SwiftKey Predictive Text Keyboard Learns Your Lingo

SwiftKey recently introduced a new application for Android users that promises to predict your next word with "incredible accuracy." To do this, SwiftKey uses TouchType’s Fluency technology in order to predict the next word you are likely to type.

SwiftKey has been two years in the making. Now, it's available in beta version through the Android Market. According to the British start-up company TouchType, SwiftKey can increase writing speeds by up to 50% with its powerful prediction engine.

Although SwiftKey uses your own messages to help predict future text, the company assures users it does not collect personal data or log your conversations. Instead, the software "simply observes how you use individual words next to each other and builds a personal language model from this, which it stores on your phone’s SD card."

To give you a better idea how SwiftKey Beta works and how fast it can be, check out this video. We have to admit: It really does look impressive.

Jennifer Johnson

Jennifer Johnson

Jennifer grew up around technology. From an early age, she was curious about all things related to computers. As a child, Jennifer remembers spending nights with her dad programming in BASIC and taking apart hard drives to see what was inside. In high school, she wrote her senior term paper on her experiences with building custom computers.

Jennifer graduated from the Jeffrey S. Raikes School of Computer Science and Management at the University of Nebraska at Lincoln. After college, she began writing full-time for various PC and technology magazines. Later, she transitioned to the Web. In these roles, Jennifer has covered a variety of topics including laptops, desktops, smartphones, cameras, tablets, and various consumer electronics devices. When she's not playing with or writing about the latest gadget, Jennifer loves to spend time with her family, capture memories with her camera, and scrapbook.

Opinions and content posted by HotHardware contributors are their own.