×

Systems and methods of natural language processing and statistical analysis to identify matching categories

  • US 9,691,085 B2
  • Filed: 04/20/2016
  • Issued: 06/27/2017
  • Est. Priority Date: 04/30/2015
  • Status: Active Grant
First Claim
Patent Images

1. A method, comprising:

  • receiving, in a computing apparatus, a plurality of sets of itemized data, wherein each respective set of the itemized data includes a plurality of data fields containing an item description of a product purchased by a customer from a merchant;

    performing, by the computing apparatus, natural language processing on item descriptions in the itemized data;

    classifying, by the computing apparatus, items described by the item descriptions into item tiers based on the natural language processing, including;

    determining presence of predetermined keywords in the item descriptions, andscoring the item tiers of the items based on predetermined weights of the predetermined keywords; and

    generating, by the computing apparatus, profiles of merchants and profiles of customers based at least in part on the item tiers classified from the natural language processing of the itemized data, wherein the merchants and the customers are identified in the itemized data.

View all claims
  • 1 Assignment
Timeline View
Assignment View
    ×
    ×