Abstract: Quick tasks for on-screen keyboards are described in which an on-screen keyboard supports a plurality of quick tasks configured to edit selected text in defined ways. Quick task functions may be assigned to keys of the on-screen keyboard in addition to the default actions or “normal” character entry functions of the keys. During text input, the keys operate normally to perform a function related to character entry, such as to cause input of corresponding text character. When text is selected or in other designated interaction scenarios, operation of a key associated with a quick task may automatically trigger the quick task to modify selected text in a corresponding way. Thus, the same key of an on-screen keyboard may be employed to initiate a function related to character entry or apply of a quick task to text depending upon the interaction scenario.

Abstract: In a mobile device, the text entered by users is analyzed to determine a set of responses commonly entered by users into text applications such as SMS applications in response to received messages. This set of responses is used to provide suggested responses to a user for a currently received message in a soft input panel based on the text of the currently received message. The suggested responses are provided before any characters are provided by the user. After the user provides one or more characters, the suggested responses in the soft input panel are updated. The number of suggested responses displayed to the user in the soft input panel is limited to a total confidence value to reduce user distraction and to allow for easier selection. An undo feature for inadvertent selections of suggested responses is also provided.

Abstract: Techniques are described to generate text prediction candidates corresponding to detected text characters according to an adaptive language model that includes multiple individual language model dictionaries. Respective scoring data from the dictionaries is combined to select prediction candidates in different interaction scenarios. In an implementation, dictionaries corresponding to multiple different languages are combined to produce multi-lingual predictions. Predictions for different languages may be weighted proportionally according to relative usage by a user. Weights used to combine contributions from multiple dictionaries may also depend upon factors such as how recently a word is used, number of times used, and so forth. Further, the dictionaries may include interaction-specific dictionaries that are learned by monitoring a user's typing activity to adapt predictions to corresponding usage scenarios.

Abstract: Assistive buffer usage techniques are described. In one or more implementations, audio generated through text-to-speech conversion is output, the audio corresponding to text portioned in one of a plurality of buffers from an item of content. An input is received to rewind or fast forward the output of the audio. Responsive to the input, additional audio is output that was generated through text-to-speech conversion from text portioned in another one of the plurality of buffers.

Abstract: In a mobile device, a context is determined for the mobile device. The context is determined based on a variety of characteristics of the mobile device environment including, for example, the current application being used, any contacts that a user of the mobile device is interacting with or having a conversation with, the current date and/or time, a current topic of the conversation, a current style of the conversation, etc. Based on a set of strings associated with the determined context and user generated text, one or more string predictions are generated for the user generated text. The string predictions may be presented to the user as suggested completions of the user generated text.

Abstract: In a mobile device, the text entered by users is analyzed to determine a set of responses commonly entered by users into text applications such as SMS applications in response to received messages. This set of responses is used to provide suggested responses to a user for a currently received message in a soft input panel based on the text of the currently received message. The suggested responses are provided before any characters are provided by the user. After the user provides one or more characters, the suggested responses in the soft input panel are updated. The number of suggested responses displayed to the user in the soft input panel is limited to a total confidence value to reduce user distraction and to allow for easier selection. An undo feature for inadvertent selections of suggested responses is also provided.

Abstract: Assistive buffer usage techniques are described. In one or more implementations, audio generated through text-to-speech conversion is output, the audio corresponding to text portioned in one of a plurality of buffers from an item of content. An input is received to rewind or fast forward the output of the audio. Responsive to the input, additional audio is output that was generated through text-to-speech conversion from text portioned in another one of the plurality of buffers.