Penerapan Data Mining Untuk Menganalisis Data Bencana Gempa Bumi Di Kepulauan Maluku Pada BMKG Menggunakan Naïve Bayes Algorithm

Abstract

Earthquakes are natural phenomena that cannot predict the location, scale, depth of the hypocenter and the place affected by the earthquake. Until now, there is no precise theory that can be used to predict this. The collection of data on earthquake events in the Maluku Islands provides an opportunity for writers to take part in problem problems, namely by applying the big data theorem (data mining) by applying the naïve Bayes algorithm application technique using a tool that can know the level of accuracy in making earthquake predictions. The research phase begins with the earthquake data process. Then the big data is normalized and the training dataset and testing dataset are generated. Then upload the training data set on the tool and do modeling using classification techniques with the application of Hebrew by testing the training data which is then evaluated with the test data set so that the final results of the research are obtained. Based on the research it can be seen that the annual earthquake disaster in 2019-2020 in the Maluku Islands, namely North Maluku Province with the truth of the prediction data of 71.1%, in North Maluku Province there was an earthquake disaster on Ternate Island with the correctness of the prediction data of 78, 6%, In Maluku Province, there was an earthquake on Seram Island with the accuracy of the prediction data of 92%. Based on research, it can be denied that Yahoo Naïve Bayes can be used properly for earthquakes.

Keywords: Ring Of Fire, Data Mining, Normalisasi, Data Set, Data training, Data testing, Weka, Classification, Algoritma Naïve Bayes.