Text classification guide for Android

The MediaPipe Text Classifier task lets you classify text into a set of defined categories, such as positive or negative sentiment. The categories are determined the model you use and how that model was trained. These instructions show you how to use the Text Classifier with Android apps.

You can see this task in action by viewing the demo. For more information about the capabilities, models, and configuration options of this task, see the Overview.

Code example

The example code for Text Classifier provides a simple implementation of this task for your reference. This code help you test this task and get started on building your own text classification app. You can browse the Text Classifier example code on GitHub.

Download the code

The following instructions show you how to create a local copy of the example code using the git version control command line tool.

To download the example code:

1. Clone the git repository using the following command:

   git clone https://github.com/google-ai-edge/mediapipe-samples
   

2. Optionally, configure your git instance to use sparse checkout, so you have only the files for the Text Classifier example app:

   cd mediapipe
   git sparse-checkout init --cone
   git sparse-checkout set examples/text_classification/android
   

For instruction on how to setup and run an example with Android Studio, see the example code setup instructions in the Setup Guide for Android.

Key components

The following files contain the crucial code for the text classification example app:

• TextClassifierHelper.kt - Initializes the text classifier and handles the model selection.
• MainActivity.kt - Implements the application, including calling TextClassifierHelper and ResultsAdapter.
• ResultsAdapter.kt - Handles and formats the results.

Setup

This section describes key steps for setting up your development environment and code projects specifically to use Text Classifier. For general information on setting up your development environment for using MediaPipe Tasks, including platform version requirements, see the Setup guide for Android.

Dependencies

Text Classifier uses the com.google.mediapipe:tasks-text libraries. Add this dependency to the build.gradle file of your Android app development project. You can import the required dependencies with the following code:

dependencies {
    implementation 'com.google.mediapipe:tasks-text:latest.release'
}

Model

The MediaPipe Text Classifier task requires a trained model that is compatible with this task. For more information on available trained models for Text Classifier, see the task overview Models section.

Select and download a model, and then store it within your project assets directory:

<dev-project-root>/src/main/assets

Use the BaseOptions.Builder.setModelAssetPath() method to specify the path of the model to use. For a code example, see the next section.

Create the task

Use one of the Text Classifier TextClassifier.createFrom...() functions to prepare the task for running inferences. You can use the createFromFile() function with a relative or absolute path to the trained model file. The code example below demonstrates using the TextClassifier.createFromOptions() function. For more information on the available configuration options, see Configuration options.

The following code demonstrates how to build and configure this task.

// no directory path required if model file is in src/main/assets:
String currentModel = "text_classifier_model.tflite";

fun initClassifier() {
    val baseOptionsBuilder = BaseOptions.builder()
        .setModelAssetPath(currentModel)
    try {
        val baseOptions = baseOptionsBuilder.build()
        val optionsBuilder = TextClassifier.TextClassifierOptions.builder()
            .setBaseOptions(baseOptions)
        val options = optionsBuilder.build()
        textClassifier = TextClassifier.createFromOptions(context, options)
    } catch (e: IllegalStateException) { // exception handling
    }
}

You can see an example of how to create a task in the code example TextClassifierHelper class initClassifier() function.

Configuration options

This task has the following configuration options for Android apps:

Prepare data

Text Classifier works with text (String) data. The task handles the data input preprocessing, including tokenization and tensor preprocessing.

All preprocessing is handled within the classify() function. There is no need for additional preprocessing of the input text beforehand.

String inputText = "The input text to be classified.";

Run the task

The Text Classifier uses the TextClassifier.classify() function to run inferences. Use a separate execution thread for executing the classification to avoid blocking the Android user interface thread with your app.

The following code demonstrates how to execute the processing with the task model using a separate execution thread.

    fun classify(text: String) {
        executor = ScheduledThreadPoolExecutor(1)

        executor.execute {
            val results = textClassifier.classify(text)
            listener.onResult(results)
        }
    }

You can see an example of how to run a task in the code example TextClassifierHelper class classify() function.

Handle and display results

The Text Classifier outputs a TextClassifierResult which contains the list of possible categories for the input text. The categories are defined by the the model you use, so if you want different categories, pick a different model, or retrain an existing one.

The following shows an example of the output data from this task:

TextClassificationResult:
  Classification #0 (single classification head):
    ClassificationEntry #0:
      Category #0:
        category name: "positive"
        score: 0.8904
        index: 0
      Category #1:
        category name: "negative"
        score: 0.1096
        index: 1

This result has been obtained by running the BERT-classifier on the input text: "an imperfect but overall entertaining mystery".

You can see an example of how to display results in the code example ResultsAdapter class and ViewHolder inner class.