| CPC G10L 15/01 (2013.01) [G10L 15/26 (2013.01); G10L 2015/225 (2013.01)] | 20 Claims |
|
1. A cellular phone configured to perform automated text correction using correction probability scores, the cellular phone comprising:
a touch-sensitive display configured to display text and detect touch input;
a microphone configured to receive voice input provided to the microphone;
one or more processors; and
one or more computer-readable media storing instructions that, when executed, cause at least one of the one or more processors to perform operations comprising:
receiving voice input of one or more sentences spoken by a user and captured by the microphone;
converting the received voice input of the one or more sentences to text data representing the one or more sentences;
causing display, on the touch-sensitive display, of the text data representing the one or more sentences converted from the received voice input;
subsequent to display, on the touch-sensitive display, of the text data representing the one or more sentences, receiving a text correction of a first portion of the text data representing the one or more sentences, the text correction being entered by the user through touch input detected by the touch-sensitive display and resulting in a change from the first portion of the text data to corrected text within the one or more sentences displayed on the touch-sensitive display;
based on receipt of the text correction of the first portion of the text data, analyzing the text data representing the one or more sentences to assess similarity of the first portion of the text data to other portions of the text data representing the one or more sentences by:
comparing the first portion of the text data to the other portions of text data to assess similarity of the first portion of the text data to each of the other portions of text data,
reevaluating the voice input with the text correction as an input considered in conversion of the voice input corresponding to the other portions of the text data,
evaluating context of the text correction in assessing similarity of the first portion of the text data to other portions of the text data, and
determining a correction probability score based on the comparison of the first portion of the text data to the other portions of text data, the reevaluation of the voice input with the text correction as an input, and the evaluation of the context;
accessing, from electronic storage, data relating correction actions to ranges of correction probability scores;
comparing the determined correction probability score to the ranges of correction probability scores;
selecting a correction action based on the comparison of the determined correction probability score to the ranges of correction probability scores; and
performing the selected correction action.
|