Akademik Veri Yönetim Sistemi
The International Conference of Materials and Engineering Technology, Gaziantep, Türkiye, 10 - 12 Ekim 2019, ss.1-9, (Tam Metin Bildiri)
In recent years, social media analytics has become important research fields, which lead to improving many approaches such as crisis management systems. In addition, impressive attention has been given for mining the publically available huge amount of data to gain situational awareness, which may help in preventing or decrease the effect of some disaster through taking the correct responses. The main purpose of this study is to investigate the performance of some well-known machine learning Classifiers, i.e., K-Nearest Neighbor (KNN), Naïve Bayes (NB), Random Forest (RF), AdaBoost (AdaBoost), GradientBoosting (GBC), for enhancing the task of classifying the information available before or even during any crisis. This comparison is conducted to analyze the performance of the classification framework and provide recommendations related to improving crises management systems.
Several experiments have been conducted, and it has been observed that the performance of all the studied algorithms is somehow similar to each other. In addition, the accuracy is very much affected by the quality of the processed data. Overall, none of the studied algorithms has a stability in the performance, however, by building an ensemble system of the studied algorithms, the performance, robust and stability of the classifying process has been significantly improved.
Keyword: Crises Management Systems; K-Nearest Neighbors; Naïve Bayes; Random Forest; AdaBoost Classifier; GradientBoosting Classifier.