Skip to main content

Tutorial 4C: Building an NLP model for customer complaints with H2O Driverless AI

Introduction​

Overview​

In this tutorial, we will explore how to use H2O Driverless AI to build a Natural Language Processing (NLP) model for analyzing customer complaints in the banking sector. We will use the dataset from Kaggle's Consumer Complaint Resolution Dataset, with the target column Customer disputed?. This target indicates whether a customer disputed the resolution of their complaint.

Objectives​

The banking sector often receives a high volume of customer complaints, and resolving them effectively is important for maintaining customer satisfaction and loyalty. Understanding the root causes of disputes and predicting whether a resolution will be disputed is a challenge. By leveraging Natural Language Processing (NLP), we aim to develop a predictive model that analyzes complaint narratives and other features to predict if a customer will dispute the resolution.

This can help banks to identify potential issues in resolutions, improve customer service processes, and reduce dispute rates proactively. By the end of this tutorial, you will learn how to,

  • Streamline the process of analyzing textual complaint data using NLP techniques in H2O Driverless AI.
  • Develop a predictive model that identifies patterns in customer complaints to forecast disputes.
  • Interpret the key factors contributing to disputes.
  • Provide hands-on experience in automating NLP workflows using H2O Driverless AI.

By the end of this tutorial, you will have a practical understanding of how to use H2O Driverless AI to create an NLP model for dispute prediction and gain insights into improving customer service in the banking sector.

Prerequisites​

  • Access to the Lab 23 in Aquarium containing H2O Driverless AI v1.10.7 (LTS).
  • Familiarity with H2O Driverless AI is essential. Alternatively, completion of the following tutorial is also acceptable: Tutorial 1A: Introduction to H2O Driverless AI.
Feedback