New approaches to origin and destination and no-show forecasting: Excavating the passenger name records treasure

Abstract

This paper describes the differences between traditional and passenger name records (PNR)-based origin and destination forecasting. It shows the opportunities for information gain and improvement in forecasting accuracy which can be generated by using PNRs as a data source. The second part of the paper addresses PNR-based no-show forecasting (PNRPRO), which uses a machine learning algorithm as the prediction model. It depicts a business case concerning the challenges to successful implementation at Lufthansa German Airlines. The main success factor there was the development of an outperformance model which mixes the results of exponential smoothing and the PNRPRO model for no-show prediction.