Kanker Serviks merupakan salah satu jenis kanker yang disebabkan oleh Human Papillomavirus atau HPV. Di Indonesia, kasus kemunculan kanker serviks berada di peringkat ke-2 dibawah kanker payudara. Kebutuhan akan deteksi dini kanker serviks sangat diperlukan, terlebih kemunculan kanker serviks dapat dikenali ketika kondisi kanker memasuki stadium akhir. Dengan pemanfaatkan teknologi serta berdasarkan data perilaku, penelitian ini mengusulkan identifikasi dini kanker serviks menggunakan kombinasi seleksi fitur information gain dan data mining. Implementasi dilakukan pada dataset Cervical Cancer Risk Behavioral. Metode information gain mampu menghasilkan 9 fitur utama yang akan digunakan dalam tahap evaluasi dengan data mining. Dari hasil evaluasi beberapa model data mining diketahui metode Naive Bayes mampu memberikan performa terbaik dengan pencapaian 93,21% akurasi, 96% sensitivitas, dan 83,33% spesifisitas. Sehingga dapat disimpulkan bahwa penggunakan skema 9 atribut berdasarkan bobot information gain mampu memberikan peningkatan kemampuan model data mining dalam mengidentifikasikan kemunculan kanker serviks.

Cervical cancer is a type of cancer caused by the Human Papillomavirus or HPV. In Indonesia, the incidence of cervical cancer is ranked 2nd under the breast cancer. The need for early detection of cervical cancer is necessary because the emergence of cervical cancer can only be identified when the cancer is in a late stage. By utilizing technology and based on behavioral data, this study proposes early detection of cervical cancer using a combination of information gain feature selection and data mining. The implementation is carried out on the Cervical Cancer Risk Behavioral dataset. The information gain method provides 9 main features which will be used in the evaluation phase. From the evaluation results, it can be seen that the Naive Bayes method is able to provide the best performance with the value of 93.21% accuracy, 96% sensitivity, and 83.33% specificity. Therefore, it can be concluded that the 9 attributes scheme is able to increase the capability of data mining models in identifying the cervical cancer.