background image
C H A P T E R 9
About the Recognition System
When the recognition system returns a word unit to a view based on the
classes, the view displays or uses the best
interpretation of the original input strokes. Paragraph views display words directly;
edit views create a
child automatically to display text that the
recognition system returns. Additionally, the recognition system constructs a
correction information frame from the word unit and saves learning data as
appropriate. For more information, see "Correction and Learning" (page 9-13)
and "Accessing Correction Information"(page 10-23). Your view can provide an
method to perform application-specific processing of
the word unit.
The set of possible interpretations that the text recognizer returns to a view is
affected by
the text recognizer that the view uses to interpret the input strokes
options you have specified for the text recognizer in use
the dictionaries that are available to the view for recognition use
A dictionary is a system construct against which the user's input strings are
matched, as a means of ensuring the validity of the text recognizer's output. The
system supplies dictionaries that define names, places, dates, times, phone numbers,
and commonly used words to the text recognizers. The user can expand the
system's built-in vocabulary by adding new words to a RAM-based user dictionary
accessed from the Personal Word List slip. In addition, you can provide custom
dictionaries for the recognition system's use. For example, you might create a
custom dictionary to supply specialized vocabulary, such as medical or legal
terminology. The section "System Dictionaries" beginning on page 9-11 describes
the system-supplied dictionaries in more detail. The use of custom dictionaries for
recognition is described in "Using Custom Dictionaries" beginning on page 10-24.
Although the interpretations returned by the printed recognizer are never limited to
dictionary words, its output is influenced strongly by the set of dictionaries
available for its use. The interpretations returned by the cursive recognizer can be
restricted to those words appearing in the set of dictionaries available for its use;
however its default behavior is to return non-dictionary words in addition to words
appearing in available dictionaries.
Options specified for the currently enabled recognizer may also influence the
interpretations it returns to the view. For example, the cursive recognizer's default
settings enable its letter-by-letter recognition option, to increase the likelihood of
its returning strings not in the currently available set of dictionaries. The user can
control this option and others from within the Handwriting Settings preferences slip.
Note that even when the cursive and printed recognizers are configured similarly,
the results they return for the same input may differ. For example, using the cursive
© 2007-2017, o7 studio » при воспроизведении материала сайта ссылка обязательна