> For the complete documentation index, see [llms.txt](https://dots.gitbook.io/dots-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://dots.gitbook.io/dots-docs/analysis/highlights-and-its-types.md).

# Highlights & its types

### **What are Highlights?**

Highlights are tagged excerpts - sentences, phrases, or words - within large bodies of text such as interviews, reports, and survey responses. Each highlight can be associated with both thematic tags and attribute tags as defined in your Themes Manager, making it easier to organize and analyze qualitative data.  

Once created, the platform provides a centralized view to explore the highlights by thematic and attribute tags, geography, time and more.

**Types of Highlights**

Highlights can be created in three ways: manually, through AI-assisted workflows where the AI suggests and you approve, or through auto highlights where the AI tags entire documents on its own. 

<table><thead><tr><th width="268">Method</th><th>Best for</th></tr></thead><tbody><tr><td>Manual Highlights</td><td>Precise coding and complete control</td></tr><tr><td>AI-Assisted Highlights</td><td>Faster coding with human review</td></tr><tr><td>Auto Highlights</td><td>Large-scale annotation across many documents</td></tr></tbody></table>

### **Manual Highlights**

To create a highlight manually, follow these steps:

{% stepper %}
{% step %}
Open the published page of the document and scroll to the relevant text.
{% endstep %}

{% step %}
Select the text you want to annotate. An annotate button will appear above the selected text.
{% endstep %}

{% step %}
Click *Annotate* to open the annotation panel and select the appropriate tags from your Thematic and Attribute tabs.
{% endstep %}

{% step %}
Once you have chosen the relevant tags, click the apply button to save the highlight.
{% endstep %}
{% endstepper %}

<figure><img src="/files/PNycjAr8CKHIsSEaanfi" alt="" width="533"><figcaption></figcaption></figure>

### **AI - Assisted Highlights**

Next, let’s see how to highlight using AI suggestions:

{% stepper %}
{% step %}
Open the published page of the document
{% endstep %}

{% step %}
Click the options button (three dots icon) and select *Review*

<figure><img src="/files/OCrkSO3sUd19gbUWXv2N" alt="" width="563"><figcaption></figcaption></figure>
{% endstep %}

{% step %}
Scroll to the text field where you want AI suggestions and click on the text. List of options show up at the bottom of the screen.&#x20;
{% endstep %}

{% step %}
Click the magic wand icon to generate highlight suggestions. Each suggestion displays the excerpt along with the suggested thematic and attribute tags.
{% endstep %}

{% step %}
Review the suggestions
{% endstep %}

{% step %}
Click *Apply* to create the highlight or *Dismiss* to reject the suggestion.
{% endstep %}
{% endstepper %}

<figure><img src="/files/AME0xbjeYAucfpBR7fNq" alt="" width="563"><figcaption></figcaption></figure>

### **Auto Annotations**

Auto highlights allow AI to tag entire documents without requiring you to review each suggestion individually. This approach is best suited for scaling up i.e. when you have already validated the structure of your Themes Manager and want to apply it across a larger dataset without having to go document by document.

Auto highlights use the tags from your Themes Manager and runs in two phases:&#x20;

1. First, it applies **thematic tags** wherever they are relevant.
2. Next, it applies **attribute tags** only to the excerpts that have already received a thematic tag.

This ensures every auto-tagged excerpt always carries both a subject (thematic) and a data cue (attribute).

You can run auto highlights directly from the document listing page:

{% stepper %}
{% step %}
Go to the **document listing page** and select the documents you want tagged. You can pick one, many, or all.
{% endstep %}

{% step %}
Click the ***Annotate*** button. The selected documents are added to the auto-annotation queue.
{% endstep %}

{% step %}
Wait while the AI processes them. **Auto highlighting takes some time**, depending on the number of documents in the queue and the length of each document. You can continue working elsewhere on the platform while the annotation process runs.
{% endstep %}
{% endstepper %}

<figure><img src="/files/MPkPDyOT7o5VzVulMO7o" alt="" width="563"><figcaption></figcaption></figure>

Once a document is annotated, the highlights appear on it just like manual or AI-assisted highlights , with the same tags from your Themes Manager.&#x20;

During the process, the AI can also suggest new themes and tags that emerge from the data. These inductive suggestions may help surface findings that may not be captured by your existing theme structure.

<br>