Automatically identifying similar purchasing opportunities

• US 8,117,075 B1
• Filed: 08/12/2008
• Issued: 02/14/2012
• Est. Priority Date: 03/30/2000
• Status: Expired due to Fees

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Linden, Greg
• Primary Examiner(s): FADOK, MARK A
• Application Number:
  US12/190,321
• Time in Patent Office: 1,281 Days
• Field of Search: 705/26.1, 705/26, 705/27
• US Class Current: 705/26.1
• CPC Class Codes: G06Q 30/00 Commerce
• CPC Class Codes: G06Q 30/0601 Electronic shopping [e-shopping]
• CPC Class Codes: G06Q 30/0623 Item investigation
• CPC Class Codes: G06Q 30/08 Auctions
• CPC Class Codes: G06Q 40/04 Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange

Abstract

A facility for identifying purchasing opportunities within a set of purchasing opportunities that are similar to a distinguished purchasing opportunity is described. The distinguished purchasing opportunity has descriptive information associated with it. For each of several terms occurring in this descriptive information, the facility generates a term score. Each term score reflects the extent to which the occurrence of the term and the descriptive information associated with the distinguished purchasing opportunity differentiates the distinguished purchasing opportunity from auto purchasing opportunities in the set. The facility then selects as key words the terms occurring in the descriptive information associated with the distinguished purchasing opportunity that have the highest term scores. The facility identifies purchasing opportunities of the set containing the selected key words, and establishes a purchasing opportunity score for each identified purchasing opportunity by summing the term score of the key words occurring in information associated with the identified purchasing opportunities.

Claims

1. A method in a computer system for identifying documents in a set of documents relevant to a distinguished document, comprising the steps of:

determining, in the computer system, an inverse document frequency for each one of a plurality of key words within the distinguished document with respect to the set of documents;

identifying, in the computer system, the key words within the distinguished document having the highest ones of the inverse document frequencies;

for each of the identified key words, conducting, in the computer system, a search for at least a subset of the documents containing at least one of the identified key words;

generating, in the computer system, a score associated with each of the at least a subset of the documents, the score calculated by summing the inverse document frequency of each identified key word contained within each of the at least a subset of the documents;

ranking, in the computer system, the at least a subset of the documents of the set based at least upon the score.

2. The method of claim 1 wherein the ranking is further based at least upon, for each key word, the fraction of occurrences of the key word in the set of documents that occur in the distinguished document.

3. The method of claim 1 wherein the distinguished and ranked documents each describe a purchasing opportunity.

4. The method of claim 1 wherein the distinguished and ranked documents each describe an online auction.

5. The method of claim 1, wherein the step of identifying the key words within the distinguished document having the highest ones of the inverse document frequencies further comprises the step of identifying, in the computer system, the key words having one of the inverse document frequencies greater than a fixed threshold.

6. The method of claim 1, wherein the step of identifying the key words within the distinguished document having the highest ones of the inverse document frequencies further comprises the step of identifying, in the computer system, a fixed number of the key words having the highest ones of the inverse document frequencies.

7. The method of claim 1, wherein the step of identifying the key words within the distinguished document having the highest ones of the inverse document frequencies further comprises the step of identifying, in the computer system, a fixed percentage of the key words having the highest ones of the inverse document frequencies.

8. A system for identifying documents in a set of documents relevant to a distinguished document, comprising:

a server computer system;

logic executed in the server computer system that determines an inverse document frequency for each one of a plurality of key words within the distinguished document with respect to the set of documents;

logic executed in the server computer system that identifies the key words within the distinguished document having the highest ones of the inverse document frequencies;

logic that, for each of the identified key words, conducts a search for documents from the set of documents containing at least one of the identified key words; and

logic that, for each document identified in the search, identifies a relevant document where a sum of the inverse document frequencies of the identified keywords contained in the document exceeds a threshold.

9. The system of claim 8, further comprising logic executed in the server computer system that performs a separate search for each of the identified keywords.

10. The system of claim 8, further comprising logic that ranks the documents identified in the search based at least upon the sum of the inverse document frequencies of the identified key words, wherein the ranking is further based upon, for each key word, the fraction of occurrences of the key word in the set of documents that occur in the distinguished document.

11. The system of claim 8, wherein the distinguished document describe a purchasing opportunity.

12. The system of claim 8, wherein the distinguished document describe an online auction.

13. The system of claim 8, wherein the logic executed in the server computer system that identifies the key words within the distinguished document having the highest ones of the inverse document frequencies further comprises logic that identifies the key words having one of the inverse document frequencies greater than a fixed threshold.

14. The system of claim 8, wherein the logic executed in the server computer system that identifies the key words within the distinguished document having the highest ones of the inverse document frequencies further comprises logic that identifies a fixed number of the key words having the highest ones of the inverse document frequencies.

15. The system of claim 8, wherein the logic executed in the server computer system that identifies the key words within the distinguished document having the highest ones of the inverse document frequencies further comprises logic that identifies a fixed percentage of the key words having the highest ones of the inverse document frequencies.

16. A system for identifying documents in a set of documents relevant to a distinguished document, comprising:

means for determining an inverse document frequency for each one of a plurality of key words within the distinguished document with respect to the set of documents;

means for identifying the key words within the distinguished document having the highest ones of the inverse document frequencies;

means for conducting, for each of the identified key words, a search for at least a subset of the documents containing at least one of the identified key words;

means for generating a score associated with each of the at least a subset of the documents, the score calculated by summing the inverse document frequency of each identified key word contained within each of the at least a subset of the documents;

means for ranking the at least a subset of the documents of the set based at least upon the score.

17. The system of claim 16, wherein the means for identifying the key words within the distinguished document having the highest ones of the inverse document frequencies further comprises means for identifying the key words having one of the inverse document frequencies greater than a fixed threshold.

18. The system of claim 16, wherein the means for identifying the key words within the distinguished document having the highest ones of the inverse document frequencies further comprises means for identifying a fixed number of the key words having the highest ones of the inverse document frequencies.

19. The system of claim 16, wherein the means for identifying the key words within the distinguished document having the highest ones of the inverse document frequencies further comprises means for identifying a fixed percentage of the key words having the highest ones of the inverse document frequencies.