PENERAPAN ALGORITMA KLASIFIKASI NAIVE BAYES UNTUK DATA STATUS HUNI RUMAH BANTUAN DANA REHABILITASI DAN REKONSTRUKSI PASCA BENCANA ERUPSI GUNUNG MERAPI 2010

Abstract

Bencana Erupsi gunung Merapi berikut susulan material lahar hujan pada Tahun 2010 mengakibatkan kerusakan rumah dan infrastruktur di wilayah Kabupaten Sleman D.I.Yogyakarta dan Kabupaten Magelang Jawa Tengah. Melalui Perka BNPB No.5 Tahun 2011, pemerintah menginstruksikan rencana dan aksi rehabilitasi dan rekonstruksi pasca erupsi dilakukan dengan skema program Rehabilitasi dan Rekonstruksi Masyarakat dan Permukiman Berbasis Masyarakat. Skema program ini telah membangun rumah sebanyak 2.516-unit bagi warga yang terdampak erupsi Merapi dan lahar hujan. Menurut Key Performance Indikator (KPI) The World Bank, status huni rumah terbangun merupakan salah satu indikator kinerja program rehab rekon. Pelaksanaan program rehab dan rekon ini sebagian besar didokumentasikan secara digital dan terekam ke dalam basis data. Dalam Ilmu Teknologi Informasi dibidang data mining, basis data merupakan aset yang dapat digunakan sebagai bahan pengenalan dan penemuan pola-pola data yang dapat dipelajari dan diteliti guna menyelesaikan permasalahan. Basis data yang dimiliki Satker rehab rekon merekam sebanyak 2.146-unit rumah huntap sudah dihuni dan 370 rumah belum dihuni. Hasil penelitian/eksperimen menunjukkan bahwa penerapan algoritma klasifikasi Naive Bayes dapat diterapkan terhadap data status huni rumah bantuan dana rehabilitasi dan rekonstruksi pasca erupsi Merapi 2010 dengan hasil nilai akurasi klasifika si mencapai sebesar 89,59% dan nilai performa klasifikasi AUC mencapai 0,826

Kata kunci : Erupsi Merapi, Data Mining, Naive Bayes, Klasifikasi, Rehab Rekon, Status huni

Disaster Eruption of Mount Merapi and the following a mixture of lava rain material in 2010 resulted in damage to homes and infrastructure in the Sleman Regency of D.I.Yogyakarta and Magelang District of Central Java. Through Perka BNPB No.5 of 2011, the government instructed plans and actions for rehabilitation and reconstruction after the eruption was carried out with the scheme of the Community Rehabilitation and Reconstruction and Community Based Settlement program. The program scheme has built 2,516-unit houses for residents affected by Merapi and rain lava eruptions. According to The World Bank's Key Performance Indicator (KPI), the occupancy status of built houses is one of the indicators of the performance of the rehabilitation and reconstruction program. The implementation of the rehabilitation and reconstruction program is mostly digitally documented and recorded in the database. In Information Technology in the field of data mining, the database is an asset that can be used as an introduction and discovery of data patterns that can be studied and researched to solve problems. The database owned by the reconstruction rehabilitation work unit recorded 2,146 housing units has been occupied and 370 houses have not been occupied. The results of the research / experiment show that the application of the Naive Bayes classification algorithm can be applied to the occupancy status data of houses for rehabilitation and reconstruction assistance after the 2010 Merapi eruption with the classification accuracy reaching 89.59% and the AUC classification perf ormance value reaching 0.826

Keywords: Merapi Eruption, Data Mining, Naive Bayes, Classification, Reconstruction Rehabilitation, Occupied status