Abstract
The purpose of this system is to find out if a flight is getting delayed during departure and arrival then what are the reasons for the delay. Therefore we intend to aid the airlines by predicting the delays by using certain data patterns from the previous information. This system explores what factors influence the occurrence of flight delay. Classification algorithm is applied to classify flights into delay categories. Using OneR, a classification algorithm, models are developed to predict delay on both arrival and departure side. Discretization was applied using Weka, a data mining tool, to divide the delays on departure and arrival side into five categories viz; Negligible, Insignificant, Nominal, Significant, Indefinite. The result thus obtained from these categories was further analyzed to predict overall reason for delay. The delay predicted can be due to Weather, Security, Carrier, National Aviation System (NAS) and Late Arrival. The models further combine this result with Meteorological Terminal Aviation Routine Weather Report (METAR) to give the report of weather conditions at origin and destination airport. The results of data analysis will suggest that flight delays follow certain patterns that distinguish them from on-time flights. We may also discover that fairly good predictions can be made on the basis on a few attributes. Classification can be used for analyzing future data trends. It is important that the classification is appropriate so that the data prediction is accurate.