ML.NET Sentiment Demo

Kim Too Flex

The code first creates a new ML context. From this context, it loads the training data and splits it into a training and test set. Next, it defines the model's training algorithm and trains the model on the training set. Finally, it uses the model to make predictions on the test set.

using Microsoft.ML;
using Microsoft.ML.Data;
using Microsoft.ML.Trainers;

namespace SentimentAnalysis
{
  class Program
  {
    static void Main(string[] args)
    {
      // Create a new ML context
      var mlContext = new MLContext();

      // Load the training data
      var data = mlContext.Data.LoadFromTextFile<SentimentData>(
        "./sentiment.csv",
        hasHeader: true,
        separatorChar: ',');

      // Split the data into training and test sets
      var split = mlContext.Data.TrainTestSplit(data, testFraction: 0.2);

      // Define the model's training algorithm
      var pipeline = mlContext.Transforms.Text.FeaturizeText(
          outputColumnName: "Features",
          inputColumnName: nameof(SentimentData.SentimentText))
        .Append(mlContext.BinaryClassification.Trainers.FastTree(
          labelColumnName: "Label",
          featureColumnName: "Features"));

      // Train the model
      var model = pipeline.Fit(split.TrainSet);

      // Use the model to make predictions on the test set
      var predictions = model.Transform(split.TestSet);

      // Evaluate the model's performance
      var metrics = mlContext.BinaryClassification.Evaluate(
        data: predictions,
        labelColumnName: "Label",
        scoreColumnName: "Score");

      Console.WriteLine("Accuracy: {0:P2}", metrics.Accuracy);
      Console.WriteLine("AUC: {0:P2}", metrics.AreaUnderRocCurve);
    }
  }

  // Define the data schema for the input data
  class SentimentData
  {
    [LoadColumn(0)]
    public bool Sentiment;

    [LoadColumn(1)]
    public string SentimentText;
  }
}