| http://www.w3.org/ns/prov#value | - Referring also to FIG. 12, whether the user types or speaks the corrected text, the system builds (step 198) a tree structure (e.g., 200) for each of the last three utterances using the speech frames corresponding to these utterances and the speech frames (if spoken) or speech models (if typed) corresponding to the corrected text.
|
