Classifiers
Classifier agents classify data from your corpus into categories. Classified documents will be tagged with the classification they receive, allowing you to filter by generated classifications. Currently, our platform supports Topic, Flesch Reading Ease, Language, Passive Voice, and Sentiment classification. All classifiers can be run on a schedule, keeping your documents' classifications up-to-date automatically.
Common configuration
Because all classifiers have the same general goal (to classify your documents), their configuration is largely the same. Some classifiers have additional settings, but all classifiers have the following configuration options:
Filter
The filter you choose determines what content from your corpus will be classified.
Tagging threshold
A 0-100 threshold for applying a tag to a document when classifying.
Allowed tags
Specify which tags can be used for classification. If a tag passes the threshold but is not in the list of allowed tags, it will be ignored. By default, all tags are allowed.
Maximum tags to apply
Limits the number of tags that can be applied when classifying. For example, if 3 tags reach the tag threshold, but the maximum tags is set to 2, only the top 2 tags will be applied. Sensible defaults are provided for each classifier agent, so you likely won't need to change them unless you have a special use case.
Remove previous classifications
If enabled, the classifier will remove all previous classification tags when tagging a document.
Generate reports
If enabled, this agent will generate a .csv report detailing the classification scores and what tags were applied for each classified document.
Include text in report
If enabled, the text of classified documents will be included in generated reports. Report generation must be enabled first.
Last updated