On Automated Generation of Keyboard Layout to Reduce Finger-Travel Distance

On Automated Generation of Keyboard Layout to Reduce Finger-Travel Distance

Amol D. Mali (Department of Electrical Engineering and Computer Science, University of Wisconsin-Milwaukee, Milwaukee, WI, USA) and Nan Yang (Department of Electrical Engineering and Computer Science, University of Wisconsin-Milwaukee, Milwaukee, WI, USA)
DOI: 10.4018/IJMSTR.2017040103
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

The QWERTY keyboard layout can be very inefficient for one-finger typing on virtual keyboards since the letters in many common digrams are placed on opposite sides of the keyboard, resulting in a long finger travel. This paper reports on use of simulated annealing for finding alternate arrangements of the letters of the English alphabet on keyboards with different number of rows, to reduce finger-travel distance for entering text. The use of simulated annealing led to arrangements of the letters on 3 × 10, 4 × 7, and 5 × 6 layouts with a lower weighted sum of finger-travel distances for all digrams (denoted by d) compared to the QWERTY layout (lower by about 40%). The layout with the least value of d among those found in this work is a 5 × 6 layout for which the value of d is only 1.78 key widths compared to 3.31 key widths (the value of d for QWERTY). Alternate automated methods to solve this problem, connections between keyboard layouts and facility layouts, and many new applications of the ideas in this work are discussed.
Article Preview

Introduction

Text entry on a mobile device via a virtual keyboard is one of the most common activities in the mobile computing era. ExactTarget survey (www.marketingcharts.com) shows that the top activities on smartphones are accessing emails and text messaging. Both involve a lot of text entry. A study (techlandtime.com) found that cell phone users between the ages of eighteen and twenty-four years exchange 110 text messages per day on an average. Given the huge popularity of mobile text entry, any improvement in its performance would offer a huge benefit to a large number of users.

QWERTY has been the default layout for entering text for over a hundred years. However, it can be very inefficient for text entry on mobile devices. One of the rationales behind the QWERTY layout is that placing letters in common digrams on two sides of the keyboard requires a user to alternate between two hands during typing, and can avoid mechanical jams in typewriters. Such a layout turns out to be suboptimal for virtual keyboards on mobile devices, where many users enter text with only one finger. Because letters in common digrams are placed on two sides, the finger entering the text usually needs to travel back and forth multiple times between two sides of the keyboard, to enter a word. For example, Figure 1 shows the finger-travel trace for entering “open”. As shown, the input finger travels across the keyboard twice for entering this simple four-letter word.

Long finger-travel distance is detrimental to text entry. First of all, long finger-travel distance results in low input speed. Secondly, in one-handed typing scenario in which a user holds a phone with one hand and types using the thumb on the same hand, moving the thumb back and forth between two sides of the keyboard is unpleasant due to the limited range of thumb movement. Third, recent research shows that spending too much time texting can lead to neck and shoulder pain (Paddock, 2009).

Figure 1.

Entering “open” on the QWERTY layout

Faster text entry contributes to an earlier delivery of the text, enabling an earlier fulfillment of the objectives of sending the text. Faster text entry may become essential in case of an emergency or in extreme weather when one may have to type without a protective covering.

Given the widely-known problems of the QWERTY layout, can one design a layout for one-handed typing that requires a finger to travel less? In this paper, simulated annealing is used to investigate this question. The QWERTY keyboard layout is modified in this paper in order to lower the weighted sum of finger-travel distances for all digrams (denoted by d). The value of d for QWERTY is 3.31 key widths. Applying simulated annealing, 3 × 10, 5 × 6, and 4 × 7 layouts for which d is lower by about 40% compared to QWERTY were obtained. This article is a significantly extended version of (Yang & Mali, 2016). This extended version includes a much larger literature review, alternate automated methods for solving the problem of finding a keyboard layout to reduce finger-travel distance, and many new applications of the ideas used in modifying the QWERTY layout.

Work On Enhancing Keyboard Use

It has been known that the QWERTY layout is suboptimal for mobile text entry. As has been published (Rick, 2010), (Yamada, 1980), QWERTY was initially designed to reduce jamming in mechanical typewriters by placing letters in common digrams on opposite sides of the keyboard. Such a layout can be inefficient for one-finger text entry.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 5: 4 Issues (2017): 2 Released, 2 Forthcoming
Volume 4: 4 Issues (2016)
Volume 3: 4 Issues (2015)
Volume 2: 4 Issues (2014)
Volume 1: 4 Issues (2013)
View Complete Journal Contents Listing