This book starts with a heavy dose of statistics and formulas. “Oh my,” I thought, “What did I get myself into?”

Chapter 1 is an overview of text entry systems, starting with the early typewriter--interesting if you’ve never heard any of it before. Chapter 2, “Language Models for Text Entry,” by Tanaka-Ishii, one of the editors, jumps right into n-gram, hidden Markov, adaptive, and other mathematical models for predicting what comes next when a writer types a letter or two. Chapter 3, by Wobbrock, goes into mathematically measuring text entry speeds and efficiencies (in other words, if you can enter text quickly but can’t correct typographical errors easily, the method is inefficient). Chapter 4, by Mackenzie, describes how to set up experiments to evaluate text entry speed and efficiency. It contains less math, but is still abstract. I am not averse to math, but I never got a good grounding in it in school. It was therefore a struggle to pay attention to the points the authors were making in between formulas.

The effort was worth it, though, since subsequent chapters refer to these analysis methods. And why is that important? It’s important because this isn’t just a book about various text entry methods--interesting enough on their own--but about which ones work, for whom, and how well.

Several text entry systems are described in Part 2 (chapters 5 to 9). Chapter 5 covers various systems for entering text on mobile phones. To deal with the problem of having a small number of keys on a mobile device, each of which needs to carry more than one letter, different types of keyboards and different kinds of completion algorithms are needed. Chapter 6 discusses handwriting recognition interfaces for English or Latin alphabets. Shape writing that uses pen movements over an on-screen alphabet to create words is covered in chapter 7. This looks like it would be a lot of fun. Experiments seem to indicate that it is quick to learn and use. The next chapter deals with speech-based interfaces for capturing natural speech and dictation. Chapter 9 discusses eye tracking, which is used by people with motor disorders. There are a few types: text entry by gazing at a letter and pausing for a set number of seconds to select it; gazing and blinking or moving eyes to the left or right to select it; and a zooming interface that lets users navigate through the letters and select words predicted by the software.

Part 3 (chapters 10 to 13) covers non-Latin languages and character entry systems. Chinese, Japanese, Korean, Thai, Hebrew, Arabic, various Indian languages, Turkish, Cyrillic, and other interesting text entry puzzles on both keyboards and mobile devices are covered.

Part 4 (chapters 14 to 16) discusses text input for children and the elderly, as well as for those with either physical or visual impairments (for example, those trying to enter information in a moving vehicle).

Unlike many compilations, each chapter in this book is both well written and dense with information. This is remarkable because it is hard to get one good book out of many authors. Since each chapter also includes information about the research done on the efficiency of the entry method, the editors must have also had their authors back up all claims with data. If your organization has anything at all to do with text entry systems, you’ll find this book invaluable.