Azure SDK for JavaScript

Azure Text Analytics SDK for JavaScript

The Azure Text Analytics client library for JavaScript enables you to perform natural language processing tasks on unstructured text. This SDK provides a convenient way to integrate powerful AI-driven text analysis capabilities into your JavaScript applications, whether they run in the browser or on Node.js.

Key features include:

• Sentiment Analysis: Determine the overall sentiment (positive, negative, neutral, mixed) of a document or sentence.
• Key Phrase Extraction: Identify the main talking points within a document.
• Language Detection: Detect the language of the input text.
• Entity Recognition: Recognize and categorize entities (people, places, organizations, etc.) within text.
• Named Entity Recognition (NER): A more advanced form of entity recognition for specific types of entities.
• Personally Identifiable Information (PII) Detection: Identify and redact sensitive personal information.

Prerequisites

Before you can use the Azure Text Analytics SDK, you'll need:

• An Azure subscription.
• A Text Analytics resource created in your Azure subscription. You can create one via the Azure portal.
• A Text Analytics API key and endpoint. These can be found in your Text Analytics resource's "Keys and Endpoint" section.
• Node.js installed (if targeting Node.js).

Installation

To install the SDK for Node.js:

npm install @azure/cognitiveservices-textanalytics

For browser-based applications, you can use a CDN or bundle the library with your project.

Core Concepts

The SDK is built around the concept of a TextAnalyticsClient. You'll instantiate this client with your API key and endpoint to interact with the Text Analytics service.

Getting Started

Here's a basic example of how to use the SDK to detect the language of a text document.

// Import the TextAnalyticsClient and TextDocumentInput
import { TextAnalyticsClient, TextDocumentInput } from "@azure/cognitiveservices-textanalytics";
import { AzureKeyCredential } from "@azure/core-auth";

// Replace with your actual subscription key and endpoint
const subscriptionKey = "YOUR_TEXT_ANALYTICS_SUBSCRIPTION_KEY";
const endpoint = "YOUR_TEXT_ANALYTICS_ENDPOINT";

// Authenticate the client
const auth = new AzureKeyCredential(subscriptionKey);
const client = new TextAnalyticsClient(endpoint, auth);

// Prepare the document(s) for analysis
const documents: TextDocumentInput[] = [
    { id: "1", text: "Hello world!" },
    { id: "2", text: "Bonjour le monde!" }
];

async function detectLanguage() {
    try {
        const results = await client.detectLanguage(documents);
        console.log("Language Detection Results:");
        for (const doc of results) {
            if (!doc.error) {
                console.log(`  Document ID: ${doc.id}, Detected Language: ${doc.primaryLanguage.name} (ISO6391: ${doc.primaryLanguage.iso6391Name})`);
            } else {
                console.error(`  Document ID: ${doc.id}, Error: ${doc.error.message}`);
            }
        }
    } catch (error) {
        console.error("An error occurred during language detection:", error);
    }
}

detectLanguage();
                    

API Reference

Explore the full API reference for more details on all available methods such as recognizePiiEntities, analyzeSentiment, and extractKeyPhrases.