DOCUMENT RELEVANCE DETERMINING METHOD AND COMPUTER PROGRAM

• US 20110087659A1
• Filed: 12/21/2009
• Published: 04/14/2011
• Est. Priority Date: 10/14/2009
• Status: Active Grant

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: CHANDRA, Praphul, DEY, Prasenjit, MADHVANATH, Sriganesh, RAMANATHAN, Krishnan
• Granted Patent: US 9,223,854 B2
• Time in Patent Office: Days
• US Class Current: 707/728
• CPC Class Codes: G06F 16/3334 Selection or weighting of terms from queries, including natural language queries
• CPC Class Codes: G06F 16/338 Presentation of query results

Abstract

Disclosed is a method and computer program product of determining the relevance of at least a part of an electronic document comprising a plurality of terms distributed over a plurality of regions of said document, comprising displaying the electronic document to a user; determining the gaze characteristics of the person on a region of the electronic document; assigning a relevance score to an individual term in said region based on said characteristics; and generating a term relevance label for said electronic document, said term relevance label comprising relevance scores for the respective individual terms in said document The relevance scores may also be used to define a user profile for the user that can aid in retrieving future documents of relevance to the user.

Claims

1. A method of determining the relevance of at least a part of an electronic document comprising a plurality of terms distributed over a plurality of regions of said document, comprising:

displaying the electronic document to a user;

determining the gaze characteristics of the person on a region of the electronic document;

assigning a relevance score to an individual term in said region based on said characteristics; and

generating a term relevance label for said electronic document, said term relevance label comprising relevance scores for the respective individual terms in said document.

2. The method of claim 1, wherein said characteristics comprise at least one of the duration of the gaze and the gaze frequency.

3. The method of claim 1, wherein said assigning step comprises creating a gaze intensity map for said electronic document, said gaze intensity map comprising the gaze intensities for the respective regions of said document, and wherein assigning the relevance score comprises assigning relevance scores for the individual terms in said regions from said gaze intensities.

4. The method of claim 3, wherein the same individual term is present in different regions, and wherein the step of assigning the relevance score for the individual term comprises combining the relevance scores of the individual term in the different regions.

5. The method of claim 3, wherein the electronic document comprises at least one image, the method further comprising:

segmenting said image into multiple image regions; and

mapping the gaze intensities onto said image regions.

6. The method of claim 3, wherein the electronic documents belongs to a set of electronic documents, the method further comprising assigning a weight to each relevance score based on the frequency of the individual term in the document and the frequency of the electronic documents in the set in which the individual term occurs.

7. The method of claim 1, wherein the determining, assigning and generating steps are repeated for different pages of the electronic document.

8. The method of claim 1, wherein the displaying, determining, assigning and generating steps are repeated for a different user.

9. The method of claim 1, further comprising:

recognizing a gesture and/or facial expression made by the user when gazing at one of said regions, wherein the step of assigning (34) a relevance score to an individual term in said region is based on a combination of the gaze characteristics and the recognized gesture and/or facial expression.

10. The method of claim 1, further comprising extracting a user profile for said user from the generated term relevance label.

12. A method of retrieving at least one electronic document from a database comprising a plurality of electronic documents ranked in accordance with the method of claim 1, the method comprising:

receiving a user-defined search query including at least one individual term;

accessing the electronic database;

accessing the term relevance label of the respective electronic documents to retrieve the respective relevance scores of the at least one individual term in the search query for said documents,

selecting the electronic documents from said electronic database for which the relevance score of the at least one individual term in the search query meets a predefined criterion; and

displaying said selection of electronic documents to the user.

13. The method of claim 12, wherein the database comprises the Internet.

14. A computer program product comprising computer program code for, if executed on a computer, causing the computer to implement the steps of the method of claim 1.

15. A computer system comprising a processor coupled to a gaze detection device and the computer program product of claim 14, said processor being configured to execute said computer program product.

11. A method of updating an electronic document, comprising:

displaying the electronic document to a person;

determining the gaze characteristics of the person on a region of the electronic document;

assigning a relevance score to an individual term in said region based on said characteristics; and

updating at least a part of the electronic document based on the relevance score.