Skripsi
DETEKSI MALICIOUS URL PADA FILE BERBASIS FITUR LEKSIKAL MENGGUNAKAN METODE RANDOM FOREST.
With the existence of attack models such as phishing and malware distribution, these actions begin by accessing a Uniform Resource Locator (URL) or files containing harmful links within them. A Uniform Resource Locator (URL) is a specific identifier used to locate resources through the internet. URLs can pose threats to availability, control, confidentiality, and data integrity, with one of the threats being malicious URLs. To differentiate between malicious URLs and normal URLs, feature extraction is employed to identify important characteristics of malicious URLs. The extraction features used are lexical features consisting of 18 attributes. After extraction, due to the imbalanced dataset, resampling is performed using oversampling with SMOTE. To classify the dataset, this research utilizes a machine learning algorithm known as random forest. Random Forest is an algorithm that constructs multiple decision trees. This algorithm can achieve high classification accuracy and provide good results. In this study, the evaluation yields results with an accuracy value of 90.97%, precision of 99.05%, recall of 85.39%, and an f1-score of 91.71%.
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2307003994 | T127806 | T1278062023 | Central Library (Referens) | Available |
No other version available