Penggunaan Algoritma Nearest Neighbor Pada Sistem Penalaran Berbasis Kasus Untuk Diagnosis Penyakit ISPA

Miswar Papuangan (Prodi Teknik Informatika, Fakultas Teknik, Universitas Pasifik Morotai, Maluku Utara)
Munazat Salmin (Prodi Teknik Informatika, Fakultas Teknik, Universitas Khairun, Maluku Utara)

Jurnal Serambi Engineering
Vol 5, No 1 (2020)

Acute Respiratory Infection  is defined as an acute respiratory disease caused by an infectious agent that is transmitted from human to human. Symptoms include shortness of breath, difficulty breathing, sore throat, fever, wheezing, runny nose, and cough. This research implements CBR to help diagnose ARI. The diagnosis process is done by entering a new problem that contains the symptoms and risk factors that will be diagnosed in the system. The normalized nearest neighbor method with expert confidence is used to calculate the similarity between new problems and cases stored on a case basis. The results of testing the system using 112 case data with 78 case stored in a case base and 34 data used as new cases data, the system has identified four types of ARI disease with a system performance measurement 97.06%.