Multivariate transaction classification
View Patent ↗Embodiments relate to classification of transactions based upon analysis of multiple variables. For a purchase transaction, such variables can include but are not limited to: buying location, source system, line of business, cost center, functional area, supplier capabilities, item description, account description, organization, department, custom parameters, and others. Embodiments may rely upon one or more classification schemes, such as statistical classification, semantic classification, and/or knowledge base classification, taken alone or in combination. In a purchase transaction, classification based on multivariate analysis facilitates identification of a purchased item or service, and hence accuracy in classifying and assigning a central classification code. Particular embodiments may include a feature allowing user review/revision of category assignments via a feedback loop linked to past classification. This revision feature may add clarity to a current transaction, allow modification of future classification for ongoing improvement, and provide a user-driven measure of system performance.
1. A computer-implemented method comprising:
causing a classification engine to receive unclassified data comprising a first variable and a second variable of a purchase transaction;
causing the classification engine to reference a first ruleset reflecting a statistical classification scheme to generate a first classification based upon the first variable, the second variable, and a rule of the first ruleset; and
causing the classification engine to communicate the first classification to a user,
wherein the first ruleset further comprises a filtered rule selected from at least one of,
a rule generating classifications exceeding a threshold;
a rule generating classifications with a lack a dominant classification choice;
a rule generating a top outcome that is not classified;
a rule generating a top outcome associated with less than a percentage of a total spend;
a rule for which a dimension value is blank.
2. A method as in claim 1 further comprising:
causing the classification engine to reference a second ruleset to generate a second classification based on the first variable, the second variable, and a rule of the second ruleset; and
causing the classification engine to determine that a confidence factor of the second classification is lower than a confidence factor of the first classification.
3. A method as in claim 1 wherein the first variable comprises buying location, source system, line of business, cost center, functional area, supplier capabilities, item description, account description, organization, or department.
4. A method as in claim 1 further comprising revising the first classification based upon feedback from review by the user.
5. A method as in claim 1 wherein the first classification is from a public taxonomy.
6. A non-transitory computer readable storage medium embodying a computer program for performing a method, said method comprising:
causing a classification engine to receive unclassified data comprising a first variable and a second variable of a purchase transaction;
causing the classification engine to reference a first ruleset reflecting a statistical classification scheme to generate a first classification based upon the first variable, the second variable, and a rule of the first ruleset; and
causing the classification engine to communicate the first classification to a user,
wherein the first ruleset further comprises a filtered rule selected from at least one of,
a rule generating classifications exceeding a threshold;
a rule generating classifications with a lack a dominant classification choice;
a rule generating a top outcome that is not classified;
a rule generating a top outcome associated with less than a percentage of a total spend;
a rule for which a dimension value is blank.
7. A non-transitory computer readable storage medium as in claim 6 wherein the method further comprises
causing the classification engine to reference a second ruleset to generate a second classification based on the first variable, the second variable, and a rule of the second ruleset; and
causing the classification engine to determine that a confidence factor of the second classification is lower than a confidence factor of the first classification.
8. A non-transitory computer readable storage medium as in claim 6 wherein the first variable comprises buying location, source system, line of business, cost center, functional area, supplier capabilities, item description, account description, organization, or department.
9. A non-transitory computer readable storage medium as in claim 6 wherein the method further comprises revising the first classification based upon feedback from review by the user.
10. A non-transitory computer readable storage medium as in claim 6 wherein the first classification is from a public taxonomy.
11. A computer system comprising:
one or more processors;
a software program, executable on said computer system, the software program configured to:
cause a classification engine to receive unclassified data comprising a first variable and a second variable of a purchase transaction;
cause the classification engine to reference a first reflecting a statistical classification scheme to generate a first classification based upon the first variable, the second variable, and a rule of the first ruleset; and
cause the classification engine to communicate the first classification to a user,
wherein the first ruleset further comprises a filtered rule selected from at least one of,
a rule generating classifications exceeding a threshold;
a rule generating classifications with a lack a dominant classification choice;
a rule generating a top outcome that is not classified;
a rule generating a top outcome associated with less than a percentage of a total spend;
a rule for which a dimension value is blank.
12. A computer system as in claim 11 wherein the software program is further configured to:
cause the classification engine to reference a second ruleset to generate a second classification based on the first variable, the second variable, and a rule of the second ruleset; and
cause the classification engine to determine that a confidence factor of the second classification is lower than a confidence factor of the first classification.
13. A computer system as in claim 11 wherein the first variable comprises buying location, source system, line of business, cost center, functional area, supplier capabilities, item description, account description, organization, or department.
14. A computer system as in claim 11 wherein the software program is further configured to revise the first classification based upon feedback from review by the user.