SocialPulse Documentation & Matrix Science

SocialPulse is not a black box. This manual details exactly how our analytical matrices are calculated, how you can customize your dynamic dashboards, and how to define your brand's unique ontology.

1. Analytical Matrices (Metrics)

The platform calculates several core matrices in real-time. Understanding these calculations is vital to interpreting your dashboard correctly.

1.1 Global Sentiment Score

The Global Sentiment Score is a weighted average of all post polarities mapped against your active ontology over a rolling 24-hour window.

Mathematical Calculation

Score = (Σ(w_i * p_i) / Σw_i) * 100

Where:
w_i = Influence weight of the author (0.1 to 1.0)
p_i = NLP polarity of the post (-1.0 to 1.0)

How to read this conclusion: If your score drops below 40%, the platform's color-coding shifts from Green to Yellow. This indicates a general malaise in brand perception, likely requiring a marketing intervention rather than immediate crisis PR.

1.2 Crisis Velocity Matrix

Velocity measures momentum. It calculates how quickly negative sentiment is spreading relative to normal baseline volume.

How to read this conclusion: The red "Active Crisis" cards on the dashboard appear only when Velocity crosses 2.5 standard deviations above the 30-day mean. A high velocity means you have hours, not days, to respond.

2. Dashboard Customizations & Views

No two analysts work the same way. The SocialPulse dashboard is fully modular.

Drag-and-Drop Widgets

Every graph, table, and matrix on the Overview page is a widget. Click the grid icon in the top right of any widget to enter "Edit Mode". You can drag widgets to reorder them or resize them to emphasize specific data points.

Save layouts as "Personal Views".

Set a specific View as your default landing page.

Tenant Admins can push a layout to all users as a "Tenant Default".

System DefaultMy WorkspaceEdit mode

Sentiment by platform

Crisis velocity

Volume

12.4k

Mentions

3,982

Drag widgets from the library, resize with handles, and the canvas reflows around them — the System Default stays locked as an always-available fallback.

3. Ontology Filters

Did you know? In Sprint 9, we introduced Ontology Versioning. You can instantly switch between different sets of keywords and sentiment rules to see how they affect historical data.

The global filter bar at the top of the application allows you to slice data by Ontology Nodes. For example, if you are a Telecom, you can filter the entire dashboard to only show posts related to the "Network Outages" node.