# Sentiment Analysis

## 😄 What is sentiment analysis

Using sentiment analysis you can generate a sentiment based on a natural free text entered by the respondent in one of your text based answers like fields for example. This way you can have a insight of the respondent's sentiment about a particular topic without having to to use a closed question to get his feedback.

## 🚀 Enabling sentiment analysis 

You can enable sentiment analysis on any [field answer type](https://docs.ngsurvey.com/form-management/form-designer/questions/question-types/standard/text-comment-field) of your questions surveys using the **Compute sentiment** property of your answer.

![](https://1025048312-files.gitbook.io/~/files/v0/b/gitbook-legacy-files/o/assets%2F-M8fLhS0bmfBRyq0HdUm%2F-MBZoNglGULJjSsdBXtQ%2F-MBZwz17edcghYW5t8GD%2Fimage.png?alt=media\&token=02a57d2e-7b5f-43f8-b4a6-a48dc8fa7a77)

{% hint style="warning" %}
Enabling the **Compute sentiment** will retroactively rate all the pre-existing respondent answers that were not rated yet.
{% endhint %}

Once you have enabled the sentiment on the answer ngSurvey will analyse all respondent answers and rate them according to a sentiment. An answer can be rated with following sentiments.

* Positive
* Neutral
* Negative
* Mixed

For example if a respondent would have been rated as Positive if the respondent entered following text.

![](https://1025048312-files.gitbook.io/~/files/v0/b/gitbook-legacy-files/o/assets%2F-M8fLhS0bmfBRyq0HdUm%2F-MBZoNglGULJjSsdBXtQ%2F-MBZxlyVgD0yrvz2IDrF%2Fimage.png?alt=media\&token=68240de8-a67b-4b49-ba7e-20d334756f94)

## 📊 Sentiments reporting

Once the sentiments have been calculated you may analyze the results in 2 different ways.

Either using the [text reports](https://docs.ngsurvey.com/form-management/reports/text-reports) which will displays "smileys" next to each response to identify the sentiment of the respondent.

![](https://1025048312-files.gitbook.io/~/files/v0/b/gitbook-legacy-files/o/assets%2F-M8fLhS0bmfBRyq0HdUm%2F-MBZzmnC3mGNzw6rB_Rk%2F-MB_-8HRLLR25mCAlW0r%2Fimage.png?alt=media\&token=e0b6564b-75f3-4455-93e9-6e49eb5394cf)

You can filter out the answers by a given sentiment category using the sentiment filter buttons

![](https://1025048312-files.gitbook.io/~/files/v0/b/gitbook-legacy-files/o/assets%2F-M8fLhS0bmfBRyq0HdUm%2F-MBZzmnC3mGNzw6rB_Rk%2F-MB_-M12Uieqrf3Sgwy-%2Fimage.png?alt=media\&token=82491051-5b86-4708-b934-5880782cc312)

You may also use the [Sentiment index report](https://docs.ngsurvey.com/form-management/reports/report-builder/report-items/sentiment-index) to display a [sentiment index](https://docs.ngsurvey.com/form-management/reports/report-builder/report-items/sentiment-index) ranging from -10 (very negative) to +10 (very positive).

![](https://1025048312-files.gitbook.io/~/files/v0/b/gitbook-legacy-files/o/assets%2F-M8fLhS0bmfBRyq0HdUm%2F-MBZybmuKsoKj9qS_uRL%2F-MBZzgsulR1szcr7E39b%2Fimage.png?alt=media\&token=1add2269-4bbf-4ec2-abb5-6fd4b369b382)

{% hint style="info" %}
The last 25 sentiments of all the enabled sentiment answer is also displayed by default in the [overview dashboard](https://docs.ngsurvey.com/form-management/overview-reports).
{% endhint %}

{% hint style="warning" %}
Sentiment analysis requires that you have linked ngSurvey [third party API settings](https://docs.ngsurvey.com/installation-setup/system-settings/third-party-apis/microsoft) to the Microsoft Text Analysis API or to enable your [AI Suite](https://docs.ngsurvey.com/ai-suite).
{% endhint %}