System and methods for constructing personalized context-sensitive portal pages or views by analyzing patterns of users' information access activities

• US 20050278323A1
• Filed: 07/27/2005
• Published: 12/15/2005
• Est. Priority Date: 04/04/2002
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Anderson, Corin Ross, Horvitz, Eric
• Granted Patent: US 7,904,439 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/954 Navigation, e.g. using categorised browsing
• CPC Class Codes: G06Q 30/0256 User search
• CPC Class Codes: Y10S 707/99948 Application of database or data structure, e.g. distributed, multimedia, or image

Abstract

The present invention relates to a system and methodology to assist users with data access activities and that includes such activities as routine web browsing and/or data access applications. A coalesced display or montage of aggregated information is provided that is focused from a plurality of sources to achieve substantially one-button access to user'"'"'s desired web or data source information/destinations in order to mitigate efforts in retrieving and viewing such information. Past web or other type data access patterns can be mined to predict future browsing sites or desired access locations. A system is provided that builds personalized web portals for associated users based on models mined from past data access patterns. The portals can provide links to web resources as well as embed content from distal (remote) pages or sites producing a montage of web or other type data content. Automated topic classification is employed to create multiple topic-centric views that can be invoked by a user.

Claims

1. A display building system, comprising:

a predictive component to determine potential topics of interest; and

a formatting component to display at least one of the potential topics of interest based at least in part on an expected utility and a current context.

2. The system of claim 1, the formatting component approximates a value of a page p to a user as a function of computed quantities defined as an interest, I(p), and a navigation savings derived by placing content on a montage, S(p).

3. The system of claim 2, the expected utility of a page is evaluated as a function, ƒ

(I(p), S(p)).

4. The system of claim 3, further comprising at least one of a weighted additive utility model and a weighted multiplicative utility model.

5. The system of claim 1, the utility models formulated as at least one of:

ƒ

(I(p),S(p))=I(p)^k1*S(p)^k2;


ƒ

(I(p),S(p))=I(p)^k1+S(p)^k2; and


ƒ

(I(p),S(p))=k1*I(p)+k2*S(p).

wherein k1 and k2 are constants.

6. The system of claim 5, further comprising at least one control to adjust at least one of the utility model and the constants.

7. The system of claim 6, the at least one control adjusts various parameters associated with the utility model and the constants, the parameters adjusted via graphical sliders that are labeled with descriptions related to an influence of adjusting the at least one control in different directions.

8. The system of claim 3, the expected utility of a page is expressed as a probability that the user will visit a page p given a current context C, Pr(p|C).

9. The system of claim 8, the expected utility of a page is expressed as:

E[U(p)]=Pr(p|C)*ƒ

(I(p),S(p)).

wherein E[U(p)] is an expected utility of a page.

10. The system of claim 9, the expected utility of the page is expressed as at least one of:

E[U(p)]=Pr(p|C)*(I(p)^k1*S(p)^k2;


E[U(p)]=Pr(p|C)*(I(p)^k1+S(p)^k2; and


E[U(p)]=Pr(p|C)*(k1*I(p)+k2*S(p));

wherein k1 and k2 are constants.

11. The system of claim 9, an expected utility in a topic is expressed as:

E[U(T)]=Pr(T|C)I(T)

wherein E[U(T)] is an expected utility of a topic T equated to a product of a probability that the user will view any page with topic T in a current context C, stated as Pr(T|C ), and I(T) is an interest in the topic.

12. The system of claim 1, the formatting component formats the display by maximizing a total expected utility, subject to the sizes of a display and in relation to the size and amount of embedded content placed in a montage.

13. The system of claim 12, further comprising at least one of a knapsack algorithm, box fitting algorithm and a geometric analysis to fit a plurality of display clippings into the montage.

14. A computer-readable medium having computer-executable instructions stored thereon to determine potential topics of interest and to display potential topics of interest of claim 1.

15. A method for displaying a montage, comprising:

employing an automated topic classifier to identify at least one topic of a display page; and

utilizing the at least one topic and statistics relating to topic access as part of a definition of a context to display the page.

16. The method of claim 15, further comprising:

utilizing inferred topics in the construction of a model; and

employing real-time inferences of the model at run-time.

17. The method of claim 15, further comprising:

determining a probabilistic training procedure comprising at least one of Support Vector Machines (SVMs), naï

ve-Bayes models, Bayesian networks, regression models, and text-similarity.

18. The method of claim 17, further comprising at least one of:

tagging a database of representative pages;

assigning the pages to different categories in a topic ontology; and

analyzing the pages to assign respective pages a likelihood of being in an associated category represented in the topic ontology.