Recognizing and combining redundant merchant designations in a transaction database

A system for determining whether a subject merchant location database entry and a candidate merchant location database entry are describing the same merchant location, the system comprising:

• a machine-readable database storing merchant location database entries, each merchant location database entry having a DBA name text field designating the doing business as (DBA) name of each respective merchant location, a street address text field designating the street address of the location of each respective merchant location, and one or more additional descriptive fields descriptive of one or more predetermined characteristics of the respective merchant location;
• a processor;
• and
• a tangible machine-readable memory device storing a program of instructions thereon which, when executed by the processor, cause the processor to:
• designate a merchant location database entry appearing in a master merchant location database or a transaction data stream to be compared as a subject merchant location database entry;
• populate a set with one or more candidate merchant location database entries located in a data warehouse database maintained by a network operator for comparison to the subject merchant location database entry, wherein each candidate merchant location database entry selected as a member of the set is chosen from the machine-readable database for having a predetermined minimum textural similarity with the subject merchant location database entry on the basis of each database entry's respective DBA name text field or street address text field;
• compare the subject merchant location database entry with each of the candidate database entries on the basis of the one or more additional descriptive fields retrieved from the data warehouse database;
• perform a logistic regression using the results of the comparing to calculate a probability that the merchant location corresponding to the subject merchant location database entry and the merchant location corresponding to one or more of the candidate merchant location database entries are the same merchant location;
• and
• output the results of the logistic regression,
• wherein the one or more additional descriptive fields retrieved from the data warehouse include a classification code, the classification code derived from a hierarchical classification, and the comparing determines whether the subject merchant location database entry or the candidate merchant location database entry includes a classification code related to an industry which is experientially known to have merchant location identification data that is either more stable than or less stable than other industries, and the logistic regression weights a merchant location classification code with regard to whether the related industry is known to have more or less stable merchant location identification data.

A non-transitory machine-readable medium storing a program of instructions thereon which, when executed by a processor, cause the processor to:

• designate a merchant location database entry appearing in a master merchant location database or a transaction data stream to be compared as a subject merchant location database entry, the subject merchant location database entry having a DBA name text field designating the doing business as (DBA) name of the merchant location, a street address text field designating the street address of the location of the merchant location, and one or more additional descriptive fields descriptive of one or more predetermined characteristics of the merchant location;
• populate a set with one or more candidate merchant location database entries located in a data warehouse database maintained by a network operator for comparison to the subject merchant location database entry, selected from a data warehouse database of merchant location entries, each candidate merchant location database entry having a DBA name text field designating the doing business as (DBA) name of the merchant location, a street address text field designating the street address of the location of the merchant location, and one or more additional descriptive fields descriptive of one or more predetermined characteristics of the merchant location, each candidate merchant location database entry being selected as a member of the set from and each candidate merchant location database entry having a predetermined minimum textural similarity with the subject merchant location database entry on the basis of respective DBA name text field or street address text field;
• compare the subject merchant location database entry with each of the candidate database entries on the basis of the one or more additional descriptive fields retrieved from the data warehouse database;
• perform a logistic regression using the results of the comparing to calculate a probability that the merchant location corresponding to the subject merchant location database entry and the merchant location corresponding to one or more of the candidate merchant location database entries are the same merchant location;
• and
• output the results of the logistic regression,
• wherein the one or more additional descriptive fields retrieved from the data warehouse include a classification code, the classification code derived from a hierarchical classification, and the comparing determines whether the subject merchant location database entry or the candidate merchant location database entry includes a classification code related to an industry which is experientially known to have merchant location identification data that is either more stable or less stable than other industries, and the logistic regression weights the merchant location classification code with regard to whether the related industry is known to have more or less stable merchant location identification data.