Knowledge Base Content Insights from AI Answers

The AI Answers feature on the search results page is heavily used in our community. I would like to see if deliver insights and observations to us in Control based on user behavior and activity on the front-end. 

Particularly things like: 

• Suggested improvements to KB section/category organization based on how users search for and interact with topics, such as: 
  • Tagging and labeling that aligns with search behavior
  • Categories and sections that align with how community members search for content, ask questions, phrase their replies, etc
• Identified gaps in KB topics based on community topics, replies, search behavior, and KB topic performance (and suggestions for closing those gaps)
• Identified high performing community topics that seem to be filling in gaps in the KB (and suggestions for closing those gaps in the KB and promoting the community content)
• Identifying miscategorized, duplicative, redundant, or conflicting content in the KB and the Community
• Prompt refinement, aka do users continue to edit or reword their search prompt until they find topics that they want to interact with, how much, how do the prompts change, etc
• Case deflection: analytics on searches that have a low deflection rate, i.e. a user enters a search prompt, investigates a few topics, but ultimately hits the “Contact Support” link to (presumably) submit a ticket