Practical Text Mining and Statistical Analysis for Non-structured Text Data Applications

Tutorial B

Text Mining Improves Model Performance in Predicting Airplane Flight Accident Outcome

Jennifer Thompson,     Woodward, OK, USA

Thomas Hill,     Tulsa, OK, USA

Contents

Introduction

The Data

Text Mining the Data

Text Mining Results

Data Preparation

Using Text Mining Results to Build Predictive Models

Introduction

Keywords from unstructured text can be powerful predictors in data mining. When text mining is used to extract these key terms, we can see substantial gains in model performance. Unstructured text is often more descriptive than the results from a one-size-fits-all questionnaire. This is particularly true of the examples used in this tutorial.

The National Transportation Safety Board is a federal agency dedicated to investigating ...