> ## Documentation Index
> Fetch the complete documentation index at: https://collony.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Sentiment

> Track community mood over time - trend charts, user-level scores, and filtered message views.

The Sentiment page gives you a detailed view of community mood across the selected time range. It breaks down the positive/neutral/negative split, shows per-user scores, and lets you filter analyzed messages by sentiment category.

<Frame>
  <img src="https://mintcdn.com/collonyai/XQ-9_MUnEZncUbcz/images/sentiment-1.png?fit=max&auto=format&n=XQ-9_MUnEZncUbcz&q=85&s=630e96e77ac297370c9f256b1f46b896" alt="Sentiment 1" width="3456" height="1930" data-path="images/sentiment-1.png" />
</Frame>

## Sentiment summary

Four stats at the top:

* **Positive** - percentage and message count
* **Neutral** - percentage and message count
* **Negative** - percentage and message count
* **Total Analyzed** - messages scored in this period

## Daily Sentiment Trend

A chart showing how the positive/neutral/negative split shifts day by day. Use this to correlate sentiment drops with specific events - a product announcement, a market move, a moderation incident - and understand how quickly the community recovers.

## Sentiment Distribution

A breakdown showing the proportion of each sentiment category across the period.

## User-level sentiment

Two ranked tables:

**Most Positive Users** - members with the highest average sentiment score in the last 7 days. Useful for identifying engaged, constructive contributors worth recognizing or promoting.

**Most Negative Users** - members with the lowest average sentiment scores. Persistent negativity from the same accounts often precedes escalating behavior - worth monitoring before it becomes an incident.

## Recent Analyzed Messages

A filterable feed of scored messages. Filter by: All, Positive, Negative, Questions, Swearing, High Risk, Sarcastic, Shilling, or FUD. Click any message to expand the full sentiment breakdown.

<Tip>
  Filter by Shilling or FUD and cross-reference with your Incidents page during market volatility. Coordinated FUD campaigns often appear in sentiment data before they trigger moderation events.
</Tip>