Identifying the items most relevant to a current query based on user activity with respect to the results of similar queries

• US 20070239713A1
• Filed: 03/28/2006
• Published: 10/11/2007
• Est. Priority Date: 03/28/2006
• Status: Active Grant

• Current Assignee: A9.com Incorporated (Amazon.com, Inc.)
• Original Assignee: A9.com Incorporated (Amazon.com, Inc.)
• Inventors: Ortega, Ruben, Leblang, Jonathan, Saunders, Colin
• Granted Patent: US 7,996,396 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

A facility for ranking in a result set is described. The facility establishes a system of ratings for combinations of a web page with the query term that reflect the level of relevance of the web page to the query term. In response to each of a number of content interactions performed by users, each with respect to a result corresponding to a web page included in a result set that was generated in response to a query comprising a set of one or more query terms, the facility adjusts a rating score for each combination of the web page and a query term among the set of query terms. For each of a number of web pages included as results in a distinguished result set produced in response to a distinguished query comprising a set of distinguished query terms, the facility (1) selects established and adjusted reading scores for each combination of the web page and a query term among the distinguished set of query terms, and (2) determines a ranking score for the result in the distinguished result set by combining the selected reading scores.

Claims

1. A method in a computing system for ranking web pages in a result set, comprising:

establishing a system of ratings for combinations of a web page with a query term that reflect the level of relevance of the web page to the query term;

in response to each of a plurality of content interactions performed by users, each content interaction performed with respect to a web page included as results in a result set that was generated in response to a query comprising a set of one or more query terms;

adjusting a rating score for each combination of the web page and a query term among the set of query terms; and

for each of a plurality of web pages included as results in a distinguished result set produced in response to a distinguished query comprising a set of distinguished query terms;

selecting established and adjusted rating scores for each combination of the web page and a query term among the distinguished set of query terms, and

determining a ranking score for the web page in the distinguished query result by combining the selected rating scores.

2. The method of claim 1, further comprising causing the web pages included in the distinguished result set to be displayed in accordance with the determined ranking scores.

3. The method of claim 1, further comprising causing the web pages included in the distinguished result set to be displayed in an order corresponding to the determined ranking scores.

4. The method of claim 1, further comprising using the determined ranking scores to select a proper subset of the web pages included in the distinguished query result for display.

5. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a positive direction in response to a content interaction in which a user prints content of the distinguished web page.

6. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a positive direction in response to a content interaction in which a user highlights content of the distinguished web page.

7. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a positive direction in response to a content interaction in which a user copies content of the distinguished web page.

8. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a positive direction in response to a content interaction in which a user pastes content copied from the distinguished web page.

9. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a positive direction in response to a content interaction in which a user bookmarks the distinguished web page.

10. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a positive direction in response to a content interaction in which a user searches within the distinguished web page.

11. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a positive direction in response to a content interaction in which a user posts form content from the distinguished web page.

12. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a positive direction in response to a content interaction in which a user annotates the distinguished web page.

13. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a positive direction in response to a content interaction in which a user follows a link in the distinguished web page to another web page in the Internet domain of the distinguished web page.

14. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a positive direction in response to at least a threshold volume of communications events between client and server while a user is viewing the distinguished web page to another web page in the Internet domain of the distinguished web page.

15. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a negative direction in response to a content interaction in which a user follows a link in the distinguished web page to another web page in the Internet domain of the distinguished web page.

16. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in a negative direction in response to a content interaction in which a user follows a link in the distinguished web page to a web page in an Internet domain other than that of the distinguished web page.

17. The method of claim 1 wherein at least one rating score for a distinguished web page is adjusted in response to a content interaction performed with respect to a web page reached by the user by follows a link in the distinguished web page.

18. A computer-readable medium whose contents cause a computing system to perform a method for scoring the relevance of documents to terms, the method comprising:

initializing a set of ratings for combinations of a document with a term that each reflect the level of relevance of the document to the term;

observing a plurality of actions, each action performed by a user with respect to a document included in a result set generated in response to a query comprising one or more query terms; and

for each observed action, adjusting a rating for each combination of the document with respect to which the action was performed and a term among the terms comprising the query in response to which the result set was generated if the action is of a type selected from the group consisting of;

printing content of the document,

highlighting content of the document,

copying content of the document,

pasting content copied from the document,

bookmarking the document,

searching within the document,

posting form contents from within the document,

securely posting form contents from within the document,

annotating the document,

following a link in the document to another document from the same source,

interacting with the document to produce at least a threshold volume of communication events with a source of the document,

submitting a new query after viewing the document, and

following a link in the document to another document from a source other than the source of the document.

19. The computer-readable medium of claim 18, further comprising, each of a plurality of documents included in a distinguished result set produced in response to a distinguished query comprising a set of distinguished query terms:

selecting ratings for each combination of the document and a query term among the distinguished set of query terms, and

determining a ranking score for the document in the distinguished result set by combining the selected ratings.

20. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type printing content of the document.

21. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type highlighting content of the document.

22. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type copying content of the document.

23. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type pasting content copied from the document.

24. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type bookmarking the document.

25. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type searching within the document.

26. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type securely posting form contents from within the document.

27. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type annotating the document.

28. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type following a link in the document to another document from the same source.

29. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type interacting with the document to produce at least a threshold volume of communication events with a source of the document.

30. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type submitting a new query after viewing the document.

31. The computer-readable medium of claim 18 wherein, for at least one observed action, a rating for each combination of the document with which the action was performed and a term among the terms comprising the query in response to which the result set was generated is adjusted where the action is of the type following a link in the document to another document from a source other than the source of the document.

32. A computer-readable medium providing a ranked search result data structure, comprising, for each of a plurality of content items satisfying a distinguished query comprising query terms:

first information identifying the content item; and

second information indicating a ranking value of the content item within the search result, the indicated ranking value reflecting the extent to which users have previously performed at least one action of a type selected from the group consisting of the following with respect to the content item when included in search results produced for queries comprising at least a portion of the query terms comprising the distinguished query;

printing content of the content item,

highlighting content of the content item,

copying content of the content item,

pasting content copied from the content item,

bookmarking the content item,

searching within the content item,

posting form contents from within the content item,

securely posting form contents from within the content item,

annotating the content item,

following a link in the content item to another content item from the same source,

interacting with the content item to produce at least a threshold volume of communication events with a source of the content item,

submitting a new query after viewing the content item,

following a link in the content item to another content item from a source other than the source of the content item,

forwarding to another user a link to the content item,

responding to an explicit inquiry about the usefulness of the content item,

viewing the content item for at least a threshold amount of time, and

playing at least a threshold percentage of a time-indexed media resource associated with the content item,

such that the content item identified by the data structure may be displayed in a manner that reflects each content item'"'"'s ranking value.

33. The computer-readable medium of claim 32 wherein the data structure further comprises, for each of the plurality of content items satisfying the distinguished query, third information specifying a first address for the content item,

and wherein the ranking value indicated by the first information further reflects the extent to which users have previously performed actions of a type selected from the group with respect to the content item when included in search results produced for queries comprising at least a portion of the query terms comprising the distinguished query and accessed either via the first address or via at least one other address at which the content item is available.

34. The computer-readable medium of claim 32 wherein the ranking value indicated by the first information reflects the extent to which users have previously performed actions of a type selected from the group with respect to the content item either (1) when included in search results produced for queries comprising at least a portion of the query terms comprising the distinguished query, or (2) when included in search results together with at least a portion of the query terms comprising the distinguished query.

35. The computer-readable medium of claim 32 wherein the computer-readable medium is a data transmission medium.

36. The computer-readable medium of claim 32 wherein the computer-readable medium is a data storage medium.