Proposal: Capture latent author interest from dwell signals #16051
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Problem ------- Engagement-based ranking under-serves "silent" users who read content but do not explicitly like, reply, or retweet. The system already logs and labels multiple high-intent dwell signals (tweet detail, profile, link, fullscreen video), and aggregates them in real time, but does not expose a persistent user preference representation derived from those signals. As a result, latent interest (interest without explicit engagement) is not captured as a first-class signal. Proposed solution ----------------- Introduce online, decayed latent-interest aggregates derived from existing dwell labels, keyed by (userId, sourceAuthorId), to represent implicit long-term interest in authors. This change: - Defines AuthorLatentInterestEngagements using high-intent dwell signals (profile dwell, tweet detail dwell, long link dwell) - Adds authorLatentInterestRealTimeAggregates with exponential decay - Registers the aggregate group in ProdAggregateGroups - Ensures outputs are not filtered via aggregates_to_drop.txt - Mirrors existing user+author aggregate flag defaults (e.g. includeAnyFeature) The change is additive and does not alter existing ranking or scoring behavior. It provides a foundation for future use in scoring, mixing, or exploration to better serve silent readers. Future work (not included): --------------------------- - Consume this aggregate in scoring or mixing - Tune decay/weights via params - Extend to topic/entity latent interest
|
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Improvement proposal: Latent Interest Aggregates from Dwell Signals
Problem
Engagement-based ranking under-serves "silent" users who read content but do not explicitly like, reply, or retweet. The system already logs and labels multiple high-intent dwell signals (tweet detail, profile, link, fullscreen video), and aggregates them in real time, but does not expose a persistent user preference representation derived from those signals.
As a result, latent interest (interest without explicit engagement) is not captured as a first-class signal.
Proposed solution
Introduce online, decayed latent-interest aggregates derived from existing dwell labels, keyed by (userId, sourceAuthorId), to represent implicit long-term interest in authors.
This change:
The change is additive and does not alter existing ranking or scoring behavior. It provides a foundation for future use in scoring, mixing, or exploration to better serve silent readers.
Future work (not included):