Predictive text is a technology that allows cell phone users to use one key to represent multiple characters when typing a message. The system considers the sequence of keys pressed and offers the most likely candidate for the word the user intended. If the word is not correct, the software can offer a selection of options, and the user chooses the one she wants. This technology learns with the user over time and can increase speed and accuracy when sending text messages.
A number of companies have developed predictive text technology for cellphones. With the rise of smartphones with full, albeit small, keyboards, users were less likely to need this feature, but some phones continue to utilize the software. One of the pioneers in this field was developer Martin King, who created the T9 predictive text system.
This software comes in a number of languages, and users can program it to add words. Some phones also have multiple dictionaries for bilingual users. Thus a user could switch between Spanish and English, for example. When the user first starts using predictive text, the program will base its suggestions on what is known about average word use in a given language. As the program learns, it will give weight to words the user chooses more frequently.
In addition to allowing users to type words more accurately, predictive text can also suggest words before the user is finished typing. As the typist works, it scans through the sequence of characters to predict the most likely word he plans to use. It can flash the word on the screen for the user to select or dismiss, if it wasn't the word he wanted. This can further increase speed, especially as the program learns more about the user's word choices and typing habits.
One inevitable consequence of predictive text was the rise of mistakes made by choosing a word in a hurry or not paying attention to the word the program selected automatically. Such errors can cloud the meaning of a sentence and on occasion can create serious misunderstandings. In programs that also autocorrect, or use a sequence of characters to generate the closest word, sometimes users end up with an unexpected insertion in a text message. Autocorrect software is also present on many smartphones, and it draws upon some of the same technology as predictive text.