Auto-tagging is a powerful way to discover new conversations and continually enrich your data.
With auto-tagging enabled, Frame will automatically apply the tag to conversations which match its trained model.
Frame's success team can assist you in training auto-tagging models. In general, here are some characteristics of tags that are ready for auto-tagging:
|What to Look For||What to Avoid|
|The tag comprises records which have language patterns directly observable in the messages themselves.||- The tag represents a higher-level intent or product category that cannot be inferred from the message text at all.|
- The tag represents the inbound channel (e.g., chat/email), not the contents of the message.
- The tag represents an aspect of the speaker's identity (e.g., "VIP"), not the contents of the message.
|The tag is specific enough to be associated with message text that is in some way similar across records.||The tag is a "catch-all" for any kind of activity, such as a tag called "Support".|
Updated about 1 year ago