IP Library Granted Patent US 7,877,384
Granted Patent B2
US 7,877,384 · App. 11/681,161 · Granted Jan 25, 2011

Scoring relevance of a document based on image text

View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 7,877,384
App. No.
11/681,161
Granted
Jan 25, 2011
Kind
B2
Abstract

A method and system for determining relevance of a document having text and images to a text string is provided. A scoring system identifies image text associated with an image of the document. The scoring system calculates an image score indicating relevance of the image text to the text string. The image score may be used in many applications, such as searching, summary generation, and document classification, image search, and image classification.

Claims (11)

1. A computing device for ranking pages of a search result for a search request, comprising:

a memory storing computer-executable instructions implementing

a component that receives an indication of pages of the search result, the pages including text and images;

a component that, for each page, determines importance of each image of the page to the page, the importance of an image being determined based on image level features derived from the image itself and page level features derived from relationships between the image and the page that includes the image, the image level features including an indication of whether the image contains a face, the page level features including percent of the page occupied by the image;

a component that, for each page, determines an image score for the page based on comparison of text associated with images of the page with an importance that is above a threshold importance to the search request, the text associated with an image being text of the page that is adjacent to the image of the page;

a component that, for each page, determines a text score for the page based on comparison of text of the page, including text not associated with an image of the page, to the search request;

a component that, for each page, determines a static score for the page, the static score derived from a static structure of documents and links between the documents and indicating importance of a page relative to other pages; and

a component that ranks the pages based on the image score, text score, and static score of the pages; and

a processor for executing the instructions stored in the memory.

2. The computing device of claim 1 including a component that outputs an indication of the ranks of the pages.

3. The computing device of claim 1 wherein the importance of an image to its page is based on an image level feature, a page level feature, and a site level feature, wherein the importance of the image is determined using a support vector machine that is trained to determine the importance of images to pages, and wherein the text associated with an image includes text from a group consisting of anchor text, URL text, ALT text, and surrounding text.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Dec 9, 2014
From: MICROSOFT CORPORATION
To: MICROSOFT TECHNOLOGY LICENSING, LLC
Reel/Frame 034542/0001 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Apr 23, 2007
From: YU, QING; SHI, SHUMING; LI, ZHIWEI; WEN, JI-RONG; MA, WEI-YING
To: MICROSOFT CORPORATION
Reel/Frame 019192/0522 →