PHP Tutorial: Quick Integration and Parsing of Baidu Natural Language Understanding API

1. Introduction

Baidu Natural Language Processing is an AI semantic technology service by Baidu, offering a variety of natural language processing features including tokenization, part-of-speech tagging, named entity recognition, sentiment analysis, dependency parsing, text classification, and deep neural network (DNN) semantic recognition.

This article demonstrates how to call Baidu's Natural Language Understanding API using PHP and efficiently parse and process the returned JSON data.

2. Obtaining Baidu AI Application Authorization

2.1 Register a Baidu Smart Cloud Account

Before using Baidu Natural Language Understanding API, you need to register a Baidu Smart Cloud account and activate the required service permissions.

Steps:

• Visit the Baidu Natural Language Processing official website and click “Start Using” at the top right to be redirected to Baidu Smart Cloud.
• Click “Register” and fill in the required information, complete verification to finish registration.

2.2 Create an Application and Obtain API Key and Secret Key

After registration, log in to the Baidu Smart Cloud Console to create an application and get your API Key and Secret Key.

Process:

• Log into the Baidu Smart Cloud Console, go to the “Console Home” page, and click “Create Application.”
• Fill in the application name, select application type and service category, then submit.
• In the application management page, locate your API Key and Secret Key for API calls.

3. PHP Integration with Baidu Natural Language Understanding API

3.1 Set Parameters

Before calling the API, set the relevant parameters, mainly including:

• appId: API Key.
• appKey: Secret Key.
• text: Text content to be processed.
• url: URL of the text data (choose either text or url).
• language: Text language, default is “Chinese,” optional “English.”
• unit: Type of data returned by the API, options include “word,” “sentence,” “paragraph,” and “document,” default is “word.”

Example code:

$appId = 'your-app-id';
$appKey = 'your-app-key';
$text = 'Text data to process';
$url = 'URL of the text data';
$language = 'Chinese';
$unit = 'word';

3.2 Send Request

Use PHP’s cURL library to send an HTTP request. Example code:

$ch = curl_init();
$options = array(
  CURLOPT_URL => 'https://aip.baidubce.com/rpc/2.0/nlp/v1/lexer?charset=UTF-8',
  CURLOPT_POST => true,
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_HTTPHEADER => array(
    'Content-Type: application/json',
  ),
  CURLOPT_POSTFIELDS => json_encode(array(
    'appId' => $appId,
    'appKey' => $appKey,
    'text' => $text,
    'url' => $url,
    'language' => $language,
    'unit' => $unit,
  )),
);
curl_setopt_array($ch, $options);
$response = curl_exec($ch);
curl_close($ch);

3.3 Parse Response

After a successful request, the API returns JSON data, which needs to be parsed:

$result = json_decode($response);
if (!empty($result->error_code)) {
  echo 'Request failed: ' . $result->error_msg;
} else {
  foreach ($result->items as $item) {
    echo $item->item . ' ' . $item->pos . "\n";
  }
}

The above code outputs each tokenized word and its part-of-speech tag to the console for further processing.

4. Conclusion

This article explained the complete process of integrating Baidu Natural Language Understanding API using PHP, including account registration, app creation, parameter configuration, request sending, and response parsing.

Developers can adjust parameters as needed in real applications to flexibly implement features like text intelligence analysis, voice interaction, and machine translation.