Computer-aided extraction of semantics from keywords to confirm match of buyer offers to seller bids

  • US 8,793,237 B2
  • Filed: 07/11/2012
  • Issued: 07/29/2014
  • Est. Priority Date: 08/31/2004
  • Status: Active Grant
First Claim
Patent Images

1. A computer-implemented method of extracting semantic characteristics from keywords in a computerized search engine the method comprising:

  • providing a search engine interface to a computer-aided search engine for capturing user input or selection of one or more entity text strings;

    providing a database interface to a first database containing a union of known entity category descriptions that may occur as an enumerated list of hierarchical category descriptions, a second database containing buyer profiles, a third database containing entity descriptions, and a fourth database containing relative scoring of entity attributes;

    receiving through the search engine interface a received text string comprising a search input by a user, the received text string including at least one lexical keyword;

    performing a lexical search on the received text string using the search engine to return search results from a search corpus comprising at least one of the first, second, third and fourth database;

    storing in the first database a plurality of category descriptions of items potentially responsive to the search;

    storing in the second database an entity description for each of the potentially responsive items, the entity description comprising two or more item attributes of a respective item, and having a value assigned to each item attribute of the one or more item attributes;

    assigning a category attribute quotient for each value assigned to each item attribute, and storing each category attribute quotient in an array that is indexed against a behavior or characteristic of a class of user regarding the respective item attribute and wherein the category attribute quotient comprises a first number for each value that determines an ordering of category attributes and respective values for display to the user;

    scoring each category description of the plurality of category descriptions by mapping the received text string against each category description and tallying a second number of lexical matches between the received text string and the item attributes of each category description to form a plurality of scored category descriptions, wherein each match increases the score of a corresponding scored category description;

    displaying to the user the scored category descriptions in descending order of score based on the second number of lexical matches and the category attribute quotient and prompting the user to select a first category description of the scored category descriptions wherein the first category description corresponds to a selection by the user corresponding to an item sought by the user;

    receiving a user selection of the first category description to generate a selected category description;

    displaying to the user suggested item attributes based on the category attribute quotient for the selected category description;

    receiving from the user, in response to iteratively displayed suggested item attributes, respective responsive selections of suggested item attributes of the selected category description;

    displaying to the user all user selected attribute values in a final order that is based on a relative order of selection by the user; and

    displaying to the user an organized array of matching items from the search corpus to enable selection by the user of a final selected item.

View all claims

    Thank you for your feedback