After the success of gesture-based keyboards such as Swype, the next obvious disruption to keyboard technology is optimisation of the legacy Qwerty layout that’s persisted since the typewriter era. Not that people haven’t tried alternatives to Qwerty already (e.g. Dvorak et al.) – and generally failed to make them stick. But that’s not stopping a group of academic researchers — including the co-inventor of the gesture IP behind Swype — from devising a new touchscreen keyboard layout in the hope that people can finally be persuaded to shift their typing habits.
KALQ, which is named, like Qwerty, after a string of its keys, is designed to speed up thumb typing on tablets and phablets (aka big phones). Its creators, who are from the University of St Andrews, the Max Planck Institute for Informatics and Montana Tech, claim that once users have accustomed themselves to the non-Qwerty layout — with about eight hours practice required to be as fast as Qwerty and 13-19 hours to surpass your Qwerty typing speed — typing performance can be about a third (34 percent) more efficient than thumb typing on split-screen Qwerty layouts.
They are planning to release KALQ as a free Android app for tablets and phablets, which will also work on smaller screen smartphones but stress their research and performance claims relate specifically to larger devices, rather than phones. They are also not directly comparing the performance of the new layout against any of the gesture keyboard input methods (Swype, SwiftKey’s Flow etc) — their performance data is based on a direct comparison with thumb typing on a split Qwerty.
Dr Per Ola Kristensson, Lecturer in Human Computer Interaction in the School of Computer Science at the University of St Andrews, who is one of the academics involved in the research, told TechCrunch they tested KALQ on a Galaxy Tab 7.7, adding that while the keyboard may also offer speed improvements on smartphones it’s not a claim they have tested. Kristensson is no stranger to keyboard disruption, being the man who wrote the pattern recognition algorithm underlying Swype, and co-founder of ShapeWriter, the startup that commercialised the gesture keyboard system in 2007 — before being acquired by Nuance in 2010 (the company that now owns Swype).
Kristensson said the KALQ researchers used a subset of publicly available emails from the Enron trial that were tagged ‘Sent from my BlackBerry’ as their data pool, analysing the mobile users’ use of language to figure out the best positions for the keys. As well as using computational optimisation techniques and looking at how devices behave when users are touch typing, they also modelled thumb movements with the aim of making a fast yet comfortable keyboard. KALQ is an English-language optimised letter layout, but the process that came up with its layout is “general,” said Kristensson: “You can feed it whatever language you want. So the layout may change, depending on your country.”
There’s been lots of crazy text input technologies proposed… The problem with a lot of them is they are not fast enough.
For English speakers, KALQ’s split-screen layout repositions the alphabet into two unequal blocks of letters, with consonants in the left block (plus Y which can be classed as either) and vowels plus the remaining consonants (including K, L and Q) in the right. A space key is included towards the edge of each block for easy reach with either thumb. The letter order is specifically designed to minimise typing long sentences with just one thumb — which is cumbersome and slows touchscreen typists down — and also places frequently used letter keys centrally close to each other to minimise thumb movements. In addition, the layout generally aims to encourage typing on alternating sides of the keyboard — which Kristensson said is a more ergonomic and comfortable way to type.
As well as learning the new letter layout, KALQ typists need to learn to move both thumbs at once to get the fastest speeds. “Experienced typists move their thumbs simultaneously: while one thumb is selecting a particular key, the other thumb is approaching its next target. From these insights we derived a predictive behavioural model we could use to optimise the keyboard,” noted Dr Antti Oulasvirta, Senior Researcher at the Max Planck Institute, in a statement.
The researchers said trained KALQ users were able to reach speeds of 37 words per minute — which they said is the highest ever reported entry rate for two-thumb typing on touchscreen devices, and “significantly higher” than the approximately 20 words per minute entry rate users can normally reach on a regular split Qwerty layout. The group will be presenting its research next month at the ACM Conference on Human Factors in Computing Systems in Paris. The Android KALQ app will be available for download in due course.
Persuading users to adopt a new keyboard layout is likely to be a tough ask but Kristensson said the problem with most of the Qwerty layout challengers to-date has been that they are not disruptive enough — in terms of the performance bump they offer users who have to go through the pain of learning how to type quickly again.
“If you want to get people to change their layout you basically have to get people to invest, you have to get them to give up the assigned cost, their previous investment in Qwerty typing. And then we have to invest new time in learning KALQ,” he said. “There’s been lots of crazy text input technologies proposed. Actually hundreds of them. Most of them have failed. I would say probably 99% of them have filed but the problem with a lot of them is actually they are not fast enough so why would people reinvest in learning a new text entry method if it doesn’t provide a substantial performance advantage so I think [KALQ] is one of the few keyboards that can provide that. So I’m hopeful.”
Asked whether the group might look to commercialise the research, he said the priority is to try to encourage people to adjust their typing behaviour and accept a Qwerty alternative but added that the group may look to monetise their algorithms in other ways — by, for example, using them to optimise other menu-based user interfaces.
“What I’m hoping here is that we will have impact,” he told TechCrunch. “I wanted to get people away from thinking about the Qwerty keyboard. And I think impact here may mean that we will release [KALQ] for free — but remember we are the ones who have all the algorithms to come up with optimal keyboards so we learn a lot about how to optimise user interfaces in general. My co-investigator, Antti Oulasvirta, he’s completely passionate about optimising any sort of user interface. So the process we use here can also be used to optimise other user interfaces like menu structures for example so there is lots of potential for the underlying technology. This is just one instantiation of that. But I think trying to sell a new keyboard — that’s a risky proposition. I’m not sure a venture capitalist would go for it.”
