Text to Words

This phase breaks down sentences and paragraphs into a sequence of words and punctuation, where the punctuation is used to denote pauses and other prosody information. If a punctuation character is not used as punctuation, it is converted to its word form.

This phase will also perform part-of-speech tagging, disambiguation of capitonyms (words that are pronounced differently when capitalized, like “My Polish neighbour bought some polish at the store.”) and other processing that affects the words spoken for the document text.

References

  1. Numbers.