Deteksi Email Phishing Berbasis LSTM yang Ditingkatkan Menggunakan Fitur Emosi yang Diekstraksi oleh DistilBERT
Abstract
Apa : Phishing email merupakan ancaman keamanan siber yang mengeksploitasi aspek psikologis untuk menipu pengguna. Sistem deteksi saat ini umumnya hanya menganalisis pola teks dan mengabaikan fitur manipulasi emosional, sehingga rentan terhadap serangan rekayasa sosial yang canggih. Masukan sistem berupa teks email dan keluarannya adalah klasifikasi phishing atau non-phishing.
Mengapa : Topik ini krusial karena serangan phishing sering menjadi pintu masuk serangan siber lainnya. Metode deteksi berbasis deep learning yang ada belum memanfaatkan pemahaman emosi (seperti urgensi atau ketakutan) sebagai fitur deteksi, menciptakan celah keamanan yang signifikan.
Bagaimana Penelitian ini mengusulkan metode deteksi yang mengintegrasikan fitur emosional hasil ekstraksi DistilBERT ke dalam klasifikasi Long Short-Term Memory (LSTM). Sistem bekerja dalam dua cabang paralel: satu cabang memproses urutan teks dan cabang lainnya memproses enam label emosi yang diekstrak dari fine-tuned DistilBERT. Kedua fitur tersebut digabungkan untuk menghasilkan prediksi yang lebih akurat.
Hasil utama. Evaluasi pada 5.809 email menunjukkan model yang diusulkan mencapai akurasi 0,98 dan recall 0,97, mengungguli model baseline tanpa fitur emosi yang hanya mencapai recall 0,87. Peningkatan recall sebesar 0,10 membuktikan bahwa integrasi analisis emosi efektif mengurangi false negative dalam deteksi phishing.
Kata kunci: phishing detection, long short-term memory, distilled bidirectional encoder representations from transformers, emotion analysis, deep learning, phishing detection using emotion
References
A. Dhakal, “The Need for a Multi-Media Approach in Agricultural Communication: Bridging Socioeconomic Gaps,” European Journal of Theoretical and Applied Sciences, vol. 3, no. 1, pp. 3–10, Jan. 2025, doi: 10.59324/ejtas.2025.3(1).01.
A. Jean and T. Alherbe, “Gid Crypto: Application for End-to-End Encrypt and Decrypt E-mail and Data,” ASEAN Journal of Scientific and Technological Reports, vol. 27, no. 2, pp. 90–102, Mar. 2024, doi: 10.55164/ajstr.v27i2.251127.
Z. Alkhalil, C. Hewage, L. Nawaf, and I. Khan, “Phishing Attacks: A Recent Comprehensive Study and a New Anatomy,” Front. Comput. Sci., vol. 3, Mar. 2021, doi: 10.3389/fcomp.2021.563060.
N. PETLIAK, Y. BEZKOROVALNYI, and N. KUPCHYK, “ANALYSIS OF MODERN METHODS OF DETECTION OF PHISHING E-MAILS,” Herald of Khmelnytskyi National University. Technical sciences, vol. 341, no. 5, pp. 510–515, Oct. 2024, doi: 10.31891/2307-5732-2024-341-5-73.
M. Benmalek, “Ransomware on cyber-physical systems: Taxonomies, case studies, security gaps, and open challenges,” Internet of Things and Cyber-Physical Systems, vol. 4, pp. 186–202, 2024, doi: 10.1016/j.iotcps.2023.12.001.
S. Eppakayala and S. K. G. Ankula, “Intelligent Spam Filtering and Analysis Using Web Automation, ReportLab, and Ensemble Machine Learning Techniques,” Feb. 11, 2025. doi: 10.21203/rs.3.rs-5941111/v1.
T. Amutha and S. Geetha, “Automated Spam Detection Using Sandpiper Optimization Algorithm-Based Feature Selection with the Machine Learning Model,” IETE J. Res., vol. 70, no. 2, pp. 1472–1479, Feb. 2024, doi: 10.1080/03772063.2023.2280663.
G. R and F. M. Philip, “Phishing Detection,” YMER Digital, vol. 21, no. 06, pp. 405–412, Jun. 2022, doi: 10.37896/YMER21.06/39.
A. G. Putrada and R. F. Aushaf, “L2 Regularization to Mitigate Overfitting in Email Phishing Attack Detection by CNN-LSTM,” in 2024 IEEE 22nd Student Conference on Research and Development (SCOReD), IEEE, Dec. 2024, pp. 416–421. doi: 10.1109/SCOReD64708.2024.10872755.
M. Jari, “A Comprehensive Survey of Phishing Attacks and Defences: Human Factors, Training and the Role of Emotions,” International Journal of Network Security & Its Applications, vol. 14, no. 5, pp. 11–24, Sep. 2022, doi: 10.5121/ijnsa.2022.14502.
Santosa Pohan, Desi Irfan, Intan Nur Fitriyani, Yusril Iza Mahendra Hasibuan, and Indah Chayani, “Simulation and Detection of Phishing Attacks on Student Academic Emails Using Social Engineering Techniques,” International Journal of Health Engineering and Technology, vol. 2, no. 4, May 2025, doi: 10.55227/ijhet.v2i4.283.
R. Karamagi and S. Ally, “Security Risk Scale: A Case of Email Phishing Detection using Text Mining,” Journal of ICT Systems, vol. 1, no. 2, pp. 93–108, Dec. 2023, doi: 10.56279/jicts.v1i2.43.
A. J. J S, “A Survey on Phishing Email Detection Techniques: Using LSTM and Deep Learning,” Int. J. Res. Appl. Sci. Eng. Technol., vol. 13, no. 8, pp. 1597–1600, Aug. 2025, doi: 10.22214/ijraset.2025.73836.
N. Altwaijry, I. Al-Turaiki, R. Alotaibi, and F. Alakeel, “Advancing Phishing Email Detection: A Comparative Study of Deep Learning Models,” Sensors, vol. 24, no. 7, p. 2077, Mar. 2024, doi: 10.3390/s24072077.
O. Senouci and N. Benaouda, “Enhancing Phishing Detection in Cloud Environments Using RNN-LSTM in a Deep Learning Framework,” Journal of Telecommunications and Information Technology, no. 1, pp. 1–9, Mar. 2025, doi: 10.26636/jtit.2025.1.1916.
S. Atawneh and H. Aljehani, “Phishing Email Detection Model Using Deep Learning,” Electronics (Basel)., vol. 12, no. 20, p. 4261, Oct. 2023, doi: 10.3390/electronics12204261.
F. A. Acheampong, H. Nunoo-Mensah, and W. Chen, “Transformer models for text-based emotion detection: a review of BERT-based approaches,” Artif. Intell. Rev., vol. 54, no. 8, pp. 5789–5829, Dec. 2021, doi: 10.1007/s10462-021-09958-2.
S. Jamal, H. Wimmer, and I. H. Sarker, “An improved transformer‐based model for detecting phishing, spam and ham emails: A large language model approach,” SECURITY AND PRIVACY, vol. 7, no. 5, Sep. 2024, doi: 10.1002/spy2.402.
V. Sanh, L. Debut, J. Chaumond, and T. Wolf, “DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter,” Mar. 2020.
M. Rokunojjaman, A. Malik, M. M. J. Sajeeb, M. Yahiduzzaman, and M. S. Islam, “Multi-Feature Hybrid LSTM-CNN Framework for Phishing Email Detection,” International Journal of Electronics and Communications Systems, vol. 5, no. 1, pp. 93–103, Jun. 2025, doi: 10.24042/ijecs.v5i1.26764.
K. F. Rafat, Q. Xin, A. R. Javed, Z. Jalil, and R. Z. Ahmad, “Evading obscure communication from spam emails,” Mathematical Biosciences and Engineering, vol. 19, no. 2, pp. 1926–1943, 2021, doi: 10.3934/mbe.2022091.
E. O. Asani, O. S. Babalola, A. E. Akinola, A. S. Barnabas, O. V. Adams, and J. O. Odumesi, “A Server-side Phishing detection API using Long-Short Term Memory (LSTM) and Multi-layer Perceptron (MLP),” in 2024 International Conference on Science, Engineering and Business for Driving Sustainable Development Goals (SEB4SDG), IEEE, Apr. 2024, pp. 1–8. doi: 10.1109/SEB4SDG60871.2024.10629792.
P. Bhatti, Z. Jalil, and A. Majeed, “Email Classification using LSTM: A Deep Learning Technique,” in 2021 International Conference on Cyber Warfare and Security (ICCWS), IEEE, Nov. 2021, pp. 100–105. doi: 10.1109/ICCWS53234.2021.9703084.
N. Elgiriyewithana, “Emotions dataset for NLP,” Kaggle. Accessed: Nov. 13, 2025. [Online]. Available: https://www.kaggle.com/datasets/nelgiriyewithana/emotions
N. A. Alam, “Phishing Email Dataset,” Kaggle. Accessed: Nov. 13, 2025. [Online]. Available: https://www.kaggle.com/datasets/naserabdullahalam/phishing-email-dataset
