Text entry software spells relief for weary BlackBerry thumbs
By Mike Langberg
Mercury News
Humans in their long history have invented only two ways for individuals to produce text: handwriting and typing on a keyboard.
Shumin Zhai, an IBM scientist, may have invented another way: SHARK, an abbreviation for ShortHand-Aided Rapid Keyboarding.
SHARK is intended for writing text with a stylus on small touch-sensitive screens, such as those found in cell phones and personal digital assistants. It uses a radically different approach that is easy to learn and fast.
Zhai gave a presentation on SHARK earlier this week during a conference at IBM's Almaden Research Center in San Jose, where he works, and his ideas led me to one of those rare eye-opening moments when you suddenly see what could be the answer to a long-festering problem.
The problem is how to comfortably enter text onto the screens of pocket-size devices. Anyone who uses a cell phone to laboriously tap out text messages, or a BlackBerry to respond to e-mail, knows what I'm talking about -- those tiny keypads cause thumb strain. At the same time, software that attempts to recognize either speech or regular handwriting isn't reliable.
SHARK in its current form -- demonstrated so far on full-size Windows XP notebooks -- puts a small grid on the screen. The grid shows a cluster of letters in what appears to be random order.
To write a word, you put the stylus on the first letter of the word and then drag the stylus to draw a line through the alphabet cluster, touching every letter in the word. When you lift up the stylus after hitting the last letter, SHARK figures out what word you want and displays it on the screen.
If SHARK makes a mistake, you tap the word and get a list of the most likely alternatives based on the path you traced through the grid.
Zhai sent me the most recent version of SHARK, and it worked like magic. (I didn't have a notebook with a touch-sensitive screen, but I found a mouse works in place of a stylus.) More than nine times out of 10, the software displayed the word I wanted. When SHARK got it wrong, the correct word was almost always the second or third choice on the drop-down list.
If you want to see for yourself, Zhai has posted an earlier version of SHARK on his Web site (www.almaden.ibm.com/u/zhai/SHARK.htm) Zhai, working in collaboration with graduate student Per-Ola Kristensson, is promising to post a much-improved SHARK Version 2 by the end of August.
Over time, SHARK users learn to make familiar words just by drawing the appropriate line on the screen -- for example, ``the'' is a simple reverse C motion.
Zhai calls these simple line drawings ``sokgraphs,'' and says users can learn sokgraphs for the 50 to 60 mostly commonly used words in four training sessions of 45 minutes each. Those words represent about 40 percent of what we write.
Think of renting a car. When you get in your own car, you instantly stick the key in the ignition, release the emergency brake and put the vehicle in gear, all without looking down. But when you get in a rental car for the first time, you have to look around to find all the controls. After a week in a rental car, you have to think twice when you get back in your own car.
In the same way, SHARK sokgraphs build up in your ``sense memory'' over time. But you're never stuck, because you can always look at the SHARK screen if you don't know where to move the stylus. You can even tap out letters one at a time if necessary.
In lab tests, Zhai says volunteers have achieved writing speeds of 50 to 80 words per minute. That's slower than a proficient touch typist, but much faster than using two thumbs on a keypad.
IBM hasn't yet decided how to commercialize SHARK, but Zhai is confident the software could be ready for paying customers with a few months notice. The processor and memory requirements are low enough, he adds, that SHARK could run on most of today's cell phones and PDAs.
As enthusiastic as I am about SHARK, I've learned to never mistake a research project for a sure thing. SHARK may be too different from conventional writing or typing to gain public acceptance. And Zhai has yet to prove his software will be fully functional on handheld devices.
But, if nothing else, SHARK stands as a challenge to all the bright people in Silicon Valley working on next-generation wireless gadgets: If you don't like SHARK, then invent something better. Our tired thumbs will salute you.
mercurynews.com |
