Language identification using recurrent neural networks

A method, comprising:

• at an electronic device:
• receiving a first typed character of a character sequence;
• providing the first typed character to a neural network;
• determining, using the neural network, a first character context of the first typed character based on the first typed character and a second typed character of the character sequence;
• determining, using the neural network, a first confidence level that the character sequence is associated with a language of a plurality of languages based on the first character context of the first typed character;
• determining whether the first confidence level exceeds a predetermined threshold;
• in accordance with a determination that the first confidence level exceeds the predetermined threshold, providing the language as a candidate language;
• in accordance with a determination that the first confidence level does not exceed the predetermined threshold, forgoing providing the language as the candidate language;
• determining whether a third typed character has been provided to the electronic device;
• in accordance with a determination that a third typed character has been provided to the electronic device:
• providing the third typed character to the neural network;
• determining, using the neural network, a second character context of the first typed character based on the first typed character and the third typed character;
• determining, using the neural network, a second confidence level that the character sequence is associated with the language of the plurality of languages based on the second character context of the first typed character;
• determining whether the second confidence level exceeds the predetermined threshold;
• and
• in accordance with a determination that the second confidence level exceeds the predetermined threshold, providing the language as the candidate language.

An electronic device, comprising:

• one or more processors;
• a memory;
• and
• one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for:
• receiving a first typed character of a character sequence;
• providing the first typed character to a neural network;
• determining, using the neural network, a first character context of the first typed character based on the first typed character and a second typed character of the character sequence;
• determining, using the neural network, a first confidence level that the character sequence is associated with a language of a plurality of languages based on the first character context of the first typed character;
• determining whether the first confidence level exceeds a predetermined threshold;
• in accordance with a determination that the first confidence level exceeds the predetermined threshold, providing the language as a candidate language;
• in accordance with a determination that the first confidence level does not exceed the predetermined threshold, forgoing providing the language as the candidate language;
• determining whether a third typed character has been provided to the electronic device;
• in accordance with a determination that a third typed character has been provided to the electronic device:
• providing the third typed character to the neural network;
• determining, using the neural network, a second character context of the first typed character based on the first typed character and the third typed character;
• determining, using the neural network, a second confidence level that the character sequence is associated with the language of the plurality of languages based on the second character context of the first typed character;
• determining whether the second confidence level exceeds the predetermined threshold;
• and
• in accordance with a determination that the second confidence level exceeds the predetermined threshold, providing the language as the candidate language.