Method and system for comparing commercial entities based on purchase patterns

Patent metadata

Field	Value
Publication number	US-9846885-B1
Application number	US-201414266719-A
Country	US
Kind code	B1
Filing date	Apr 30, 2014
Priority date	Apr 30, 2014
Publication date	Dec 19, 2017
Grant date	Dec 19, 2017

Abstract

A method for comparing purchase patterns includes matching multiple products purchased by a base company to multiple leaf nodes in a taxonomy tree to obtain multiple matching leaf nodes. The taxonomy tree is a hierarchical classification of products. The method further includes assigning, to each of the matching leaf nodes and to each parent node of the matching leaf nodes, a point value to obtain multiple point values, creating, for the base company and by a computer processor, a base feature vector including the point values, and calculating, by the computer processor, a similarity score between the base feature vector of the base company to a test feature vector of a test company. The method further includes providing, in response to the similarity score satisfying a similarity threshold, a recommendation.

First claim

What is claimed is:

1. A method for comparing purchase patterns, comprising:
matching a plurality of products purchased by a base company to a plurality of leaf nodes in a taxonomy tree to obtain a plurality of matching leaf nodes, wherein the taxonomy tree is a hierarchical classification of products;
assigning, to each of the plurality of matching leaf nodes and to each parent node of the plurality of matching leaf nodes, a point value to obtain a plurality of point values;
creating, for the base company and by a computer processor, a base feature vector comprising the plurality of point values;
calculating, by the computer processor, a similarity score between the base feature vector of the base company to a test feature vector of a test company; and
providing, in response to the similarity score satisfying a similarity threshold, a recommendation, wherein the recommendation comprises an advertisement to a customer of the base company.

2. The method of claim 1 , wherein matching the plurality of products comprises matching a plurality of product keywords to a plurality of keywords annotated on the plurality of leaf nodes in the taxonomy tree.

3. The method of claim 2 , wherein the plurality of product keywords are extracted from product information of the plurality of products.

4. The method of claim 1 , further comprising:
extracting product information of a plurality of companies, wherein the plurality of companies comprise the base company and the test company;
identifying the plurality of products for each company of the plurality of companies from keywords in the product information; and
generating the taxonomy tree based on a plurality of identified products.

5. The method of claim 1 , further comprising:
performing a clustering of a plurality of feature vectors from the plurality of companies.

6. The method of claim 5 , wherein the clustering of the plurality of feature vectors from the plurality of companies is based on the plurality of point values in each of the plurality of feature vectors.

7. The method of claim 1 , wherein the taxonomy tree comprises the plurality of parent nodes connected from each leaf node of the plurality of leaf nodes to a root node, wherein each parent node of the plurality of parent nodes represents a class of products and each leaf node of the plurality of leaf nodes represents a product.

8. The method of claim 1 , further comprising:
normalizing the point value assigned to the plurality of parent nodes.

9. The method of claim 1 , wherein the recommendation indicates that the test company is an alternate vendor to the base company.

10. A system for comparing purchasing patterns, comprising:
a computer processor; and
a classification engine, executing on the computer processor, and comprising:
a feature vector generation module configured to:
match a plurality of products purchased by a base company to a plurality of leaf nodes in a taxonomy tree to obtain a plurality of matching leaf nodes, wherein the taxonomy tree is a hierarchical classification of products,
assign to each of the plurality of matching leaf nodes and to each parent node of the plurality of matching leaf nodes, a point value to obtain a plurality of point values, and
create, for the base company, a base feature vector comprising the plurality of point values, and

a similarity score generation module configured to:
calculate a similarity score between the base feature vector of the base company to a test feature vector of a test company, and
provide, in response to the similarity score satisfying a similarity threshold, a recommendation,
wherein the recommendation comprises an advertisement to a customer of the base company, and
wherein the test company is an alternate vendor to the base company.

11. The system of claim 10 , further comprising:
a data repository configured to:
store the taxonomy tree;
store a plurality of feature vectors; and
store transaction records, wherein the transaction records comprise product information for the plurality of products from each company of a plurality of companies, wherein the plurality of companies comprise the base company and test company.

12. The system of claim 10 , further comprising:
a user interface configured to:
receive a request to display the recommendation.

13. The system of claim 10 , further comprising:
a taxonomy tree generation module configured to:
receive, from the data repository the transaction records for the plurality of products from each company of the plurality of companies;
generate the taxonomy tree from a plurality of keywords from the product information in the transaction records of the plurality of products from each company of the plurality of companies; and
store the taxonomy tree in the data repository.

14. The system of claim 10 , further comprising:
a feature vector clustering module configured to:
receive, from the data repository the plurality of feature vectors; and
perform a clustering of the plurality of companies based on the plurality of feature vectors.

15. A non-transitory computer-readable storage medium storing a plurality of instructions for comparing purchase patterns, the plurality of instructions comprising functionality to:
match a plurality of products purchased by a base company to a plurality of leaf nodes in a taxonomy tree to obtain a plurality of matching leaf nodes, wherein the taxonomy tree is a hierarchical classification of products;
assign, to each of the plurality of matching leaf nodes and to each parent node of the plurality of matching leaf nodes, a point value to obtain a plurality of point values;
create, for the base company, a base feature vector comprising the plurality of point values;
calculate a similarity score between the base feature vector of the base company to a test feature vector of a test company; and
provide, in response to the similarity score satisfying a similarity threshold, a recommendation, wherein the recommendation comprises an advertisement to a customer of the base company.

16. The non-transitory computer-readable storage medium of claim 15 , wherein matching the plurality of products comprises matching a plurality of product keywords to a plurality of keywords annotated on the plurality of leaf nodes in the taxonomy tree.

17. The non-transitory computer-readable storage medium of claim 16 , wherein the plurality of product keywords are extracted from product information of the plurality of products.

18. The non-transitory computer-readable storage medium of claim 15 , further comprises functionality to:
extract product information of a plurality of companies, wherein the plurality of companies comprise the base company and the test company;
identify the plurality of products for each company of the plurality of companies from the plurality of keywords in the product information; and
generate the taxonomy tree based on a plurality of identified products.

19. The non-transitory computer-readable storage medium of claim 15 , further comprises functionality to:
perform a clustering of a plurality of feature vectors from the plurality of companies.

20. The non-transitory computer-readable storage medium of claim 19 , wherein the clustering of the plurality of feature vectors from the plurality of companies is based on the plurality of point values in each of the plurality of feature vectors.

21. The non-transitory computer-readable storage medium of claim 15 , wherein the taxonomy tre

Assignees

• Sarshar Nima
• Intuit Inc

Inventors

• Sarshar Nima

Classifications

• G06Q30/0201 Primary

  Market modelling; Market analysis; Collecting market data · CPC title