Intelligence_Dictionary v.1.0_MANUAL
A structured breakdown of the mathematical and conceptual primitives powering TrendForge, including
keywords and concepts.
This is not marketing — this is the system language.
Vector_Space
In TrendForge, every article, social post, and news headline is converted into a Vector—a list of 1,536 unique coordinates. This mathematical representation allows us to plot ideas in a high-dimensional space.
[0.012, -0.045, 0.882, 0.124, -0.331,
...]
Ideas that are semantically similar "live" near each other in this space, regardless of the specific keywords used.
pgvector
pgvector enables PostgreSQL to store and query high-dimensional vectors. This is what makes semantic search and clustering possible at scale.
Instead of querying text, we query meaning using cosine distance operations directly in SQL.
Embeddings
Embeddings transform raw text into vectors. Instead of storing words, we store meaning.
- • Captures semantic similarity across different wording
- • Enables clustering without keywords
- • Powers all downstream intelligence
DBSCAN
DBSCAN is the clustering algorithm that groups similar articles together into Topic Clusters.
It automatically detects clusters of any shape and ignores noise — making it ideal for real-world content streams.
Centroids
A Centroid is the mathematical "Average" of a topic cluster. It represents the core consensus of a story.
Centroids are updated in real-time as new articles join a cluster. When a new article is too far from the centroid, it signals a Narrative Shift.
Cosine_Similarity
Cosine Similarity measures how close two vectors are — not by distance, but by direction.
Smaller angles = higher similarity. This is how we determine if two pieces of content belong to the same narrative.
Narrative_Arbitrage
The practice of identifying and exploiting the "Time Lag" between niche social buzz (Reddit/Discord) and mainstream media (Google News).
"Arbitrage exists when a topic's Social Velocity exceeds its News Volume by more than 40%. This is the window where you can publish and 'own' the narrative before the competition arrives."
Alpha_Angle
The Alpha Angle is the optimal narrative position — where originality, timing, and momentum intersect.
- • High originality (low competition)
- • Rising momentum (early timing)
- • Narrative gap (market inefficiency)
Signal_Convergence
Signal Convergence is the Definitive Proof. It occurs when independent data streams—social velocity, search intent, and sentiment vectors—intersect at a single coordinate, validating that a trend is no longer speculative.
Detected when multi-nodal correlation exceeds a critical confidence threshold. This is the "Bullseye" state where noise is filtered out and the strategic path becomes an execution directive.
Market Signals Explained
The Market Signal Feed is a stateful anomaly detection engine. Unlike standard aggregators, the Neural Inbox monitors the decoupling of Public Narrative from Ground Truth by identifying mathematical deviations in sentiment, volume, and semantic direction.
We track a number of different types of Market Signals. By differentiating the signal types, it allows you to adapt your strategy to different market conditions. Instead of just saying, here is a gap, we offer a more detailed analysis of what TYPE of gap and how you should respond.
Magnitude Promotion
24-Hour Cluster Lock prevents signal fatigue. Detections promote existing signal magnitude rather than duplicating alerts.
Neural Scrutiny
Thresholds governed by Engine Profiles: STEALTH (Low), BALANCED (Med), or ALPHA BURST (High-Sensitivity).
Forensic Telemetry
Calculates the "Recycling Rate" vs "Originality Index" to separate market noise from strategic opportunity.
Narrative_Shift
Detects a fundamental change in Semantic Direction. It identifies the exact moment an "old story" is replaced by a "new reality," usually triggered by a significant sentiment inversion.
TRIGGER: Cosine Distance > Profile Threshold. Measures the delta between embeddings at T-24H and T-Now to detect pivots before headlines reflect them.
Black_Swan
A Systemic Rupture. This is not a trend, but a statistical anomaly that shatters the historical baseline, violating all standard probability bounds.
TRIGGER: 3-Sigma Violation (σ). Detects when volume or sentiment punches vertically through the 99.7% probability ceiling of the historical baseline.
Blue_Ocean
Identifies Market White Space. High-velocity social interest combined with zero professional media competition—an explosive supply/demand mismatch.
TRIGGER: Scarcity Ratio > 400:1. Detects explosive social "Whispers" and niche demand before a single major news outlet generates professional coverage.
Arbitrage_Alert
Identifies the Truth Decoupling. Triggers when the mainstream "Narrative Hype" has completely detached from verified, hard data points.
TRIGGER: Delta Comparison. Measures the displacement between the Narrative_Weight and the Truth_Index index to find exploitable alpha gaps.
Consensus_Vacuum
Detects Extreme Polarization. High intensity at opposing extremes leaves a vacant middle ground—creating an opening for a factual bridge or a neutral narrative.
TRIGGER: Bimodal Distribution. Identifies near-zero volume in the center of the sentiment spectrum despite high-velocity engagement on the "tribal" edges.
Category_Surge
The Rising Tide. Triggers when an entire vertical—such as Biotech or AI—accelerates simultaneously across 60% or more of its sub-topic clusters.
TRIGGER: Synchronization Rate > 60%. Maps momentum across 25 sub-topic nodes; illuminates as clusters synchronize into a singular cross-vertical push.
Pos / Neg Sentiment
Sentiment analysis is the process of determining whether a piece of writing is positive, negative, or neutral. It's a key component in understanding the emotional tone behind a narrative.
Semantic Distance
Semantic distance is a measure of how related two concepts or terms are. In the context of TrendForge, it helps us understand how closely different narratives are related, even if they don't share keywords.
Cosine Similarity
Cosine similarity is a metric used to measure how similar two vectors are. In our system, it's used to compare the vector representations of documents to determine how similar their content is.
Sigma Deviation (σ)
Sigma deviation, or standard deviation, is a measure of the amount of variation or dispersion of a set of values. We use it to identify anomalies and significant shifts in data trends.
Bimodal Distribution
A bimodal distribution is a continuous probability distribution with two different modes. This often indicates that there are two different groups or opinions within a dataset, which is a key signal for narrative divergence.
Originality Index
The Originality Index is a proprietary metric that scores content based on its uniqueness. It helps to distinguish between recycled news and genuinely new information or perspectives.
Cluster Magnitude
Cluster Magnitude refers to the size and density of a topic cluster. It helps us quantify the scale and intensity of a narrative, providing a clear indicator of its market impact.