We present a model for predicting expert text entry rates for several input methods on a 12-key mobile phone keypad. The model includes a movement component based on Fitts ’ law and a linguistic component based on digraph, or letter-pair, probabilities. Predictions are provided for one-handed thumb and two-handed index finger input. For the traditional multi-press method or the lesser-used twokey method, predicted expert rates vary from about 21 to 27 words per minute (wpm). The relatively new T9 method works with a disambiguating algorithm and inputs each character with a single key press. Predicted expert rates vary from 41 wpm for one-handed thumb input to 46 wpm for two-handed index finger input. These figures are degraded somewhat depending on the user’s strategy in coping with less-than-perfect disambiguation. Analyses of these strategies are presented. Keywords Text entry, mobile systems, mobile phones, keypad input, human performance modeling, Fitts ' law, digraph frequencies

Mobile internet technologies, such as WAP, are important for pervasive, anytime, anywhere computing. Although much progress has been made in terms of technological innovation, many of mobile internet systems are difficult to use, lack flexibility and robustness. They give a poor user experience. Evaluation and theoretical analysis of usability combined with innovative design can achieve significant improvements in user performance and satisfaction. Using such multidisciplinary methods explains the negative reactions to WAP, and - more constructively - suggest ways of developing more effective and efficient devices and services.

Although text entry is a vital part of day-to-day computing familiar to most people, not much research has been done to enable text entry on large interactive tables. One might assume that a good approach would be to choose an existing technique known to be fast, ergonomic, and currently preferred by the general population, but there are many additional factors to consider in this specific domain. We consider a variety of existing text-entry methods and examine their viability for use on tabletop displays. We discuss these techniques not only in terms of their general characteristics, performance, and adoption, but introduce other evaluative criteria, including: environmental factors unique to large digital tables and the support for multi-user simultaneous interaction. Based on our analysis we illustrate by example how to choose appropriate text-entry methods for tabletop applications with differing requirements, whether by selection from existing methods, or through a combination of desirable elements from a variety of methods. Our criteria can also be used as heuristics during the iterative design of a completely new text-entry technique.

The graphical user interface (GUI) is today’s de facto standard for desktop computing. GUIs are designed and optimized for use with a mouse and keyboard. However, modern trends make this reliance on a mouse and keyboard problematic for two reasons. First, people with disabilities may have trouble operating those devices. Second, with the popularization of wireless communication and mobile devices such as personal data assistants, the mouse and keyboard are often replaced by other input devices. Our solution is a tool that can be used to translate a user’s input to a form recognizable by any Windows-based application. We argue that a formal model of input is necessary to support arbitrary translations of this sort. We present a model, based on Markov information sources, that extends past work in its ability to handle software-based input such as speech recognition, and to measure relative device bandwidth. We also present our translation tool, which is based on our model, along with four applications built using that tool.

We introduce the idea of partial behaviours in user interfaces. A partial behaviour can beguile users, and may be a cause of interaction problems — however, it is possible to identify and remove them early in the design process, making them a useful concept for interaction design. A characterisation of partial behaviours is presented, in terms of a matrix algebra model of interactive systems. We use the model to show some real interfaces have undesirable and apparently unnecessary partial behaviours, and we discuss how choices made when modelling affect our notion of partiality. We also briefly describe a design tool that provides automatic support for partial behaviour analysis.

An iconic text entry system using the widely entrenched numeric keypad is presented. In an iconic system, text entry is designed to remind the user of existing notions of letter shape. For example, the letter "L" is entered by pressing buttons down the center of the keypad, and then to the right, calling to mind the shape of a capital L. The letter "O" is entered by pressing the digit 0, which resembles "O".

Full keyboards are difficult to implement on small mobile devices, and are sometimes replaced by keypads, with multiple characters assigned to each key. The Multitap method is often used for text entry on devices with keypads. While conceptually simple, Multitap requires one or more key presses to enter each desired letter, and is relatively inefficient from the standpoint of the number of keystrokes required to enter each word. It also requires a significant amount of visual searching to find a needed letter on a key. Fortunately, newer methods based on Multitap (such as LetterWise) have been shown to increase users ’ text entry efficiency. This paper presents an enhanced Multitap method that uses predictive next-letter highlighting to aid visual searching. Testing shows that this method, when compared to LetterWise, offers increased text entry speeds, fewer errors, and greater novice user satisfaction. Author Keywords Predictive keypad text entry, mobile device interface

The problem of optimizing keypad designs for predictive text entry methods for handheld devices has received significant interest in recent years. The necessity of such optimization lies in the ever-increasing difficulty of inputting text into mobile devices, particularly those of extremely small size. Predictive keypad text entry methods, such as Tegic’s T9, only require one key press for each letter of a word to be entered. However, such methods are ambiguous since one key sequence may correspond to multiple matching words because more than one letter is mapped to each key. Various keypad designs have been proposed in the past to optimize the placement of letters on keys to minimize ambiguous key sequences. In contrast to past research that allowed mapping of any letters to any keys, we hypothesize that maintaining alphabetical ordering across the keys of optimized keypad designs will result in higher novice usability and ease of learning, while still providing a more immediate benefit in terms of improved performance. Results from usability testing we conducted are encouraging and support our hypothesis. 1

We introduce the idea of partial behaviours in user interfaces. A partial behaviour can beguile users, and may be a cause of interaction problems — however, it is possible to identify and remove them early in the design process, making them a useful concept for interaction design. A characterisation of partial behaviours is presented, in terms of a matrix algebra model of interactive systems. We use the model to show some real interfaces have undesirable and apparently unnecessary partial behaviours, and we discuss how choices made when modelling affect our notion of partiality. We also briefly describe a design tool that provides automatic support for partial behaviour analysis.

Dictation using speech recognition could potentially serve as an efficient input method for touchscreen devices. However, dictation systems today follow a mentally disruptive speech interaction model: users must first formulate utterances and then produce them, as they would with a voice recorder. Because utterances do not get transcribed until users have finished speaking, the entire output appears and users must break their train of thought to verify and correct it. In this paper, we introduce Voice Typing, a new speech interaction model where users’ utterances are transcribed as they produce them to enable real-time error identification. For fast correction, users leverage a marking menu using touch gestures. Voice Typing aspires to create an experience akin to having a secretary type for you, while you monitor and correct the text. In a user study where participants composed emails using both Voice Typing and traditional dictation, they not only reported lower cognitive demand for Voice Typing but also exhibited 29 % relative reduction of user corrections. Overall, they also preferred Voice Typing. Author Keywords Speech recognition; dictation; multimodal; error correction;