Select the text fields to analyze and which indicators to analyze.
Text analytics relies on specific steps in the data
collection jobs. The
Collect parameter in the relevant jobs must
be set to collect text analytics. A job is relevant if it collects data for the same
indicators that are subject to text analysis. See
Create or schedule a data collection job.
Role required: pa_power_user or
pa_admin
-
Navigate to .
- To create a new text analytics configuration, click
New.
- To edit an existing text analytics configuration, click the information
icon in that row and select Open record. Clicking the
indicator source name opens the record for that indicator source, not the
text analytics configuration.
-
Select the Indicator source whose input text you want to
analyze.
-
Unlock Fields to analyze and select the fields whose
input text you want to analyze.
You must select at least one field. Usually you want to include the
Short description field.
- (Optional)
Select Use system stop words to include the Zing stop
words.
Selected by default. The Zing global stop words apply to the indicator source.
If you do not use the Zing stop words, you must select all stop words
yourself.
-
Save the text index configuration.
The Indicators related list appears at the bottom of the
form.
-
In the Indicators list, click
Edit and select the indicators to analyze from the
slushbucket.
-
Run a relevant historical data collection job to collect initial data.
You may be able to run a job from this form, or you may have to run a
historical data collection job. See the next topic for details.
