System and method for personalized snippet generation

• US 8,631,006 B1
• Filed: 04/14/2005
• Issued: 01/14/2014
• Est. Priority Date: 04/14/2005
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Haveliwala, Taher H., Kamvar, Sepandar D.
• Primary Examiner(s): Mofiz, Apu
• Assistant Examiner(s): BIBBEE, JARED M
• Application Number:
  US11/107,490
• Time in Patent Office: 3,197 Days
• Field of Search: 707/3, 707/5, 707/722, 707/723, 707/732
• US Class Current: 707/732
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/334 Query execution G06F16/335 takes precedence
• CPC Class Codes: G06F 16/337 Profile generation, learning or modification
• CPC Class Codes: G06F 16/345 Summarisation for human users
• CPC Class Codes: G06F 16/93 Document management systems
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation

Abstract

Snippets of text provided are generated based in part on a user'"'"'s profile. An item, such as a document, is examined to identify terms related to the user'"'"'s profile. A term profile for an identified term is compared to a user'"'"'s profile. The more closely related the identified term is to the user'"'"'s profile, the higher a similarity score will be. Alternatively, terms found in a document may have a user profile score which may be obtained by looking the term up in the user'"'"'s profile. Terms having high profile similarity scores or high user profile scores are used in identifying snippets which may be relevant to a user. The high scoring terms may be added to search terms and provided to a snippet generator.

Claims

1. A computer-implemented method of producing search results, comprising:

at a computerized search engine system distinct from a client system;

receiving a search request associated with a user from the client system, the search request having one or more search terms;

obtaining search results for the search request;

generating a personalized snippet for at least one of the search results in accordance with profile information associated with the user, the snippet comprising a text portion of the search result chosen based on at least one or more search terms and one or more terms of the profile information; and

transmitting the search results and personalized snippet to the client system for display;

wherein the generating comprises;

identifying content associated with one of the search results;

determining a profile similarity score for at least one term in the content; and

generating a snippet based in part on the at least one term if the profile similarity score is above a threshold; and

wherein determining the profile similarity score includes;

identifying a respective term profile associated with the at least one term; and

determining a similarity between the profile information associated with the user and the respective term profile.

2. The method of claim 1, wherein each of the profile information associated with the user and the respective term profile are represented as a vector of a plurality of profile categories and respective weights, and wherein determining the similarity between the profile information associated with the user and the respective term profile comprises computing a distance between the vector corresponding to the profile information associated with the user and the vector corresponding to the respective term profile.

3. A computer-implemented method of providing snippets, comprising:

at a server system distinct from a client system;

identifying at least one item having a content;

generating a personalized snippet for the item in accordance with information associated with a user, the snippet comprising a portion of the content chosen based at least in part on one or more items of the information associated with the user; and

transmitting the personalized snippet to the client system for display;

wherein the generating comprises;

determining an information similarity score for at least one term in the content; and

generating a snippet based in part on the term if the information similarity score is above a threshold; and

wherein determining the information similarity score includes;

identifying a respective term profile associated with the term; and

determining a similarity between the information associated with the user and the respective term profile.

4. The method of claim 3, wherein the snippet is further based on one or more query terms of a search query.

5. The method of claim 3, wherein each of the information associated with the user and the respective term profile are represented as a plurality of profile categories and respective weights.

6. The method of claim 3, wherein the item is identified in response to a search request.

7. The method of 3, wherein the content includes information about the item.

8. A non-transitory computer readable storage medium storing one or more programs configured to be executed by a computerized search engine system distinct from a client system, the one or more programs comprising:

instructions for receiving a search request associated with a user from the client system, the search request having one or more search terms;

instructions for obtaining search results for the search request;

instructions for generating a personalized snippet for at least one of the search results in accordance with profile information associated with the user, the snippet comprising a text portion of the search result chosen based on at least one or more search terms and one or more terms of the profile information; and

instructions for transmitting the search results and personalized snippet to the client system for display;

wherein the instructions for generating including instructions for;

identifying content associated with one of the search results;

determining a profile similarity score for at least one term in the content; and

generating a snippet based in part on the at least one term if the profile similarity score is above a threshold; and

wherein the instructions for determining the information similarity score include instructions for;

identifying a respective term profile associated with the term; and

determining a similarity between the information associated with the user and the respective term profile.

12. The non-transitory computer readable storage medium of claim 8, wherein each of the information associated with the user and the respective term profile are represented as a plurality of profile categories and respective weights.

13. The non-transitory computer readable storage medium of claim 8, wherein each of the profile information associated with the user and the respective term profile are represented as a vector of a plurality of profile categories and respective weights, and wherein determining the similarity between the profile information associated with the user and the respective term profile comprises computing a distance between the vector corresponding to the profile information associated with the user and the vector corresponding to the respective term profile.

9. A system for producing search results, comprising:

memory;

one or more processors; and

at least one program, stored in the memory and executed by the one or more processors, the at least one program including;

instructions for receiving, from a distinct client system, a search request associated with a user, the search request having one or more search terms;

instructions for obtaining search results for the search request;

instructions for generating a personalized snippet for at least one of the search results in accordance with profile information associated with the user, the snippet comprising a text portion of the search result chosen based on at least one or more search terms and one or more terms of the profile information; and

instructions for transmitting the search results and personalized snippet to the client system for display;

wherein the instructions for generating including instructions for;

identifying content associated with one of the search results;

determining a profile similarity score for at least one term in the content; and

generating a snippet based in part on the at least one term if the profile similarity score is above a threshold; and

wherein the instructions for determining the information similarity score include instructions for;

identifying a respective term profile associated with the term; and

determining a similarity between the information associated with the user and the respective term profile.

10. The system of claim 9, wherein each of the information associated with the user and the respective term profile are represented as a plurality of profile categories and respective weights.

11. The system of claim 9, wherein each of the profile information associated with the user and the respective term profile are represented as a vector of a plurality of profile categories and respective weights, and wherein determining the similarity between the profile information associated with the user and the respective term profile comprises computing a distance between the vector corresponding to the profile information associated with the user and the vector corresponding to the respective term profile.