How to Perform Sentiment Analysis Using IBM Watson API in Java

Integrate IBM Watson API

• First, include IBM Watson's Java SDK in your project. You can use a build tool like Maven for this. Add the following dependency in your `pom.xml`:

<dependency>
  <groupId>com.ibm.watson.developer_cloud</groupId>
  <artifactId>java-sdk</artifactId>
  <version>9.9.0</version>
</dependency>

• After adding the dependency, update your project to download the required libraries.

Authenticate with IBM Watson

• Obtain your API key and URL from the IBM Cloud dashboard for the Language Translator service.

 

• Use this information to authenticate your application. Here's a sample code snippet for authentication:

IamAuthenticator authenticator = new IamAuthenticator("your-api-key");
NaturalLanguageUnderstanding service = new NaturalLanguageUnderstanding("2021-08-01", authenticator);
service.setServiceUrl("https://api.us-south.natural-language-understanding.watson.cloud.ibm.com/instances/your-instance-id");

• The `IamAuthenticator` utilizes your API key to authenticate requests to the Watson API services.

Set Up Sentiment Analysis Parameters

• Configure the sentiment analysis parameters using the `SentimentOptions` object. This allows you to customize the analysis, such as setting the target text or analyzing specific features:

Features features = new Features.Builder()
    .sentiment(new SentimentOptions.Builder()
    .targets(Arrays.asList("product", "service"))
    .build())
    .build();

• This configuration targets specific keywords during sentiment analysis.

Perform Sentiment Analysis

• Define the text you wish to analyze and create an `AnalyzeOptions` object. This object contains the text and features you specified:

String text = "I absolutely love this product. The service is excellent, but the price is a bit high.";
AnalyzeOptions parameters = new AnalyzeOptions.Builder()
    .text(text)
    .features(features)
    .build();

• Pass these parameters to the `analyze` method of the `NaturalLanguageUnderstanding` service to receive the analysis results:

AnalysisResults results = service.analyze(parameters).execute().getResult();
System.out.println(results);

• The response will contain the sentiment analysis result indicating whether the sentiment is positive, negative, or neutral.

Handle and Interpret Results

• Once you receive the results, you can process and interpret the sentiment data:

if (results.getSentiment() != null) {
    System.out.println("Document sentiment: " + results.getSentiment().getDocument().getLabel());
    System.out.println("Score: " + results.getSentiment().getDocument().getScore());
}

• The label indicates the sentiment category (e.g., positive, neutral, negative), and the score represents the confidence level.