Computer-based method and system for linking records in data files

• US 6,658,412 B1
• Filed: 06/30/1999
• Issued: 12/02/2003
• Est. Priority Date: 06/30/1999
• Status: Expired due to Fees

• Current Assignee: Educational Testing Service
• Original Assignee: Educational Testing Service
• Inventors: Jenkins, Laura, Jirele, Thomas J.
• Primary Examiner(s): Trammell, James P.
• Assistant Examiner(s): CHEUNG, MARY DA ZHI WANG
• Application Number:
  US09/345,825
• Time in Patent Office: 1,616 Days
• Field of Search: 707/1-6
• US Class Current: 1/1
• CPC Class Codes: G06F 16/10 File systems; File servers
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access
• CPC Class Codes: Y10S 707/99936 Pattern matching access

Abstract

The present invention relates to computer-based technology for linking or matching records in data files, based on at least one identifier in common, with a threshold probability that records are linked, the method uses a Bayesian probabilistic approach to determine the likelihood that the identified records are linked.

Claims

1. A computerized method for linking records in data files, based on at least one identifier in common, with a threshold probability that records are linked, the method comprising the steps of:

a. identifying records in the data files having a first identifier in common,

b. using a Bayesian probabilistic approach to determine likelihood that identified records, having a first identifier in common, are linked,

c. linking identified records, having a first identifier in common, whose likelihood exceeds the threshold for linking identified records having a first identifier in common.

2. The method of claim 1, further comprising the steps of:

a. identifying records in the data files, not already linked, having a second identifier in common,

b. using a Bayesian probabilistic approach to determine likelihood that identified records, not already linked and having a second identifier in common, are linked,

c. linking identified records, not already linked and having a second identifier in common, whose likelihood exceeds the threshold for Linking identified records, not already linked and having a second identifier in common.

3. The method of claim 2, further comprising the steps of:

a. identifying records in the data files not already linked having a third identifier in common,

b. using a Bayesian probabilistic approach to determine likelihood that identified records, not already linked and having a third identifier in common, are linked,

c. linking identified records, not already linked and having a third identifier in common, whose likelihood exceeds the threshold for linking identified records, not already linked and having a third identifier in common.

4. The method of claim 3, further comprising the steps of continuing the steps of claim 3 for additional identifiers until either all records are linked or all identifiers have been used.

5. The method of claim 1, wherein the identifier is one of the following items:

social security number;

name (last name, first name and middle initial);

day and month of birth;

or gender.

6. A computerized system for linking records in data files, based on at least one identifier in common, with a threshold probability that records are correctly linked, comprising:

a. means for identifying records in the data files having a first identifier in common,

b. means for using a Bayesian probabilistic approach to determine likelihood that identified records, having a first identifier in common, are linked,

c. means for linking identified records, having a first identifier in common, whose likelihood exceeds the threshold for linking identified records having a first identifier in common.

7. A computerized system for linking records in data files according to claim 6, which is implemented using ACESMAIN.F90;

THREE FILE MAIN PROGRAM;

COHORT MAIN PROGRAM;

MATFILES;

CHECKHOLD;

MATCHSCR;

GET2REX;

EXPAND and DDT.