Different search mode: podcast alignment with topic #3

Open
opened 2026-04-04 22:35:47 +00:00 by taf · 0 comments
Owner

Process all chunks in all feeds against query embedding and return the average score? Slow, obviously. Alternative might be to take averages of embedding vectors on an episode and/or feed basis?

Could even make a cool viz, bubble-size controlled at an episode level by average distance to query of chunks in that episode, bubbles within same feed are grouped together into feed-bubbles, and these will have a visual size that corresponds to the relatedness of that feed's content with the query embedding

Process all chunks in all feeds against query embedding and return the average score? Slow, obviously. Alternative might be to take averages of embedding vectors on an episode and/or feed basis? Could even make a cool viz, bubble-size controlled at an episode level by average distance to query of chunks in that episode, bubbles within same feed are grouped together into feed-bubbles, and these will have a visual size that corresponds to the relatedness of that feed's content with the query embedding
Sign in to join this conversation.
No labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
taf/podcast_search#3
No description provided.