Systems, apparatus, and methods of programmatically determining unique contacts

Abstract

Systems, apparatus, and methods for determining unique contacts from a collection or pool of merchant data are discussed herein. Some embodiments may provide for an apparatus including circuitry configured to: access first merchant data associated with a first merchant; access second merchant data associated with a second merchant; determine a match score based the first merchant data and the second merchant data indicating a likelihood of the first merchant being the same as the second merchant; determine a match score threshold; determine whether the match score exceeds the match score threshold; and in response determining the match score fails to exceed the match score threshold, determine the first merchant as being different from the second merchant. Some embodiments may provide for techniques for machine learning with merchant data training sets to determine match scores.

1 - 23 . (canceled) 
     
     
         24 . An apparatus for programmatically generating a contact database of unique contacts, the apparatus comprising at least one processor and at least one memory including computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to:
 access first merchant data associated with a first merchant, the first merchant data comprising one or more first merchant attributes, wherein each of the one or more first merchant attributes is associated with a first attribute type;   access second merchant data associated with a second merchant, the second merchant data comprising one or more second merchant attributes, wherein each of the one or more second merchant attributes is associated with a second attribute type;   generate a merchant data training set based on the first merchant data and the second merchant data, the generated merchant data training set having known match and known mismatch patterns;   train a machine learning model using the merchant data training set;   identify one or more attribute pairs, wherein each attribute pair of the identified one or more attribute pairs comprises a first merchant attribute of the one or more first merchant attributes and a second merchant attribute of the one or more second merchant attributes, and wherein the first attribute type of the first merchant attribute corresponds to the second attribute type of the second merchant attribute;   for each attribute pair of the identified one or more attribute pairs, determine a similarity score of the first merchant attribute associated with the attribute pair and the second merchant attribute associated with the attribute pair;   for each similarity score associated with each attribute pair of the identified one or more attribute pairs, determine a predictive score using the trained machine learning model;   generate an adjusted similarity score for each attribute pair of the identified one or more attribute pairs based on the similarity score associated with each attribute pair of the identified one or more attribute pairs and the predictive score associated with each attribute pair of the identified one or more attribute pairs;   determine a match score by combining each of the adjusted similarity scores; and   in an instance in which the match score meets a match score threshold, determine the first merchant as being the same as the second merchant.   
     
     
         25 . The apparatus of  claim 24 , wherein in an instance in which the match score does not meet the match score threshold, the apparatus is further caused to:
 determine the first merchant as being different than the second merchant.   
     
     
         26 . The apparatus of  claim 24 , wherein the apparatus is further caused to:
 determine that the first merchant is a potential lead, wherein the determination of the first merchant as a potential lead is based on one or more of a capability of the first merchant to meet demand at a geographic area, a capability of the first merchant to meet demand for an item, and consumer evaluations of the first merchant.   
     
     
         27 . The apparatus of  claim 24 , wherein each similarity score defines an attribute based variance estimation for the first merchant and the second merchant. 
     
     
         28 . The apparatus of  claim 24 , wherein the first attribute type may comprise one or more of a name attribute, an address attribute, a phone number attribute, or a webpage attribute. 
     
     
         29 . The apparatus of  claim 24 , wherein the apparatus is further caused to:
 programmatically update the contact database of unique contacts, wherein programmatically updating the contact database of unique contacts comprises the apparatus being further caused to:   determine that one or more of the second merchant attributes of the second merchant defines a different attribute as compared to the one or more first merchant attributes of the first merchant; and   store the different attribute to the contact database of unique contacts in associated with the one or more first merchant attributes of the first merchant.   
     
     
         30 . The apparatus of  claim 24 , wherein the machine learning model is a radial basis function support vector machine algorithm or a linear vector support vector machine algorithm. 
     
     
         31 . A method for programmatically generating a contact database of unique contacts, the method comprising:
 accessing first merchant data associated with a first merchant, the first merchant data comprising one or more first merchant attributes, wherein each of the one or more first merchant attributes is associated with a first attribute type;   accessing second merchant data associated with a second merchant, the second merchant data comprising one or more second merchant attributes, wherein each of the one or more second merchant attributes is associated with a second attribute type;   generating a merchant data training set based on the first merchant data and the second merchant data, the generated merchant data training set having known match and known mismatch patterns;   training a machine learning model based on the merchant data training set;   identifying one or more attribute pairs, wherein each attribute pair of the identified one or more attribute pairs comprises a first merchant attribute of the one or more first merchant attributes and a second merchant attribute of the one or more second merchant attributes, and wherein the first attribute type of the first merchant attribute corresponds to the second attribute type of the second merchant attribute;   for each attribute pair of the identified one or more attribute pairs, determining a similarity score of the first merchant attribute associated with the attribute pair and the second merchant attribute associated with the attribute pair;   for each similarity score associated with each attribute pair of the identified one or more attribute pairs, determining a predictive score using the trained machine learning model;   generating an adjusted similarity score for each attribute pair of the identified one or more attribute pairs based on the similarity score associated with each attribute pair of the identified one or more attribute pairs and the predictive score associated with each attribute pair of the identified one or more attribute pairs;   determining a match score by combining each of the adjusted similarity scores; and   in an instance in which the match score meets a match score threshold, determining the first merchant as being the same as the second merchant.   
     
     
         32 . The method of  claim 31 , wherein in an instance in which the match score does not meet the match score threshold, further comprising:
 determining the first merchant as being different than the second merchant.   
     
     
         33 . The method of  claim 31 , further comprising:
 determining that the first merchant is a potential lead, wherein the determination of the first merchant as a potential lead is based on one or more of a capability of the first merchant to meet demand at a geographic area, a capability of the first merchant to meet demand for an item, and consumer evaluations of the first merchant.   
     
     
         34 . The method of  claim 31 , wherein each similarity score defines an attribute based variance estimation for the first merchant and the second merchant. 
     
     
         35 . The method of  claim 31 , wherein the first attribute type may comprise one or more of a name attribute, an address attribute, a phone number attribute, or a webpage attribute. 
     
     
         36 . The method of  claim 31 , further comprising:
 programmatically updating the contact database of unique contacts, wherein programmatically updating the contact database of unique contacts comprises:   determining that one or more of the second merchant attributes of the second merchant defines a different attribute as compared to the one or more first merchant attributes of the first merchant; and   storing the different attribute to the contact database of unique contacts in associated with the one or more first merchant attributes of the first merchant.   
     
     
         37 . The method of  claim 31 , wherein the machine learning model is a radial basis function support vector machine algorithm or a linear vector support vector machine algorithm. 
     
     
         38 . At least one non-transitory computer-readable storage medium for programmatically generating a contact database of unique contacts, the at least one non-transitory computer-readable storage medium having computer-executable program code portions stored therein, the computer-executable program code portions comprising program code instructions configured to:
 access first merchant data associated with a first merchant, the first merchant data comprising one or more first merchant attributes, wherein each of the one or more first merchant attributes is associated with a first attribute type;   access second merchant data associated with a second merchant, the second merchant data comprising one or more second merchant attributes, wherein each of the one or more second merchant attributes is associated with a second attribute type;   generate a merchant data training set based on the first merchant data and the second merchant data, the generated merchant data training set having known match and known mismatch patterns;   train a machine learning model based on the merchant data training set;   identify one or more attribute pairs, wherein each attribute pair of the identified one or more attribute pairs comprises a first merchant attribute of the one or more first merchant attributes and a second merchant attribute of the one or more second merchant attributes, and wherein the first attribute type of the first merchant attribute corresponds to the second attribute type of the second merchant attribute;   for each attribute pair of the identified one or more attribute pairs, determine a similarity score of the first merchant attribute associated with the attribute pair and the second merchant attribute associated with the attribute pair;   for each similarity score associated with each attribute pair of the identified one or more attribute pairs, determine a predictive score using the trained machine learning model;   generate an adjusted similarity score for each attribute pair of the identified one or more attribute pairs based on the similarity score associated with each attribute pair of the identified one or more attribute pairs and the predictive score associated with each attribute pair of the identified one or more attribute pairs;   determine a match score by combining each of the adjusted similarity scores; and   in an instance in which the match score meets a match score threshold, determine the first merchant as being the same as the second merchant.   
     
     
         39 . The computer-readable storage medium of  claim 38 , wherein in an instance in which the match score does not meet the match score threshold, the computer-executable program code portions further comprise program code instructions configured to:
 determine the first merchant as being different than the second merchant.   
     
     
         40 . The computer-readable storage medium of  claim 38 , wherein the computer-executable program code portions further comprise program code instructions configured to:
 determine that the first merchant is a potential lead, wherein the determination of the first merchant as a potential lead is based on one or more of a capability of the first merchant to meet demand at a geographic area, a capability of the first merchant to meet demand for an item, and consumer evaluations of the first merchant.   
     
     
         41 . The computer-readable storage medium of  claim 38 , wherein the first attribute type may comprise one or more of a name attribute, an address attribute, a phone number attribute, or a webpage attribute. 
     
     
         42 . The computer-readable storage medium of  claim 38 , wherein the computer-executable program code portions further comprise program code instructions configured to:
 programmatically update the contact database of unique contacts, wherein programmatically updating the contact database of unique contacts the computer-executable program code portions further comprise program code instructions configured to:   determine that one or more of the second merchant attributes of the second merchant defines a different attribute as compared to the one or more first merchant attributes of the first merchant; and   store the different attribute to the contact database of unique contacts in associated with the one or more first merchant attributes of the first merchant.   
     
     
         43 . The computer-readable storage medium of  claim 38 , wherein the machine learning model is a radial basis function support vector machine algorithm or a linear vector support vector machine algorithm.