Configuring Autocomplete Dictionary Outputs

Usually, the autocomplete dictionary output does not need to be modified. However, if your autocomplete phrases contain non-segmented languages (such as, Japanese or Thai), the user may be trying to query a word from the middle of a phrase. If this phrase has not been output segmented in the autocomplete dictionary using a method other than by spaces (the default which doesn't apply here), the autocomplete will not return the suggestion.

To add an output segmenter, select the Outputs tab of your autocomplete dictionary, and click edit on the Autocomplete Collection output. Then enter <vse-index-stream segmenter="name of the segmenter" />, replacing name of the segmenter with the segmenter that you wish to use. If necessary, the output can be further modified to include advanced index stream options for linguistic configuration. By clicking the edit button, you can overwrite the default vse-index-stream node. For more information see Index Streams.

In most cases, adding other outputs to an autocomplete dictionary would result in the production of dictionaries (wildcard, segmenter, etc.) that are of little practical use.

Tip: By adding a spelling dictionary output to your autocomplete dictionary, a list of words (in this case phrases) is added to the dictionary tabs when the autocomplete dictionary is built. This can be useful for identifying the phrases and related counts that were collected by the autocomplete dictionary without having to query the project.

Once all settings, input sources, and outputs have been configured for an autocomplete dictionary, go to the overview tab of your dictionary and press the build button to build the autocomplete dictionary.