AUTOMATIC, PERSONALIZED ONLINE INFORMATION AND PRODUCT SERVICES

• US 20130073390A1
• Filed: 09/14/2012
• Published: 03/21/2013
• Est. Priority Date: 12/28/1999
• Status: Active Grant

• Current Assignee: Personalized User Model LLP
• Original Assignee: Personalized User Model LLP
• Inventors: Konig, Yochai, Twersky, Roy, Berthold, Michael R.
• Granted Patent: US 8,626,911 B2
• Time in Patent Office: Days
• US Class Current: 705/14.54
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: H04L 67/306 User profiles
• CPC Class Codes: H04L 67/53 using third party service providers
• CPC Class Codes: H04L 67/55 Push-based network services
• CPC Class Codes: H04L 67/75 Indicating network or usage conditions on the user display
• CPC Class Codes: H04N 21/44224 Monitoring of user activity on external systems, e.g. Internet browsing

Abstract

A method for providing automatic, personalized information services to a computer user includes the following steps: transparently monitoring user interactions with data during normal use of the computer; updating user-specific data files including a set of user-related documents; estimating parameters of a learning machine that define a User Model specific to the user, using the user-specific data files; analyzing a document to identify its properties; estimating the probability that the user is interested in the document by applying the document properties to the parameters of the User Model; and providing personalized services based on the estimated probability. Personalized services include personalized searches that return only documents of interest to the user, personalized crawling for maintaining an index of documents of interest to the user; personalized navigation that recommends interesting documents that are hyperlinked to documents currently being viewed; and personalized news, in which a third party server customized its interaction with the user.

Claims

1. A computer-implemented method for providing information to a user, the method comprising:

displaying a web page using a browser program running on a computer;

selecting a plurality of documents related to said web page, said relation established according to an estimated probability that the documents are of interest to the user; and

presenting, via said browser, information regarding at least some of said plurality

of documents to the user.

2. The method of claim 1, further comprising:

prior to presenting the information regarding at least some of said documents,

monitoring user interactions with the web page while the user is engaged in use of said browser program; and

analyzing the monitored interactions to estimate parameters of a userspecific learning machine,

wherein selecting said plurality of documents comprises, for each subject one of the plurality of documents;

identifying properties of the subject document; and

applying the identified properties of the subject document to the userspecific

learning machine to estimate a probability that the subject document is of interest to the user.

3. The method of claim 2, further comprising analyzing the monitored interactions to determine documents not of interest to the user, and wherein estimating parameters of the user-specific learning machine comprises estimating parameters of the user-specific learning machine based at least in part on the documents not of interest to the user.

4. The method of claim 2, wherein monitoring user interactions with the web page comprises monitoring at least one type of interaction selected from the group consisting of:

information about the user'"'"'s interaction with the web page, context information, the user'"'"'s degree of interest in the web page, time spent by the user viewing the web page, whether the user followed at least one link contained in the web page, and a number of links in the web page followed by the user.

5. The method of claim 2, wherein said plurality of documents correspond to a respective plurality of products.

6. The method of claim 2, wherein, for each respective one of the plurality of documents, estimating the probability that the subject document is of interest to the user comprises estimating the probability that the subject document is of interest to the user according to a relevance of the subject document to the web page displayed to the user.

7. The method of claim 2, wherein identifying properties of the subject document comprises identifying properties selected from the group consisting of:

a topic associated with the subject document, at least one product feature extracted from the subject document, an author of the subject document, and an age of the subject document.

8. The method of claim 1, wherein the estimated probability that the documents are of interest to the user is determined according to information regarding the user.

9. The method of claim 8, wherein the information regarding the user is maintained remotely from the computer running the browser program.

10. A computer-implemented method for providing personalized information services to a user, the method comprising:

displaying a web page using a browser program running on a computer;

selecting a plurality of documents related to said web page, said relation based on

said web page and, for each respective one of the plurality of documents, an estimated probability that the respective document is of interest to the user; and

presenting information concerning at least some of the plurality of documents to the user via the browser.

11. The method of claim 10, further comprising:

monitoring user interactions with the web page while the user is engaged in use of

said browser program; and

analyzing the monitored interactions to determine documents of interest to the user; and

estimating parameters of a user-specific learning machine based at least in part on

the documents of interest to the user,

wherein selecting said plurality of documents comprises, for each respective one of the plurality of documents;

identifying properties of the respective document, and

applying the identified properties of the respective document to the user specific learning machine to estimate the probability that the respective document

is of interest to the user.

12. The method of claim 11, further comprising analyzing the monitored interactions to determine documents not of interest to the user, and wherein estimating parameters of the user-specific learning machine further comprises estimating parameters of the user-specific learning machine based at least in part on the documents not of interest to the user.

13. The method of claim 11, wherein monitoring user interactions comprises monitoring at least one type of interaction selected from the group consisting of:

whether the user viewed the web page, information about the user'"'"'s interaction with the web page, context information, the user'"'"'s degree of interest in the web page, time spent by the user viewing the web page, whether the user followed at least one link contained in the web page, and a number of links in the web page followed by the user.

14. The method of claim 11, wherein said plurality of documents correspond to a respective plurality of products.

15. The method of claim 11, wherein presenting information concerning at least some of the plurality of documents comprises, for each respective document, using the estimated probability for that the respective document is of interest to the user and the relevance of the respective document to the web page displayed to the user to present at least a portion of the respective document to the user.

16. The method of claim 11, wherein identifying properties of the respective document comprises identifying properties selected from the properties consisting of:

a topic associated with the respective document, at least one product feature extracted from the respective document, an author of the respective document, and an age of the respective document.

17. A computer-implemented method for providing personalized advertisements to a user, the method comprising:

selecting a plurality of advertisements for display to a user of a computer system,

said advertisements selected according to an estimated probability that the advertisements are of interest to the user;

providing, from a server to the computer, one or more of the selected advertisements; and

presenting, via a browser running on said computer, the one or more selected advertisements.

18. The method of claim 17, wherein the one or more advertisements are presented via the browser in conjunction with a web page.

19. The method of claim 18, further comprising:

prior to presenting the one or more advertisements,

monitoring user interactions with the web page while the user is engaged in use of said browser; and

analyzing the monitored interactions to estimate parameters of a userspecific learning machine,

wherein selecting said plurality of advertisements comprises, for each subject one

of the plurality of advertisements;

identifying properties of the subject advertisement; and

applying the identified properties of the subject advertisement to the users specific

learning machine to estimate a probability that the subject advertisement is of interest to the user.

20. The method of claim 19, further comprising analyzing the monitored interactions to determine advertisements not of interest to the user, and wherein estimating parameters of the user-specific learning machine comprises estimating parameters of the user-specific learning machine based at least in part on the advertisements not of interest to the user.

21. The method of claim 19, wherein monitoring user interactions with the web page comprises monitoring at least one type of interaction selected from the group consisting of:

information about the user'"'"'s interaction with the web page, context information, the user'"'"'s degree of interest in the web page, time spent by the user viewing the web page, whether the user followed at least one link contained in the web page, and a number of links in the web page followed by the user.

22. The method of claim 17, wherein, for each subject one of the plurality of advertisements, estimating the probability that the subject advertisement is of interest to the user comprises estimating the probability that the subject advertisement is of interest to the user according to a relevance of the subject advertisement to a web page displayed to the user.

23. The method of claim 19, wherein identifying properties of the subject advertisement comprises identifying properties selected from the group consisting of:

a topic associated with the subject advertisement, at least one product feature extracted from the subject advertisement, an author of the subject advertisement, and an age of the subject advertisement.

24. The method of claim 17, wherein the estimated probability that the advertisements are of interest to the user is determined according to information regarding the user.

25. The method of claim 24, wherein the information regarding the user is maintained remotely from the computer running the browser.

26. A computer-implemented method for providing personalized

advertisements to a user, the method comprising:

monitoring user interactions with search results while the user is engaged in use of

a browser program running on a computer;

receiving, at a server communicatively coupled to the computer via one or more networks, a search query from the user;

retrieving a plurality of advertisements based on the search query;

for each retrieved advertisement of said plurality of retrieved advertisements;

identifying properties of the retrieved advertisement and, based thereon,

estimating a probability that the retrieved advertisement is of interest to the user;

and

using the estimated probabilities for the respective plurality of retrieved advertisements to present at least some of the retrieved advertisements to the user.

27. The method of claim 26, further comprising presenting to said user a list of said retrieved advertisements ranked according to respective estimated probabilities of the plurality of retrieved advertisements being of interest to the user.

28. The method of claim 26, wherein estimating the probability that the retrieved advertisements are of interest to the user comprises estimating parameters of a user-specific learning machine based, at least in part, on previously retrieved advertisements not of interest to the user.

29. The method of claim 26, wherein monitoring user interactions with search results comprises one or more of:

monitoring whether the user viewed the search results, information about the user'"'"'s interaction with the search results, context information, and time spent by the user viewing individual ones of the search results.

30. The method of claim 26, wherein for each retrieved advertisement, the estimated probability that the retrieved advertisement is of interest to the user is determined according to information regarding the user.

31. The method of claim 30, wherein the information regarding the user is maintained remotely from the computer running the browser program.

32. A computer-implemented method, comprising:

responsive to a present search query submitted by a user via a client-based application, returning personalized advertisements for the user from a server-based application to the

client-based application, wherein the personalized advertisements are selected for return to the client-based application according to information regarding the user'"'"'s interests.

33. The method of claim 32, wherein the personalized advertisements are ranked in order of estimated user interest.

34. The method of claim 33, wherein the interest of the user is estimated by applying identified properties of each respective one of the personalized advertisements to a user-specific learning machine to determine a probability of user interest for each respective personalized advertisement.

35. The method of claim 33, further comprising displaying, by the client-based application, the personalized advertisements in ranked order.

36. The method of claim 33, wherein the personalized advertisements comprise one or more of:

text, images, and non-text media.

37. The method of claim 33, wherein the information regarding the user'"'"'s interests is determined, in part, based on monitored user interactions with prior advertisements.

38. The method of claim 33, wherein the client-based application comprises a Web browser application.

39. The method of claim 38, wherein the personalized advertisements are determined according to information regarding monitored user interactions with prior search results from prior search queries.

40. The method of claim 33, wherein the personalized advertisements are selected according to monitored user interactions with prior search results from prior search queries.

41. The method of claim 33, wherein the personalized advertisements comprise advertisements estimated to be of likely interest to the user in the context of the present search query.

42. The method of claim 33, wherein the personalized advertisements are ranked for presentation to the user.

43. The method of claim 33, wherein the personalized advertisements are presented according to an estimated user interest therein.

44. The method of claim 43, wherein the estimated user interest is based, at least in part, on the present search query.

45. The method of claim 44, wherein the estimated user interest is further based, at least in part, on an assessment of potential advertisements in the context of the present search query.

46. A computer-implemented method for providing personalized

advertisements to a user, the method comprising:

monitoring user interactions with search results while the user is engaged in use of

a browser program running on a computer;

receiving, at a server communicatively coupled to the computer via one or more networks, a search query from the user;

retrieving a plurality of advertisements based on the search query and, for each subject advertisement, an estimate of a probability that the subject advertisement is of interest to the user; and

using the estimated probabilities for the subject retrieved advertisements to present at least some of the retrieved advertisements to the user.