Metode Modified Weighted Minkowski Untuk Pengembangan Sistem Penalaran Berbasis Kasus

Faizal, Edi

Jurnal Saintekom : Sains, Teknologi, Komputer dan Manajemen
Vol 7 No 1 (2017): March 2017

Knowledge acquisition process is not easy, because of the different levels of expertise even though all true. Computer experts had tried other methods to resolve the problem of the acquisition, which is known as case-based reasoning. Representation of knowledge in CBR is a collection of previous case. This research focus is the application of CBR for diagnosing womb diseases. The level of similarity is calculated by using the modified weighted Minkowski. Methods of data collection are interviews, observation and study of literature. The test results show the system can be recognize the womb disease correctly is 94.44% (sensitivity), specitifity rate of 57.14%, PPV of 85.00% and 80.00% NPV. The system have an accuracy rate of 84.00% with an error rate of 16.00%.