NMSU branding

New Mexico State University

New Mexico State University

News Center




NSF grant will help NMSU researchers develop assistive technologies

Researchers at New Mexico State University are working to help visually impaired users overcome some of the obstacles in navigating World Wide Web sites in the classroom. With help from a 3-year, $600,000 National Science Foundation grant, they will develop assistive technologies to help audio browsers recognize and read specific pieces of Web pages.


Visually impaired users often navigate the Web with audio browsers that read the text out loud, but the software doesn't recognize or know how to read forms, frames and tables.

"These aspects of Web pages are three-dimensional in nature," said Art Karshmer, head of NMSU's computer science department and the grant's project director. "The current tools for reading Web text to visually impaired users are not capable of conveying the complex information on many Web pages."

So Karshmer, Gopal Gupta and Enrico Pontelli from computer science and Doug Gillan from psychology, have set out to build two new tools. One will enable these browsers to better understand the structure of the pages. The other will present the information on the pages, including the tables, to the visually impaired user in a comprehensible format.

The goal is to enable teachers to use these tools in classrooms. Instructors may have the control to make their Web pages accessible, but if they want to use pages that include tables written by others, their visually impaired students need the ability to read them, Karshmer said.

When students use audio browsers, the information is presented mostly in a linear or sequential fashion, Gillan pointed out. His challenge as a cognitive psychologist is to determine how to read information from a table so that it makes sense to someone who can't see it.

"We're trying to figure out how to translate from one perceptual format to another," he said. "We'll be asking what kind of information they need and why, and how do you present it?"

To answer those questions, Gillan will give sighted users the same information disadvantages as visually impaired users. For example, they may see the table's information as a string of text. Then he will begin adding components to see what helps.

"They may be able to search in different directions, get an overview of the columns or rows, or mark a particular cell that is relevant to return to it easily," Gillan said.

Once they have determined which devices provide the most help, developers will implement them with tools Pontelli and Gupta already have developed for reading the structure of Web pages and for filtering information between different formats.

"We're providing the technological part of the project," Pontelli said. "People read tables differently for different purposes. We want to classify the different approaches and design schemes that will allow users to quickly and easily navigate them."

He expects to use quite a bit of what he and Gupta already have developed, which will give them a chance to test in the real world some of the ideas they've been working on. He hopes to have something to test by the end of the year.

Once they have developed working systems, Gillan will become involved again in user testing. The researchers want to be sure that the implementation works as well as they thought it would, that all the parts work together and that the tools are easy to use, Gillan said.

Because there are relatively few visually impaired users at NMSU, the researchers will collaborate with colleagues around the world to help them test their ideas.

Karshmer has been studying assistive technologies for 15 years, but until recently funding wasn't available in this area, he said. This is his second substantial grant for research in this field.

"NSF is the largest and most prestigious science funding agency in the world," Karshmer said. In this round the NSF funded only four out of 111 proposals, which indicates the agency considers research in this area to be important, he said.

The number of disabled users is relatively small, which is why the situation has scarcely been addressed by industry, Karshmer said. "But the disabled population is growing, and as more elderly people buy computers we'll have to build better interfaces for abled' people," he said.

In computer science research, people often ask about a user product, Karshmer said. This is not the goal of their research. "We want to give people the knowledge they need to build the software," he said. "We want to find proof of what will and won't work. We'll provide the research, so someone else can develop the product."