Skripsi

SISTEM TANYA JAWAB DETEKSI KANKER DINI MENGGUNAKAN METODE BERT DAN TF-IDF

---

Penilaian

---

The increasing mortality rate due to cancer, particularly in developing countries like Indonesia, highlights the urgency of developing an effective question-answering detection system. According to data from Globocan 2020, Indonesia recorded 396,914 new cancer cases with 234,511 cancer-related deaths. Additionally, Riskesdas data shows that the prevalence of cancer in Indonesia increased from 1.4 per 1,000 population in 2013 to 1.79 per 1,000 population in 2018. This study aims to develop a cancer early detection question-answering system using BERT (Bidirectional Encoder Representations from Transformers) and TF-IDF (Term Frequency-Inverse Document Frequency) methods. The combination of these two methods is expected to improve the accuracy in understanding cancer symptoms, diagnosis, and treatment. The system was tested using a dataset from Kaggle containing clinical data on various types of cancer, with preprocessing techniques such as case folding, stop word removal, stemming, and tokenization applied to enhance data quality. The system’s performance evaluation showed the highest accuracy of 98.85%, achieved with a fine-tuned BERT model. In comparison with the BERT-only model (94.70%) and TF-IDF-only model (96.55%), these results demonstrate that the integration of BERT and TF-IDF is more effective in providing accurate and relevant responses. This study also involved interviews with 10 medical students from Universitas Sriwijaya, class of 2021-2022, to test the validity of the system. Of the 20 questions asked, the system successfully answered 19 correctly, resulting in an accuracy of 95%. The findings of this study contribute to the development of artificial intelligence (AI)-based health technology and support early cancer detection efforts in Indonesia by providing an efficient and reliable cancer detection system.

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2507001514	T168832	T1688322025	Central Library (Reference)	Available but not for loan - Not for Loan

Detail Information

Series Title

-

Call Number

T1688322025

Publisher

Indralaya : Prodi Teknik Informatika, Fakultas Ilmu Komputer Universitas Sriwijaya., 2025

Collation

xvi, 189 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

004.07

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Teknik Informatika

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment