Automated Labeling of Helicopter Maintenance Records with Text Classification

Separating plaintext documents into defined classes, known in the data science literature as text classification, is a time-consuming task that is essential to the processing of correspondence, safety records, and maintenance documents in the aviation industry. This work discusses a machine learning approach to this problem, and demonstrates how to construct an end-to-end document classification solution. This system is capable of classifying documents with high accuracy, thus alleviating the need for labor dedicated to this task.