A mechanism for distinguishing user profiles in web content for the visually challenged (17-Aug-2009)

Options

IP.com Prior Art Database Disclosure (Source: IPCOM)

Disclosure Number IPCOM000186369D dated 17-Aug-2009

Originally published in Prior Art Database

Disclosed by: IBM

Country: Undisclosed

Disclosure File: 6 pages / 117.5 KB / English (United States)

Authors: Ravi S Gorawade, Ananth Chakravarthy, Indiver N Dwivedi Background: Social Networking is the latest trend in the society. Applications like orkut, facebook, linked-in etc have taken it to a different level all together. People can make friends all over the world, which may have similar interests or belonging to any particular institute or anything. They can also convey their thought on various topics through blogs, communities and so on; and interact with people all over the globe. Problem Statement: Major aspects of the social networking sites are very difficult for a visually impaired person to actually use. It becomes very difficult for him/her to actually distinguish between 2 persons with same names by just listening to what the browser has just read. Also associating contents or correlating the content with the actual user or getting some extra information for the particular user while listening to the context become difficult. Normal users usually hover over a particular link which is usually embedded in the web content itself to get some more information about a particular topic, which is practically impossible in case of the visually impaired user. For example, consider a social networking site like orkut, there is a possibility that multiple users share same names, for a visually impaired person it becomes very difficult to distinguish between them as the browser just reads out the contents. This problem can be extended to all web content in general. Known Solutions: IBM HTTP Reader is a tool that assists the visually impaired person to actually read the contents of a web page. This tool only reads out the content, it does not distinguish it. If 2 users with duplicate names are present on the page there is no way to distinguish these users for the visually impaired.

Options

This text was extracted from a PDF file.

At least one non-text object (such as an image or picture) has been suppressed.

This is the abbreviated version, containing approximately 37% of the total text.

Page 1 of 6

A mechanism for distinguishing user profiles in web content for the visually challenged

Claims:

1. The mechanism for a visually challenged user to automatically obtain a context based content vocally without modifying the original visual content.

2. Mechanism of pushing the entire content of the current page to the context resolving service and getting a new tagged content without the content developer having to code anything specific.
a. Context resolution is based on the current access time of the page. For example, the developer will not code for the definition look up of "Sam" inside the content. But it may happen that the context resolution service might getting the additional tagged content as a latest news item related to the word "Sam" associated with the technical domain.

3. The domain of application of this disclosure is the screen reading software. For example, the disclosure could be blocked for the screen reading software like "IBM HTTP Reader" software or any other Operating system.

4. The mechanism of using the social data of the user, which he/she registers with the context resolution service to fetch the most appropriate content to be read out, based on the current content.

The implementation for the disclosure can be visualized from the two following flow charts. It may be noted that the implementation is spread across two components.

The server side which is responsible for generating the social profile of the names of people present in the web content.

The client side which is the browser side which is responsible for reading out the profile of the user.

Page 2 of 6

Flowchart for Server

ART

Yes

Is User Login Re

ire

Fetch Social/LDAP Profile for Logged in

Use text analysis for building context

Identify proper names in web content

Fetch proper names

Fetch commenter's Social/LDAP Profile

Resolve duplicates usin

context

Is Duplicate name found in commented contents?

Fetch Social/LDAP Profile for user basing on the text anal

tics

Yes

Add profile content to web content

Build nearest profile using commenter's profile

Fetch Social/LDAP Profile of the matched user

STOP

Yes

More names present?

Page 3 of 6

Flowchart for Browser

START

Is Hotkey pressed?

Yes