demographic data, demographic characteristics identified within the demographic data of the thin-file records having credit data reflective of a credit risk;

assigning, by a computer processor, a value to each located demographic characteristic in accordance with the weight associated with the demographic characteristic; and

combining, by a computer processor, the values assigned to the located demographic characteristics to generate the score,

wherein the demographic data comprises one or more of the following;

residence address data;

age data;

household data;

marital status data;

delinquent data for consumers in a geographic area; and

data related to length of residency.

8. A computer implemented method for predicting the likelihood of a consumer having limited credit history, the method comprising:

receiving, by a computer processor, a data record of a consumer, the data record comprising consumer demographic data;

creating, by a computer processor, a data record filter that identifies a plurality of demographic characteristics in known thin-file records and a correlation for each of the plurality of demographic characteristics to a likelihood of a record being a thin-file record;

generating, by a computer processor, a score indicating a likelihood of the data record of the consumer being a thin-file record by at least;

locating, by a computer processor, within the consumer demographic data, demographic characteristics identified by the data record filter as present in the demographic data of known thin-file records;

assigning, by a computer processor, a value to each located demographic characteristic in accordance with a weight associated with each of the demographic characteristics, the weight being based at least in part on the correlation of the demographic characteristic with a likelihood of a record being a thin-file record;

combining, by a computer processor, the values assigned to the located demographic characteristics to generate the score,

wherein the demographic data comprises one or more of the following;

residence address data;

age data;

household data;

marital status data;

delinquent data for consumers in a geographic area; and

data related to length of residency.