Method, apparatus and article for assigning a similarity measure to names
View Patent ↗A method, article and apparatus for calculating a measure of similarity between names is disclosed. Generally, name elements from an inbound name may be matched to name elements in a candidate matching name. A measure of similarity for the inbound name and candidate matching name may be calculated based on the number of matching name elements, an inbound name word count, a candidate name word count, and the number of transpositions that occur in matching name elements with one another.
1. A method of determining a measure of similarity between an inbound name and a candidate matching name, comprising:
identifying one or more matches between name elements of the inbound name and name elements of the candidate matching name, wherein each match is identified based on a correlation measure between the matched name element of the inbound name and the matched name element of the candidate matching name;
for the identified matches, determining a number of transpositions in an order of name elements of the inbound name elements relative to the name elements of the candidate matching name; and
calculating a measure of similarity based on the number of matching name elements, an inbound name word count, a candidate matching name word count, and the number of transpositions.
2. The method of claim 1 , wherein the correlation measure between the matched name element of the inbound name and the matched name element of the candidate matching name is determined using the Jaro-Winkler measure of string correlation.
3. The method of claim 1 , further comprising standardizing the name elements of the inbound name or the candidate matching name using a name dictionary.
4. The method of claim 3 , wherein the measure of similarity between the inbound and name and the candidate matching name is reduced to reflect a standardization penalty.
5. The method of claim 1 , wherein the measure of similarity is calculated according to the equation of:
(⅓)*(Number of matching name elements/inbound name word count)+(⅓)*(Number of matching name elements/candidate matching name word count)+(⅓)*((Number of matching name elements−Number of transpositions)/(Number of matching name elements)).
6. The method of claim 1 , wherein the name elements include a first name element, a middle name element, and a surname name element.
7. The method of claim 1 , wherein the measure of similarity between the inbound name and the candidate matching name is calculated based on a subset of the name elements in the inbound name or in the candidate matching name.
8. A computer readable storage medium containing a program which, when executed, performs an operation for determining a measure of similarity between an inbound name and a candidate matching name, comprising:
identifying one or more matches between name elements of the inbound name and name elements of the candidate matching name, wherein each match is identified based on a correlation measure between the matched name element of the inbound name and the matched name element of the candidate matching name;
for the identified matches, determining a number of transpositions in an order of name elements of the inbound name elements relative to the name elements of the candidate matching name; and
calculating a measure of similarity based on the number of matching name elements, an inbound name word count, a candidate matching name word count, and the number of transpositions.
9. The computer readable storage medium of claim 8 , wherein the correlation measure between the matched name element of the inbound name and the matched name element of the candidate matching name is determined using the Jaro-Winkler measure of string correlation.
10. The computer readable storage medium of claim 8 , wherein the operations further comprise, standardizing the name elements of the inbound name or the candidate matching name using a name dictionary.
11. The computer readable storage medium of claim 10 , wherein the measure of similarity between the inbound and name and the candidate matching name is reduced to reflect a standardization penalty.
12. The computer readable storage medium of claim 8 , wherein the measure of similarity is calculated according to the equation of:
(⅓)*(Number of matching name elements/inbound name word count)+(⅓)*(Number of matching name elements/candidate matching name word count)+(⅓)*((Number of matching name elements−Number of transpositions)/(Number of matching name elements)).
13. The computer readable storage medium of claim 8 , wherein the name elements include a first name element, a middle name element, and a surname name element.
14. The computer readable storage medium of claim 8 , wherein the measure of similarity between the inbound name and the candidate matching name is calculated based on a subset of the name elements in the inbound name or in the candidate matching name.
15. A system, comprising:
a processor; and
a memory containing a program configured to determine a measure of similarity between an inbound name and a candidate matching name by:
identifying one or more matches between name elements of the inbound name and name elements of the candidate matching name, wherein each match is identified based on a correlation measure between the matched name element of the inbound name and the matched name element of the candidate matching name;
for the identified matches, determining a number of transpositions in an order of name elements of the inbound name elements relative to the name elements of the candidate matching name; and
calculating a measure of similarity based on the number of matching name elements, an inbound name word count, a candidate matching name word count, and the number of transpositions.
16. The system of claim 15 , wherein the correlation measure between the matched name element of the inbound name and the matched name element of the candidate matching name is determined using the Jaro-Winkler measure of string correlation.
17. The system of claim 15 , wherein the name elements of the inbound name or the candidate matching name are standardized using a name dictionary.
18. The system of claim 17 , wherein the measure of similarity between the inbound and name and the candidate matching name is reduced to reflect a standardization penalty.
19. The system of claim 15 , wherein the measure of similarity is calculated according to the equation of:
(⅓)*(Number of matching name elements/inbound name word count)+(⅓)*(Number of matching name elements/candidate matching name word count)+(⅓)*((Number of matching name elements−Number of transpositions)/(Number of matching name elements)).
20. The system of claim 15 , wherein the name elements include a first name element, a middle name element, and a surname name element.
21. The system of claim 15 , wherein the measure of similarity between the inbound name and the candidate matching name is calculated based on a subset of the name elements in the inbound name or in the candidate matching name.