Ready to take your Mendix app to the next level? In this blog, I’ll walk you through the process of integrating Amazon Comprehend into your Mendix app, enabling you to instantly analyze and respond to customer feedback. With this integration, you’ll gain deeper insights from user responses and be able to act accordingly to better meet their needs.  

Let’s explore how to seamlessly implement this feature using the Amazon Comprehend Connector, available in the Mendix Marketplace. 

• The Amazon Comprehend connector integrates your application with Amazon Comprehend, enabling your web apps to analyse and derive insights from user-generated text.  

• It supports key features such as Entity Recognition, Sentiment Analysis, Key Phrase Extraction, Language Detection, Syntax Extraction, and PII Entity Recognition.  

• With this connector, you can effortlessly identify entities (like people, organizations, locations, and dates), detect sentiment, extract key phrases, recognize PII entities, analyse syntax, and determine the language of the text. 

Steps to Integrate the Connector in a Mendix Application:

• Create a new app using Mendix Studio Pro version 10.13.1 or higher.

• Download the Amazon Comprehend Connector from the Mendix Marketplace. from the Mendix Marketplace.

• Download the AWS Authentication module from the Marketplace.

• Before exploring the implementation files, ensure your AWS credentials are properly configured. This includes the Access Key and Secret Access Key, which are set within the AWS Authentication module.

The Amazon Comprehend Connector enables access to key features such as Entity Recognition, Sentiment Analysis, Key Phrase Extraction, Language Detection, Syntax Extraction, and PII Entity Recognition.

• Entity Recognition – Identifies entities such as people, places, dates, quantities, organizations, etc.
• Sentiment Analysis – Determines the sentiment of the text as Positive, Negative, Neutral, or Mixed.
• Key Phrase Extraction – Extracts the key phrases or concepts from a given text.
• Language Detection – Automatically identifies the language of the input text.
• Syntax Extraction – Analyzes the syntax of the input text, including tokens (words), parts of speech, and sentence structure.
• PII Entity Recognition – Detects Personally Identifiable Information (PII) like names, addresses, credit card numbers, SSNs, and more.

The connector provides six Java actions that allow integration with Amazon Comprehend services. These actions support the detection of entities, PII entities, sentiment, key phrases, syntax, and the language of the input text.

Additionally, a set of prebuilt microflows is included, each demonstrating how to utilize one of these Java actions to fully leverage the features offered by Amazon Comprehend.

To detect sentiment from user-provided text, follow the steps below. These steps are common for all key features of Amazon Comprehend:

1. Create an object for the entity listed below to utilize the microflows and Java actions provided by the connector.

2. The microflow should be designed as shown below and is available in the example folder included with the connector. The page (ComprehendAnalysis) referenced below is used to capture the user-provided text.

If the text is already stored in an entity, or if you’re processing text extracted from audio or video, you can pass it directly to this object during its creation.

3. To utilize any of the features available in the connector, you need the following three values as shown below:

a. ComprehendAnalyser Object – holds the user-provided text

b. Credentials Object – contains AWS credentials

c. Region – specifies the AWS region

4. As stated in Step 1, we can create the ComprehendAnalyser object via microflow. Moving forward, make sure that you have configured the AWS credentials (AccessKey, SecretAccessKey) in the respective constants, and use the GetStaticCredentials microflow from the AWSAuthentication module(Refer to the ACT_ComprehendAnalyser_StartAnalysis microflow from the connector.)

5. Once the microflow is ready, we can pass the user-provided text to retrieve the sentiment from the user feedback.

6. The output we receive will be similar to the one shown below.

Example Input Text

Sentiment Analysis Output

Language Detection Output

Key Phrases Detection Output

Refer to the example folder of the connector for a detailed implementation of sentiment analysis and other Amazon Comprehend features.

The Bottom Line:

Integrating Amazon Comprehend with your Mendix app empowers you to tap into real-time sentiment insights—driving smarter decisions and more personalized user experiences. It’s a simple add-on with powerful impact. Ready to build more intuitive apps? Start analyzing what your users are really saying.