FollowGraph
Module for recording on-chain user-to-user follow relationships. Powers feed routing, curation visibility, and AI-based personalization.
🧩 Purpose
The FollowGraph contract manages the social layer of the TRN platform. Every time a user follows another, the relationship is recorded on-chain for durability, transparency, and to enable deterministic behavior for content routing.
It plays a critical role in:
Determining what shows in a user's follow feed
Routing retrned posts from curators to their audience
Enhancing AI targeting for content and ad delivery
Supporting resonance weighting based on user relationships
⚙️ Core Logic
The FollowGraph contract uses on-chain mappings to store all follow relationships:
solidityCopyEditmapping(address => mapping(address => bool)) public isFollowing;
mapping(address => address[]) public followers;
mapping(address => address[]) public following;These mappings enable constant-time queries like:
Is user A following user B?
Who follows user A?
Who is user A following?
The contract is append-only for followers, with removals (unfollows) tracked via a separate flag or nullification to maintain event integrity and indexing compatibility.
All write operations are recorded on-chain, but bulk read queries are offloaded to the indexer for efficient personalization and feed hydration.
✅ Key Rules
Users may follow/unfollow any other public user
FollowGraph
Follow events are immutable and recorded as events
FollowGraph
A retrned post shows up in all followers’ feeds
RetrnIndex + FG
Content visibility is routed based on these relationships
Feed System
AI uses these relationships to cluster resonance and targeting
AI Targeting
🧠 AI & Personalization Integration
The AI distribution logic uses FollowGraph in tandem with:
User embeddings (semantic interest vectors)
Retrn history
Boosting metadata
Flag history
To determine what content a user should see next, especially when:
Viewing trending posts
Receiving AI-recommended boosted posts
Being selected for ad impressions
Follow relationships weight relevance but are not absolute determinants — content still rises or falls based on real-time engagement and resonance.
🔁 Integration with Retrns
When a user retrns a post, the retrn is indexed in the RetrnIndex, and their followers (from FollowGraph) are notified and shown the retrned post in their feeds.
The system ensures a user cannot retrn their own post
Only one retrn per user per post per day is allowed
The retrn points back to the original post, not the boosted wrapper
This enables a non-hierarchical social curation system where value is distributed based on trust and resonance, not follower count alone.
🚫 Abuse Prevention
While the platform allows open follows, AI bot detection and trust-weighted moderation infrastructure monitor for patterns such as:
Sudden bot-driven mass follows
Engagement pods
Fake influencer clusters
If detected, these patterns can be suppressed algorithmically in feeds or moderation tools, though the raw follow mappings remain public and unaltered.
🔗 Related Modules
RetrnIndex
Records retrns, tied to follower routing via FollowGraph
BoostingModule
Distributes campaign posts, enhanced by social graph relevance
AI Targeting
Uses social graph to improve recommendation precision
FlagEscalator
May incorporate social trust metrics in future extensions
GeoOracle
Enforces regional blocks after social mapping
ModerationLog
Ensures users don’t see flagged or burned content, regardless of follows
🔒 Privacy
All follow relationships are public and on-chain. This transparency:
Enables open ecosystem innovation (analytics, alternate frontends)
Prevents shadowbanning or algorithmic manipulation
Supports social accountability for curated content
Last updated