Identifying an item based on data inferred from information about the item

• US 8,244,666 B2
• Filed: 06/14/2010
• Issued: 08/14/2012
• Est. Priority Date: 02/09/2006
• Status: Active Grant

• Current Assignee: eBay Inc.
• Original Assignee: eBay Inc.
• Inventors: Loftus, Josh, Sundaranatha, Venkat, Monier, Louis Marcel Gino, Leon, Jean-Michel, Arora, Aditya, Foster, Benjamin David, Shah, Seema
• Primary Examiner(s): Le, Uyen T.
• Application Number:
  US12/815,371
• Publication Number:
  US 20100250535A1
• Time in Patent Office: 792 Days
• Field of Search: None
• US Class Current: 707/602
• CPC Class Codes: G06F 16/35 Clustering; Classification
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation

Abstract

There are provided methods and systems to identify an item based on data inferred from information about the item. Information is received that is descriptive of an item capable of being listed on an information storage and retrieval platform. The information is associated with an aspect-value pair, which is an aspect of the item and a value of the aspect and is inferred from the information. An expression of an interest of a user in an additional item is received. The expression is associated with the aspect-value pair. In response to the associating of the expression with the aspect-value pair, the item is identified as the additional item.

Claims

1. A method of identifying an item listed on an information storage and retrieval platform as a candidate for an item sought, the method comprising:

receiving information describing the item listed on the information storage and retrieval platform;

inferring, using a processor, an aspect-value pair from the information, the aspect-value pair being an aspect of the item listed on the information storage and retrieval platform and a value of the aspect of the item listed on the information storage and retrieval platform;

receiving a search query;

inferring an additional aspect-value pair from the search query, the additional aspect-value pair being an aspect of the item sought and a value of the aspect of the item sought;

determining a correspondence between the aspect-value pair and the additional aspect value pair; and

identifying the item listed on the information storage and retrieval platform as the candidate for the item sought based on the determining of the correspondence between the aspect-value pair and the additional aspect-value pair.

2. The method of claim 1, wherein the determining of the correspondence between the aspect-value pair and the additional aspect-value pair is based on a rule.

3. The method of claim 1, wherein the determining of the correspondence between the aspect-value pair and the additional aspect-value pair is based on a rule.

4. The method of claim 1, wherein the inferring of the aspect-value pair from the information is based on a rule.

5. The method of claim 4, wherein a condition clause of the rule includes an abbreviation of a word and a predicate clause of the rule specifies the aspect-value pair based on the abbreviation of the word.

6. The method of claim 5, wherein the abbreviation of the word is an abbreviation of a color.

7. The method of claim 4, further comprising:

determining the aspect of the item listed on the information storage and retrieval platform;

generating a candidate value corresponding to the value of the aspect of the item listed on the information storage and retrieval platform, the generating based on a search of a database;

receiving an acceptance of the candidate value; and

generating the rule in response to the acceptance of the candidate value, the rule including the aspect-value pair in a predicate clause.

8. A system of identifying an item listed on an information storage and retrieval platform as a candidate for an item sought, the system comprising:

at least one processor;

a listing module configured to receive information describing the item listed on the information storage and retrieval platform;

a classification service engine implemented by the at least one processor and configured to infer an aspect-value pair, the aspect-value pair being an aspect of the item listed on the information storage and retrieval platform and a value of the aspect of the item listed on the information storage and retrieval platform;

a communication module configured to receive a search query;

an aspect extractor module configured to infer an additional aspect-value pair, the additional aspect-value pair being an aspect of the item sought and a value of the aspect of the item sought; and

a search index engine configured to;

determine a correspondence between the aspect-value pair and the additional aspect-value pair; and

identify the item listed on the information storage and retrieval platform as the candidate of the item sought based on the correspondence between the aspect-value pair and the additional aspect-value pair.

9. The system of claim 8, wherein the inferring of the aspect-value pair is based on a rule.

10. The system of claim 9, wherein a condition clause of the rule includes an abbreviation of a word and a predicate clause of the rule specifies the aspect-value pair based on the abbreviation of the word.

11. The system of claim 10, wherein the abbreviation of the word is an abbreviation of a color.

12. The system of claim 9, further comprising:

a scrubber module to determine the aspect of the item listed on the information storage and retrieval platform;

a string analyzer module to generate a candidate value corresponding to the value of the aspect of the item listed on the information storage and retrieval platform, the generating based on a search of a database; and

an authoring module to;

receive an acceptance of the candidate value; and

generate the rule in response to the acceptance of the candidate value, the rule including the aspect-value pair in a predicate clause.

13. A non-transitory machine readable medium storing a set of instructions that, when executed by a machine, cause the machine to perform a method of identifying an item listed on an information storage and retrieval platform as a candidate for an item sought, the method, the method comprising:

receiving information describing the item listed on the information storage and retrieval platform;

inferring an aspect-value pair from the information, the aspect-value pair being an aspect of the item listed on the information storage and retrieval platform and a value of the aspect of the item listed on the information storage and retrieval platform;

receiving a search query;

inferring an additional aspect-value pair from the search query, the additional aspect-value pair being an aspect of the item sought and a value of the aspect of the item sought; and

identifying the item listed on the information storage and retrieval platform as the candidate for the item sought based on a determining of a correspondence between the aspect-value pair and the additional aspect-value pair.

14. The non-transitory machine readable medium of claim 13, wherein the aspect-value pair is inferred from the information based on a rule.

15. The non-transitory machine readable medium of claim 14, wherein a condition clause of the rule includes an abbreviation of a word and a predicate clause of the rule specifies the aspect-value pair based on the abbreviation of the word.

16. The non-transitory machine readable medium of claim 15, wherein the abbreviation of the word is an abbreviation of a color.

17. The non-transitory machine readable medium of claim 14, further comprising:

determining the aspect of the item listed on the information storage and retrieval platform;

generating a candidate value corresponding to the value of the aspect of the item listed on the information storage and retrieval platform, the generating based on a search of a database;

receiving an acceptance of the candidate value; and

generating the rule in response to the acceptance of the candidate value, the rule including the aspect-value pair in a predicate clause.