How to use the opinion mining tool in ATLAS.ti

Who this article is for

This article is for ATLAS.ti Desktop users who want to use the Opinion Mining tool to identify sentiments, explore positive and negative opinions in textual data, and accelerate coding and thematic analysis workflows.

What is the Opinion Mining tool?

Opinion Mining is an AI-powered analysis tool in ATLAS.ti Desktop that identifies sentiments in textual data.

The tool analyzes the data and identifies:

• positive sentiments
• negative sentiments
• keywords
• phrases related to opinions
• quotations connected to those sentiments

Opinion Mining can help researchers:

• identify attitudes and perceptions
• analyze positive and negative feedback
• explore emotional responses
• accelerate coding workflows
• analyze interview or survey data
• identify recurring sentiment patterns

The tool also proposes relevant codes that can be applied to quotations automatically or manually.

Availability of the Opinion Mining tool

The Opinion Mining tool is currently available only in:

• ATLAS.ti Windows
• ATLAS.ti Mac

ATLAS.ti Web does not currently support Opinion Mining.

How Opinion Mining works

The Opinion Mining tool analyzes selected data and identifies:

• keywords
• sentiment polarity
• phrases connected to sentiments
• related quotations

The preliminary results display:

• positive sentiment keywords on the left side
• negative sentiment keywords on the right side

Each keyword includes phrases showing how often they appear in the data.

When selecting a keyword:

• related quotations appear
• proposed codes become available
• researchers can apply suggested codes or their own codes

How to use the Opinion Mining tool in ATLAS.ti Desktop

The Opinion Mining tool can analyze:

• documents
• document groups
• codes
• code groups

Step 1: Open the Opinion Mining tool

• In ATLAS.ti Windows

There are several ways to open Opinion Mining:

Option 1: Use the Search & Code tab

1. Open your project.
2. Open the Search & Code tab.
3. Select Opinion Mining.

Option 2: Use the Document or Code Manager

1. Open the Document Manager or Code Manager.
2. Select one or more entities.
3. Click Analyze
4. Select Opinion Mining.

Option 3: Use the analysis dropdown

1. Use the dropdown menu in the top-left corner.

   Select:

   • documents
   • document groups
   • codes
   • code groups
2. Select one or more entities.
3. Click Analysis.
4. Select Opinion Mining.

• In ATLAS.ti Mac

Option 1: Use the Analysis menu

1. Open your project.
2. Click Analysis.
3. Select Opinion Mining.

Option 2: Analyze selected entities

1. Use the dropdown menu in the top-left corner.

   Select:

   • documents
   • document groups
   • codes
   • code groups
2. Select one or more entities.
3. Click Analysis.
4. Select Show Opinion Mining.

Step 2: Select the data to analyze

• In ATLAS.ti Windows and Mac

1. Select the documents, document groups, codes, or code groups.
2. Confirm the selection.

The Opinion Mining tool processes the selected data and generates preliminary sentiment results.

Step 3: Review the sentiment results

• In ATLAS.ti Windows and Mac

1. Review the positive sentiment keywords on the left side.
2. Review the negative sentiment keywords on the right side.
3. Review the phrases connected to each keyword.

Each phrase includes information about how often it appears in the data.

Step 4: Review quotations connected to sentiments

• In ATLAS.ti Windows and Mac

1. Click a keyword.
2. Review the quotations connected to the sentiment.
3. Review the proposed codes.

Step 5: Apply suggested or custom codes

• In ATLAS.ti Windows and Mac

1. Toggle the proposed codes to apply them.
2. Or apply your own custom codes manually.

When to use the Opinion Mining tool

Researchers commonly use Opinion Mining to:

• analyze positive and negative feedback
• study attitudes and perceptions
• analyze customer or user experiences
• identify emotional responses
• explore survey responses
• analyze interview data
• accelerate thematic coding workflows

The tool can be especially useful for:

• customer feedback analysis
• social research
• market research
• user experience studies
• opinion-based qualitative studies
• exploratory analysis

Best practices for Opinion Mining

To improve analysis quality:

• review quotations before applying codes
• verify that sentiments match the context
• refine codes after automatic suggestions
• combine Opinion Mining with manual qualitative analysis
• use focused datasets when possible

Researchers should interpret sentiment results carefully, especially when analyzing:

• sarcasm
• humor
• mixed opinions
• context-dependent statements

Common issues and mistakes

• Expecting Opinion Mining in ATLAS.ti Web
  • Opinion Mining is currently available only in ATLAS.ti Desktop and is not currently supported in ATLAS.ti Web.
• Applying suggested codes without reviewing quotations
  • Always review quotations and context before accepting automatically suggested codes.
• Interpreting sentiments without context
  • AI-generated sentiment analysis may not fully understand sarcasm, irony, or nuanced meaning.
• Analyzing overly broad datasets
  • Large or highly varied datasets may generate overly broad or less focused sentiment results.
• Expecting Opinion Mining to replace qualitative interpretation
  • Opinion Mining supports analysis workflows, but researchers should still interpret findings and themes critically.

When to contact support

Contact ATLAS.ti Support if:

• the Opinion Mining tool does not open
• sentiment results fail to appear
• quotations do not display correctly
• proposed codes are missing
• analysis results appear incomplete
• the software behaves differently from the documentation

When contacting support, include:

• your platform: Windows or Mac
• your ATLAS.ti version
• screenshots or error messages
• a description of your workflow
• the type of data being analyzed