How to Use MonkeyLearn (now Medallia)

Getting started with MonkeyLearn

With MonkeyLearn, you'll be able to classify and extract valuable insights from customer feedback, surveys, and support tickets without writing any code. This guide will help you set up your account and create your first text analysis model quickly.

Step 1: Sign up and set up

1. Go to MonkeyLearn's website.
2. Click on the Sign Up button in the top right corner.
3. Enter your email address and create a password, or sign up using Google or Microsoft.
4. After confirming your email, log in to your new account. You will be directed to the dashboard.
5. Familiarise yourself with the interface; you’ll see options for Classifiers and Extractors.

Step 2: Your first text classifier

1. From the dashboard, click on Create Model.
2. Select Text Classifier from the options provided.
3. Name your model (e.g., "Customer Feedback Classifier") and click Next.
4. Choose a pre-built model or start from scratch. For this example, select Sentiment Analysis.
5. Click on Train your model and upload a CSV file containing sample data (you can find sample datasets online).
6. Once the model is trained, click on Test to see how it classifies your text.
7. Use the API or Export options to integrate the model into your existing systems or export the results.

Step 3: Get better results

1. To improve accuracy, regularly update your training data with new feedback.
2. Use the Feedback feature to retrain your model based on incorrect classifications.
3. Explore the Dashboard to visualise your results and adjust your models as needed.
4. Experiment with different pre-built models like Topic Classifier to find the best fit for your needs.

Pro tip

After creating your model, use the Auto-Train feature to automatically retrain your model with new data. This saves you time and ensures your model stays relevant.

Common mistake to avoid

Avoid skipping the step of testing your model with real data after training. Failing to test can lead to inaccurate classifications, affecting your insights and decision-making. Always validate your model's performance before deploying it.