A Data Mining Application for Performance Evaluation Cabin Crew Members in an Airline Company

Nowadays, the companies are in competition fiercely both for keeping market and catching the developing technology. The fierce competition environment aims to keep current customers and gain new customers. The impact of cabin crew is great that serving which is beyond the expectations of passengers in airline company. Performance evaluations of 3764 cabin crew members were examined in 2015 in an airline company. The levels of scorecard are determined as a result of making these evaluations. The objective of this study is making meaningful rule between evaluation scores based on competence and demographic features for the levels of scorecard in 2015. In this study, WEKA was used, which is developed in open source code JAVA, and decision tree algorithms which is one of data mining methods. It was explored that Random Forest algorithm was the best algorithm and second one was J48 algorithm in terms of true positive rate in generated decision tree algorithms. This study was interpreted according to the J48 algorithm because Random Forest algorithm output was not suitable for this study due to nonvisual output and complex structure in steps of the algorithm. In addition, it was used for attribute selection with Ranker method in “InfoGainAttributeEval” algorithm and the results was detected similar to J48 algorithm outputs. In this regard, it was determined that the most important attribute affecting cabin crew scorecard levels was “Continuous learning and personal development” and no meaningful rule between demographic attributes and scorecard level