Archive for June, 2008

Listen up! Technology, Materials, and strategy for non-Visual Reading

June 22, 2008

My adoption of the audio reading mode

This post describes how this vision Loser reads on a daily basis. sighted readers of this blog should gain some insight into alternative ways technology delivers what you read visually on printed pages or screens. Those now in transition with vision loss can get a snapshot of a specific combination of reading technology, web delivery systems, and kinds of reading materials.


I consider myself an effective reader at this point in my vision loss. Three years ago I would have had no way of describing how I would be reading now. Partially, this was from the inability to know how my sensory apparatus would be working. For the record, I see pages where the text is mostly smudges. Computer screens have reasonably clear outlines with text that can be enlarged on a monitor or text size setting but remains often more like those irritating CAPTCHA boxes, all wobbly and sliced up. Partial sight can be minimally used by magnification, contrast, and eccentric viewing but for any reasonable way of consuming information, one must step over into the audio world. That means a screen reader or self-voiced reading devices, all using synthetic speech. After 2 years of hard work, a lot of technology evaluation, and countless hours of practice, the audio world now seems natural. I have no problem reconciling myself with this way of reading for the rest of my life, trusting that my hearing and hands will not give out on me.

My portfolio of reading devices

Another reason I would not have been able to predict how I read now, in 2008, is that several products I use constantly had yet to be invented in 2005. Processing power, miniaturization, wireless, and blind-driven inventiveness have produced a stable of devices that complement the PC (or MAC, whatever).

  • The Levels tar Icon is a screen-less Linux hand-held that reads all its menus and text as I cycle through email, news, and web content. The Mobile Manager hand-held fits into a docking station with keyboard and augmented speakers, power, and ports. I use the Icon for email by pop3 from gmail, occasional recordings,RSS feeds of news and podcasts, web browser, and special access to books and newspapers.
  • The American Printing House for the blind book port is another hand-held box with its only user interface a keypad, requiring ear buds or external speakers. Its memory card is loaded from a PC with books, mp3 files, and text. The book port is designed for easy navigation through books and its file systems. Like the Icon, it can also record memos. The APH book port is currently available only used, as the upgrade is having manufacturing problems. I use the book port primarily for books and lengthy synthetic spoken versions of files. A competitor Humanware victor reader stream offers similar reading capabilities, but I have never become comfortable with its navigation techniques, primarily just not my way of working.
  • The latest marvel of reading technology is the > Kurzweil NFB reader that has shrunk the scanner-OCR-reader architecture onto the Nokia N82 platform. well, it could be used also to make phone calls if attached to a phone service. This little guy is great for on-the-fly reading like room service menus, TSA notices stuffed in your luggage, mail, and printed pages lying around. one of the greatest frustrations of print disability is the difficulty of performing normal inter-human transactions where a sighted person hands you a business card or information sheet or agenda and you need that information to take the next step toward your goal. Another frustration is the profusion of junk materials surrounding the little piece of critical action, like amount to pay on a bill, but that’s where family members can be called upon. The KNFB Reader illustrates Kurzweil’s mantra that exponentiation dominates linearity , urging us to think about potentially using far more computing power to overcome our neural deficiencies.
  • The NVDA screen reader , discussed in earlier posting on selection of NVDA , is my PC work horse. It shows amazingly high quality and functionality for a young product, deriving from its free, open source origins driven by a generation of blind tech savvy developers and users seeking an alternative to the proprietary screen readers forged into the rehab-industrial complex. Note: I donate to NVAccess. Unless you need specialized scripts for complex or barely accessible products, such as many enterprise data management systems, NVDA will do well, especially in conjunction with Mozilla products.
  • Another supporting tool necessary for full reading is the Kurzweil 1000 for simplifying and managing scanners, which may have inaccessible and photo-oriented interface managers. Scanned material for submission to a service like Bookshare.org requires considerable editing that is well supported in K1000. I used the K1000 for general editing and spell checking as well as scanner management. Note that the K1000 has its own nice self-voicing practice to assist its operations and editing.


So, that’s all new technology I’ve learned in the past 2 years, ranging from my Identity cane to a suite of talking devices.

Sources of reading materials

What about the representation of the reading materials and where to they come from?

Human narrated audio books

Of course, we are all familiar with humanly recorded audio books, basically a long stream of bits, possibly with some embedded strings that reader technology can identify as section or information markers. The blind-serving organizations like NLS (National Library service) has long provided human narrators, recording media, reading tools, and a library coordinated distribution system. I personally have not tapped into this because the NLS format has only recently become available on the Icon, and, besides, I have a little problem with its paperwork to get myself certified. audible.com is the commercial system integrated with book port and soon the Icon, but I have yet to find the book that compels me to subscribe.

Note added December 08: Other sources of narrated materials are available in podcast format. Librabox podcast delivered book chapters is prolific and well done. Assistive Media Magazine readings extracts popular New Yorker style articles. State services like Arizona Sun Sounds offers books, newspapers, government information.

DAISY, Digital Talking Books and Bookshare.org library

The core technology for representing reading materials is XML, for extended markup Language, in the family of HTML for web pages. Text files have human or automatically added tags, like <title>,which the reader tool interprets for the user, which could be another computer or a human. A special version, DAISY (Digital Accessible Information System) is the interchange format for books. I get most of my books from bookshare.org which uses a copyright exemption to allows volunteers and publishers to contribute texts for distribution to members certified with a print disability who agree not to distribute further, but with free choice of reading tools and locations of materials. For me, this meant I could rebuild my personal library faster than I could donate or throw away my printed books.


The beauty of the bookshare distribution system was immeasurably enhanced by the Icon’s integration of a book search and download capability. If I hear about a New York Times best seller , a classic or a Reader’s choice, I can pull up the Icon book search by title or author, automatically log on to bookshare, download the book, if available, and start reading — in about a minute! Of course, if the book is not available, I can look for an audio at the public library or a commercial service or get a printed copy to scan. Indeed, I am now contributing books selected by my monthly AAUW book club, which takes several hours of work as I learn to expedite scanning and editing with the Kurzweil 1000 system. But it’s gratifying to know this process offers good readings to thousands more people like me. I carry my entire library on my easy reading Book Port categorized as Fiction, Biography, etc. and can also search these books in full text format. This pipeline of easily retrieved and stored books has truly broadened my reading choices with more than enough entertainment, enlightenment, and information.

Not yet available digital book collections

What about all those mass scanned book collections by Google, amazon, Microsoft, etc.? And those PDF e-books? too bad, most of these are not available to me, or very hard to use. The popular Gutenberg and Google book search do provide out-of-copyright materials, but I personally rarely need these. And, as I commented in post on “seeing through Google book search” , I am limited in my research by the image-only presentation of pages from a book search. While PDF is a nearly universal viewable distribution format, the adobe Acrobat reader is always changing its read out loud capabilities, insists on updating itself every use, and generally makes me feel out of sorts, like “when good technologies go bad”, with apologies to the adobe co-founder who was my grad school office mate. PDF accessibility is such a mixed bag, I just convert all PDF files to TXT and live with what I can get out of the results using the Icon, book port, or screen reader. My pet peeve is the need to convert PDF newsletters into TXT when the content could just as well been delivered as the more easily readable HTML. Like many other people, I thought I could buy an ebook and apply a synthetic voice reader but this mode of distribution is verboten by DRM (Digital Rights Management).


Whew, this is getting long, as I inventory my reading experience, but here some the happier parts.

More news than ever from NFB via Bookshare and Icon News Stand

As my vision faded so I could no longer read newsprint comfortably, I kept my NY Times subscription to retain access to the web site. I learned to find the sections of interest, like Editorials and business, and navigate a link path while reading the articles I wanted by the Text Aloud browser toolbar. Ouch, was this cumbersome! Now, I use the NFB Newswire newspaper delivery service offered in the bookshare membership and facilitated by the Icon News Stand application. With one “get new issues” click, I have not only the NY times, but also wall street Journal, Washington Post, San Francisco chronicle, economist, New Yorker, and more. All are structured for reading by publication, issue, section, title, and text. this means I can scan and selectively reads 100s of pages of newsprint in a half hour, an unpredictable benefit of print disability.

Local news, the gaping hole in the infrastructure

Of course, there’s a down side to news reading in that my local newspaper uses a convoluted content management system that seems to split every article into paragraphs that intertwine with advertisements and obituaries. Luckily, there is an RSS that delivers titles and a city feed that offers more official news, but I have yet to find a way to keep up on local events, even using the radio. This is one of the gaping holes in the information infrastructure for print disabled readers. I avidly track Jon Udell’s blog on strategies for Internet citizens for improving community networked information.

RSS feeds as supplementary and primary news sources

Along the lines of the DAISY representation for books is the RSS (real simple syndication) format for feeds that deliver articles and podcasts. This is the key technology for the rest of most of my reading, with over 80 feeds in my Icon RSS client. These bring CNN, Inside Higher Ed, science daily, slate, and many more magazine and news headline style materials. These are complemented by my evolved collection of news, recreational, and technical podcasts. While I really do not know what I am missing, I am thoroughly comfortable that I am keeping up with technology trends through itconversations.com with its interviews with innovators, technation, IEEE spectrum, etc. Rarely is a podcasts a time-waster and I feel myself obligated to listen to keep up. Similarly, a judicious selection of blogs help me track what’s going on in my areas of interest, including accessibility, podcasting media, and, especially this year,politics.


Two cool things about RSS are the ability to hierarchically structure feeds and to exchange feeds among readers. If you want mine, here’s susan’s reading sources , a file that can be imported into your choice of RSS reader or cribbed from in a text editor. Since all navigation in the Icon RSS reader is within a tree, I have a hierarchy of News into general, technology, Politics, and science categories, then further in places into trees of blog or other special content. Since feed updating is time consuming, maybe half an hour, the tree structure allows updating only a single feed or group of feeds, e.g. if I need a politics fix late on a Tuesday primary day. Of course, I also have several mailing lists with associated folders in the Icon email client, keeping up on mdsupport.org,book port, bookshare, NVDA, and icon user discussion lists.

Progressive reading productivity and quality

How progressive are these reading tools? I have been an Internet user since around 1970s. Indeed I found myself on the mailing list of the very first spam message 30 years ago. I subscribed to and made some embarrassing posts in Usenet groups and mailing lists in the 1980s and 1990s and had my first web page around 1993. To me, this surfeit of information is a natural progression. However, when my beloved Icon had to go to the shop for repair, I realized how important were the advances of the past year. I found the web-based RSS readers clumsy and never did get any setup comparable to my Icon trees, menus, and quick read articles.


To be provocative, I estimate my reading productivity now, compared to a few years ago, as about 10:1 in retrieving content available via Internet, wireless, RSS and other clients. Once retrieved, I feel about a 10:1 gain in ability to scan, filter, selectively read or listen to the content. Of course, I cannot get everything I need and occasionally rev up the Icon or PC Firefox web browser for searching and surfing. I’ll discuss my feelings about information overload and reading habits and brain plasticity in the companion post on “Hyperlinks considered Harmful”.


One of the greatest benefits of exploiting vision loss and using these reading tools is that advertising fades into the noise. Given the current economic model for most information services, this makes me a lousy consumer. Well, too bad, I really would like to kick in for a low-cost subscription, say $10, but do not have that opportunity. I’d like to pay $3 for each book I read with funds to the author and publisher, like is occurring for music. But my guilt is assuaged by taking every opportunity to tell in person and virtually about resources I like in hopes that enough people will click the ad links and buy the resources directly. And, much as I love my reading tools, losing vision is costly, nearly $10k for the above tools.

Advice for both sighted and impaired readers

so, you still fully sighted readers should now have a sense of how one vision Loser has replenished her reading vessels with forms of content, like DAISY, and tools that you never heard of and would consider primitive compared to iPhones and quicktime. But, if my claim of 10:1 increased retrieval and 10:1 improved reading hold true, this step over into the audio world is hardly a loss of reading capability. Limited access to certain kinds of material are offset by opportunities to access special content not available to the sighted world, like the bookshare library and the NFB News Line.


For those losing vision, as I have for three years, I urge you to begin tapping into this audio world sooner than your denial and hopes might lead you. Try using a free screen reader and audio conversion tools and get used to gaining more information by audio whenever you feel discomfort with your eyeballs glued to your screens. I hope this article assures you there are many ways to adapt your reading styles to meet your needs, and even to find gains you never dreamed of. You might visit a disability services department at a local university or an assistive technology demo exhibit hall. But beware, that the rehab and disability services personnel are themselves grappling with technology learning curves and are locked into vendor distribution practices that lag behind some of the tools I advocate in this blog. A good starting point, whatever your level of sightedness, are the user stories in nextup.com text to speech blog

For More Information on Assistive Technology

.

Need a second medical opinion? Try the Controversy Discovery Engine.

June 3, 2008

A better way to search for analytic web content??

This post offers a way of searching for more diverse and analytic results using a simple web form interface to Google. This approach is especially useful when you are looking for a second opinion, evidence, or authorities on topics like we sometimes face with vision loss. It can also make querying and searching more efficient for our weary fingers by slicing off less useful results from searches. Please give it a try and let me know if it improves your searching.

Searching for better information on ‘myopic degeneration’

First, some background. My recent Retinal Specialist appointment provoked my curiosity as my Myopic Macular Degeneration (MMD) seems to have stabilized. I have been wondering about origins and distributions of this condition, as I have only met other MMD people on the more comprehensive Macular Degeneration mdsupport.org earlier post. There’s always a sliver of hope for improvement, possibly from research driven out of the U.S. by stem cell policies. And, always, looms the now effective intervention of repair surgery or injections for retinal detachments or so-called “wet” conditions.

Time to update myself, so I go to Google and find the usual results for the query "myopic macular degeneration". Top results are mostly generic overviews "MMD is related to AMD", but I also find a lengthy Myopic Manual.

Embellishing searches with controversy-related terminology

Fourteen years of searching has taught me I might need to go quite fa r down the Google results list to get into more in-depth discussions. I really wanted to know about the controversies, debates, arguments, and even spats in the related field of ophthalmology, genetics, nutrition etc. So, why not just add the word "controversy" to the query. Indeed, I see different results, but why stop there? Speaking linguistically, and assuming Google is fairly literal, I might want to use variations such as "controversial" or "controversies". Then the thesaurus adds synonyms such as "debate", "argument", "disagreement", and many more, each with variants. Now, I also want supporting material so I might ask for "evidence”, “proof”, “hypotheses”, “opinion” and all these variants. This is a lot of decision making on synonyms and support and variant, typing each and saving for reading those interesting results.

A simple form customizes controversy-related content

Primarily, I am getting deeper and faster into the subject matter. Is there a better way to query Google to achieve these goals? Well, yes, as I tried 5 years ago and dubbed the Controversy Discovery Engine. Go ahead and try it. Type your query into the search box, choose a controversy synonym, optionally select a kind of support, and hit the button. Your embellished query will be sent to Google, asking for 50 results. That’s all there is to it. You might or might not get better results than your hand-crafted queries but at least you now have a lot of packaged queries with just a few extra clicks.

An experiment on ‘Do search engines suppress controversy?’

Why do I claim this approach often works better? Well, driven by curiosity, I performed an empirical study on "Do Search Engines Suppress Controversy?" that was published in First Monday January 2004 online. Now, it’s not that search engines or search engineers have political agendas, but rather just an effect of the link popularity strategy that makes Google search work so well. The web splits into an Organizational web which links the promoters, explainers, and associations for a topic apart from the Analytic Web that includes scholarly papers, blogs, white papers, individuals, etc. The Organizations link among each other and people link to organizations more than the Analytic Web pages are linked to from the Organizational Web or within the Analytic Web. This pushes controversies down the list of search results. Usually controversies are hard to name in queries and you need to know the controversy exists by some name to query for it.

For example, one controversial aspect of Albert Einstein was whether the first wife he dumped had contributed rather more to his research career than was acknowledged. Query for "Albert Einstein AND Mileva Maric" and, voila, the controversy is revealed in various levels of details and with arguments on both sides of the story. Bet you didn’t know that! Using a synonym for controversy raises pages that discuss his personal life and produce the names, like Serbian physicist Mileva Maric, for additional searches. This particular revelation ebbs and flows with the tide of publications on his work and life. So, our approach is to use the language of human endeavors that involve research and the give-and-take of the intellectual marketplace to morph our searches more into the Analytic Web.

More seriously, for medical conditions, people facing surgical decisions want all and the most authoritative information they can get as fast as it can be found. So we offer the Controversy Discovery Engine as a kind of “second opinion” information seeker. Please provide feedback and suggestions to slger123@gmail.com. This web page may also be modified for similar uses with appropriate link and acknowledgement. If you’re intrigued with this topic, which won me the proverbial 15 minutes of fame in the blogosphere, read the paper and its five examples: St. John’s Wort, female astronauts, Albert Einstein, Belize, and distance learning. For the real search gurus, the software instrument used in this experiment, dubbed twURL, is available for licensing.

For visually impaired readers, here is a bit more advice. The web page has four form elements with the search query edit box at the top and submit button at the bottom and two list boxes with multi-selection in between for synonyms and support. You can multi-select from the list or select NONE as the last list item. Remember to turn on the virtual buffer in a screen reader to type in the query and select from the lists. Using sight, you might want to pump up the text size using your browser, e.g. Control + in Firefox. If you use this page a lot and know how to edit HTML, save the page and customize its style to your taste.

Try searching more diversely and deeply into the Analytic Web

So, nothing to lose and possibly lots to gain, check out the Controversy Discovery Engine at http://apodder.org/ControversyDiscoveryEngine.html and let me know how it works for you.