Transparent policies
Arctera Insight Classification introduces new category of policies called Transparent policies. Transparent policies provide visibility into the policy keywords and logic. These policies are fully transparent allowing users to view the exact conditions that trigger a hit during classification. Thus, they provide full control and defensibility over classification.
Users can granularly modify, add, or remove classification criteria within a transparent policy by simply creating a copy and then editing the copy.
Users can use this editable copy to:
-
Add nested conditions with \<Any of\>, \<None of\>, or \<All of\> criteria
-
Remove keywords that trigger too many false positives
-
Add keywords to reduce false negatives
-
Add new criteria such as sentiment score, patterns, language, entity (if supported by the application) or regular expression
-
Modify criteria to include min/max counts, change proximity criteria, specify string match, or case match
For more information on use cases via copied editable policies, See Creating policies.
For built-in policies belonging to the Transparent category, users can get following details about when the transparent policy logic was last updated by Arctera. Refer to the sample image below.
-
Last Updated: Month and year of the last update to the policy.
-
Version: A version of Arctera Insight Classification in which the policy is updated.
Related information