Vendor matching engine and method of use

Abstract

A vendor matching engine and method of use, which may be used by a prospective bidder on a government contract or another prospective selectee in order to determine which bids or applications are most worthwhile to pursue. The engine may perform steps of retrieving transactional or vendor registry information, such as Federal Procurement Data System data or System for Award Management data, and may then use this data to generate matchability scores characterizing the compatibility between the selector and selectee. To generate these scores, the distances between selector and selectee coordinates in n-dimensional space may be used in order to determine how similar the prospective selectee is to the selector's desired selectee, allowing a vendor to prioritize government contracts for which they may be best suited. The system may also be used for other pairings of applicants and awarding parties, such as universities and research granting organizations.

1 .- 20 . (canceled) 
     
     
         21 . A method of generating recommendations as to one or more requests for selection in a contested selection process to be pursued by a prospective selectee having one or more competitors, the method comprising:
 receiving, with a processor, an optimality index request;   receiving, with the processor, a selectee matchability score;   receiving, with the processor, a competitor bidding pool, the competitor bidding pool comprising one or more competitor selectees;   retrieving, with the processor, from a database of selectee scores, one or more competitor selectee scores, the one or more competitor selectee scores comprising a selectee matchability score for each of the one or more competitor selectees in the bidding pool;   generating, with the processor, a combined bidding pool, the combined bidding pool comprising a plurality of selectees and a plurality of selectee scores, the plurality of selectees comprising the prospective selectee and the competitor bidding pool, and the plurality of selectee scores comprising the selectee matchability score and the one or more competitor selectee scores;   generating, with the processor, based on the plurality of selectee scores, a probability density function of a score distribution of the combined bidding pool;   calculating, with the processor, for each of the plurality of selectees, a cumulative distribution function, each cumulative distribution function being based on the probability density function;   determining, with the processor, a maximum cumulative distribution function, and determining, for each of the plurality of selectees, an area ratio, wherein the area ratio is determined as a function of the cumulative distribution function of the selectee and the maximum cumulative distribution function;   determining, with the processor, for each of the plurality of selectees, a score percentage, wherein the score percentage is a function of the selectee score of the selectee and a maximum selectee score in the plurality selectee scores;   determining, with the processor, for each of the plurality of selectees, an optimality index, wherein the optimality index is a function of the area ratio and the score percentage; and   providing the optimality index of the prospective selectee and further providing at least one recommendation tailored to the prospective selectee based on the optimality index of the prospective selectee.   
     
     
         22 . The method of  claim 21 , wherein the step of determining the optimality index is calculated using a first equation 
       
         
           
             
               π 
               = 
               
                 
                   ( 
                   
                     100 
                     × 
                     
                       
                         S 
                         v 
                       
                       
                         max 
                          
                         
                             
                         
                          
                         S 
                       
                     
                   
                   ) 
                 
                 × 
                 
                   e 
                   
                     - 
                     ϕ 
                   
                 
               
             
           
         
         wherein π is a raw optimality value, S v  is the selectee score, maxS is the maximum selectee score, e is a natural log base e, and −φ is a negative of the area ratio of the selectee; and using a second equation 
       
       
         
           
             
               Ω 
               = 
               
                 trunc 
                 ( 
                 
                   
                     π 
                     
                       π 
                       max 
                     
                   
                   × 
                   100 
                 
                 ) 
               
             
           
         
         wherein Ω 0  is the optimality index, a trunc( ) function is a function for truncation of a positive real number, and π max  is a postulated optimal-case maximum possible value of the raw optimality value. 
       
     
     
         23 . The method of  claim 21 , wherein the optimality index request is provided by the prospective selectee. 
     
     
         24 . The method of  claim 21 , wherein the optimality index request is provided by a user other than the prospective selectee, and wherein the method further includes a step of designating the prospective selectee as a target of the optimality index request. 
     
     
         25 . The method of  claim 21 , wherein the step of receiving, with the processor, a competitor bidding pool comprises receiving one or more competitor names provided along with the name of a prospective selectee. 
     
     
         26 . The method of  claim 21 , wherein the step of receiving, with the processor, a competitor bidding pool comprises at least one of: generating a list of selectee names eligible to compete with the prospective selectee and receiving a user selection from the list of selectee names, generating a list of selectee names likely to compete with the prospective selectee and receiving a user selection from the list of selectee names, and retrieving a retained bidding pool for the prospective selectee from a memory. 
     
     
         27 . The method of  claim 21 , wherein at least one of the plurality of selectee scores is generated by the steps of:
 retrieving, with the processor, from a selector, selectee information for a selectee associated with the at least one of the plurality of selectee scores, the selectee information comprising at least one of a plurality of transactional variables characterizing one or more transactions of the selectee or a plurality of attribute variables characterizing one or more attributes of the selectee, the selectee information comprising a plurality of initial variables;   selecting a plurality of selected variables from the plurality of initial variables based on the completeness, imputability, and lack of redundancy of the selected variables, and discretizing and standardizing the selected variables;   imputing, with the processor, one or more missing variable values in the plurality of selected variables, and adding the missing variable values to the set of selected variables;   aggregating the selected variables by an ID variable, determining a number of levels of each of the selected variables other than one or more ID variables and dummifying the selected variables other than the one or more ID variables, and transforming the dummified variables into a table of proportions;   calculating a plurality of deviations between a plurality of observed selector values in the table of proportions and a plurality of expected selector values, generating a table of selector inertias, performing generalized singular value decomposition on the table of awarding party inertias in order to obtain selector coordinates in a coordinate space, and generating a table of selector coordinates;   calculating a plurality of deviations between a plurality of observed selectee values in the table of proportions and a plurality of expected selectee values, generating a table of selectee inertias, and performing matrix multiplication of the table of selectee inertias by the table of selector coordinates to yield a table of selectee coordinates;   calculating a plurality of distances between each selector coordinate pair in the table of selector coordinates and each selectee coordinate pair in the table of selectee coordinates;   generating a table of selectee scores based on the distances, the table of selectee scores comprising at least a score of the prospective selectee; and   storing the table of selectee scores in a database of selectee scores.   
     
     
         28 . The method of  claim 21 , wherein the step of generating at least one recommendation tailored to the prospective selectee comprises:
 performing at least one of: a comparison of the prospective selectee to one or more competitors, and a comparison of the prospective selectee to historical data of the prospective selectee;   identifying one or more reasons for a difference in suitability between the prospective selectee and the one or more competitors or the historical data; and   generating the at least one recommendation based on the one or more reasons.   
     
     
         29 . The method of  claim 21 , further comprising:
 automatically preparing a request for selection for the prospective selectee when an optimality index is above a predefined threshold.   
     
     
         30 . The method of  claim 21 , wherein the prospective selectee is a prospective government contract vendor and a selector is a government entity. 
     
     
         31 . A vendor matching system comprising a processor, a memory, a network connection, and a display, the vendor matching system configured to provide recommendations as to one or more requests for selection in a contested selection process to be pursued by a prospective selectee having one or more competitors, the vendor matching system configured to perform the steps of:
 receiving, with the processor, an optimality index request;   receiving, with the processor, a selectee matchability score;   receiving, with the processor, a competitor bidding pool, the competitor bidding pool comprising one or more competitor selectees;   retrieving, with the processor, from a database of selectee scores, one or more competitor selectee scores, the one or more competitor selectee scores comprising a selectee matchability score for each of the one or more competitor selectees in the bidding pool;   generating, with the processor, a combined bidding pool, the combined bidding pool comprising a plurality of selectees and a plurality of selectee scores, the plurality of selectees comprising the prospective selectee and the competitor bidding pool, and the plurality of selectee scores comprising the selectee matchability score and the one or more competitor selectee scores;   generating, with the processor, based on the plurality of selectee scores, a probability density function of a score distribution of the combined bidding pool;   calculating, with the processor, for each of the plurality of selectees, a cumulative distribution function, each cumulative distribution function being based on the probability density function;   determining, with the processor, a maximum cumulative distribution function, and determining, for each of the plurality of selectees, an area ratio, wherein the area ratio is determined as a function of the cumulative distribution function of the selectee and the maximum cumulative distribution function;   determining, with the processor, for each of the plurality of selectees, a score percentage, wherein the score percentage is a function of the selectee score of the selectee and a maximum selectee score in the plurality selectee scores;   determining, with the processor, for each of the plurality of selectees, an optimality index, wherein the optimality index is a function of the area ratio and the score percentage; and   providing the optimality index of the prospective selectee and further providing at least one recommendation tailored to the prospective selectee based on the optimality index of the prospective selectee.   
     
     
         32 . The vendor matching system of  claim 31 , wherein the step of determining the optimality index is calculated using a first equation 
       
         
           
             
               π 
               = 
               
                 
                   ( 
                   
                     100 
                     × 
                     
                       
                         S 
                         v 
                       
                       
                         max 
                          
                         
                             
                         
                          
                         S 
                       
                     
                   
                   ) 
                 
                 × 
                 
                   e 
                   
                     - 
                     ϕ 
                   
                 
               
             
           
         
         wherein π is a raw optimality value, S v  is the selectee score, maxS is the maximum selectee score, e is a natural log base e, and −φ is a negative of the area ratio of the selectee; and using a second equation 
       
       
         
           
             
               Ω 
               = 
               
                 trunc 
                 ( 
                 
                   
                     π 
                     
                       π 
                       max 
                     
                   
                   × 
                   100 
                 
                 ) 
               
             
           
         
         wherein Ω is the optimality index, a trunc( ) function is a function for truncation of a positive real number, and n max  is a postulated optimal-case maximum possible value of the raw optimality value. 
       
     
     
         33 . The vendor matching system of  claim 31 , wherein the optimality index request is provided by the prospective selectee. 
     
     
         34 . The vendor matching system of  claim 31 , wherein the optimality index request is provided by a user other than the prospective selectee, and wherein the method further includes a step of designating the prospective selectee as a target of the optimality index request. 
     
     
         35 . The vendor matching system of  claim 31 , wherein the step of receiving, with the processor, a competitor bidding pool comprises receiving one or more competitor names provided along with the name of a prospective selectee. 
     
     
         36 . The vendor matching system of  claim 31 , wherein the step of receiving, with the processor, a competitor bidding pool comprises at least one of: generating a list of selectee names eligible to compete with the prospective selectee and receiving a user selection from the list of selectee names, generating a list of selectee names likely to compete with the prospective selectee and receiving a user selection from the list of selectee names, and retrieving a retained bidding pool for the prospective selectee from a memory. 
     
     
         37 . The vendor matching system of  claim 31 , wherein at least one of the plurality of selectee scores is generated by the steps of:
 retrieving, with the processor, from a selector, selectee information for a selectee associated with the at least one of the plurality of selectee scores, the selectee information comprising at least one of a plurality of transactional variables characterizing one or more transactions of the selectee or a plurality of attribute variables characterizing one or more attributes of the selectee, the selectee information comprising a plurality of initial variables;   selecting a plurality of selected variables from the plurality of initial variables based on the completeness, imputability, and lack of redundancy of the selected variables, and discretizing and standardizing the selected variables;   imputing, with the processor, one or more missing variable values in the plurality of selected variables, and adding the missing variable values to the set of selected variables;   aggregating the selected variables by an ID variable, determining a number of levels of each of the selected variables other than one or more ID variables and dummifying the selected variables other than the one or more ID variables, and transforming the dummified variables into a table of proportions;   calculating a plurality of deviations between a plurality of observed selector values in the table of proportions and a plurality of expected selector values, generating a table of selector inertias, performing generalized singular value decomposition on the table of awarding party inertias in order to obtain selector coordinates in a coordinate space, and generating a table of selector coordinates;   calculating a plurality of deviations between a plurality of observed selectee values in the table of proportions and a plurality of expected selectee values, generating a table of selectee inertias, and performing matrix multiplication of the table of selectee inertias by the table of selector coordinates to yield a table of selectee coordinates;   calculating a plurality of distances between each selector coordinate pair in the table of selector coordinates and each selectee coordinate pair in the table of selectee coordinates;   generating a table of selectee scores based on the distances, the table of selectee scores comprising at least a score of the prospective selectee; and   storing the table of selectee scores in a database of selectee scores.   
     
     
         38 . The vendor matching system of  claim 31 , wherein the step of generating at least one recommendation tailored to the prospective selectee comprises:
 performing at least one of: a comparison of the prospective selectee to one or more competitors, and a comparison of the prospective selectee to historical data of the prospective selectee;   identifying one or more reasons for a difference in suitability between the prospective selectee and the one or more competitors or the historical data; and   generating the at least one recommendation based on the one or more reasons.   
     
     
         39 . The vendor matching system of  claim 31 , wherein the vendor matching system is further configured to perform the step of:
 automatically preparing a request for selection for the prospective selectee when an optimality index is above a predefined threshold.   
     
     
         40 . The vendor matching system of  claim 31 , wherein the prospective selectee is a prospective government contract vendor and a selector is a government entity.