<![CDATA[Patients with diabetes have a high risk to suffer from other complications. In Indonesia, there are top three of diabetes complications incidents, which are retinopathy, neuropathy, and proteinuria. To prevent the worsening complication level and condition, the information about how to control and to detect early symptoms of diabetes complications become crucial. The problem is that the information about complications are limited and still widely obtained by conventional sharing between health workers and diabetic patient, this makes the complications of diabetes is often too late to be handled. Therefore, in this research, we build a Knowledge Management System (KMS) using Data Mining Algorithm J48 techniques to extract diabetes complication dominant factors. This proposed system is also able to work collaboratively between diabetics, families, and health workers, to share information based on the similarity of the profile of diabetic patient. Algorithm J48 was used to classify patient profile attributes including age, sex, duration of suffering, BMI, blood pressure, blood sugar levels, and family history, to generate rules. To generated rules and a profile matching method used as model of the proposed system, we used Indonesian diabetics data. Based on evaluation of the proposed system, it results conformity 68% with the original medical record of diabetic complication. In conclusion, the system could be used as an alternative to the early detection system.Diabetes sebagai penyakit kronis, juga adanya potensi komplikasi penyakit yang disebabkannya. Dengan prevelansi komplikasi tertinggi adalah retinopati, neuropati, dan proteinuria. Pengetahun mengenai faktor-faktor risiko terkait komplikasi tersebut dapat membantu dalam deteksi dini, sehingga dapat mengurangi risiko dan mempercepat penanganan pasien. Permasalahannya, informasi mengenai komplikasi masih terbatas serta diperoleh secara konvensional dengan melakukan sharing antara tenaga kesehatan dan penderita diabetes, sehingga komplikasi diabetes seringkali terlambat untuk ditangani. Pada penelitian ini dilakukan perancangan knowledge managementSystem(KMS) komplikasi diabetes, dengan mengadopsi teknik data mining J48 dan profile matching dalam pengklasifikasian potensi komplikasi diabetes yang dialami oleh pasien tersebut. Pada sistem terdapat pula fitur berbagi informasi dan pengetahuan. Hal ini dilakukan secara kolaborasi antara penderita diabetes, keluarga dan tenaga kesehatan. Hasil evaluasi terhadap sistem usulan diperoleh 68% kemungkinan komplikasi yang sesuai dengan komplikasi berdasarkan data rekam medis penderita diabetes. Sehingga sistem dapat digunakan sebagai alternative early detection sistem.]]>
Copyrights © 2020
