> For the complete documentation index, see [llms.txt](https://docs.tickertrends.io/tickertrends/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.tickertrends.io/tickertrends/documentation-v2/2.-data-coverage-and-data-types/2.2-raw-data-types/2.3-social-discussion-analysis/topics-sentiment.md).

# Topics Sentiment

**Overview**\
The Sentiment metric measures the tone of discussion around a keyword, brand, or topic across social and web data. It classifies posts as **positive, negative, or neutral**, and aggregates them into a distribution and a single **net sentiment score**. This helps quantify *how people feel*, not just how much they are talking.

**Historical Length**\
Varies by platform and data availability, but typically spans **multiple years of historical data** where supported. Allows users to track how perception evolves over time, especially around product launches, updates, or events.

**Granularity**\
Available at **daily aggregation**, with the ability to view sentiment snapshots tied to specific time windows or spikes in discussion.

**Update Frequency**\
Sentiment data is **updated daily**, reflecting the most recent shifts in tone as new posts and discussions are ingested.

**Methodology**\
Sentiment is derived using natural language processing models that analyze the tone and context of posts. Each post is classified into one of three categories:

* **Positive**: Favorable or supportive language
* **Negative**: Criticism, complaints, or dissatisfaction
* **Neutral**: Informational or non-opinionated mentions

These are then aggregated into:

* **Percentage breakdown** (Positive **P**/ Negative **N**/ Neutral **U**)
* **Net Sentiment Score** (range: -1 to +1)

The score is calculated by weighting positive vs. negative mentions, providing a directional view of overall sentiment.

**Use Cases**

* Identify **user frustration or product pain points** early
* Track **consumer perception shifts** after launches or updates
* Validate whether **rising discussion is positive or negative**
* Compare **brand perception vs. competitors**
* Support **earnings or demand narratives** with qualitative context

**Example Visualization**

<figure><img src="/files/SfVTIMGKvZj8HyNDz3Dj" alt=""><figcaption></figcaption></figure>

In the example above we can choose a point in time on the graph and it will populate the sentiment at that time (+- 15 days):

* **P: 45% | N: 9% | U: 45% | Score: +0.26**

This indicates:

* Majority of conversation is **positive**
* Overall tone is **skewing positive**
* Likely driven by the posts talking about kpop demon hunters and the world cup