| CPC G09B 17/003 (2013.01) [G06F 40/109 (2020.01); G10L 15/063 (2013.01); G10L 15/197 (2013.01); G10L 15/22 (2013.01)] | 20 Claims |

|
1. A method, comprising:
performing speech recognition on recorded audio of a reader reading aloud a displayed first set of words;
training a machine learning language model based on the displayed first set of words and words recognized from the first set of words by the speech recognition;
obtaining a second set of words to be displayed to the reader;
predicting, using the machine learning language model, one or more words from the second set of words expected to be difficult for the reader by inputting the second set of words to the machine learning language model; and
providing the second set of words for display to the reader, wherein the one or more words of the second set of words predicted to be difficult for the reader are displayed differently from other words of the second set of words that were not predicted to be difficult for the reader, wherein at least one word from the one or more words in the second set of words is different from the one or more words in the first set of words.
|