Efficient Hostile URL Spotting using Top- K Characteristics with HTML TAGs

Shaik Irfan Babu,M.V.P.Chandra Sekhara Rao

Websites in modern era are equipped with details such as business promotion, financial transactions and online shopping. This rapid growth leads to certain problems i.e. customer safety issues. It’s because websites give complete access to personal data of client, attackers create attractive websites to draw attention of clients. Further, these clients are trapped to use these websites and give access to their personal information to the attackers. Such cyber-attacks can be prevented using of cyber security which can classify whether the Uniform Resource Locator (URL) accessed by clients is malicious or not. In this paper, Phish Tank URL ( is used for dataset collection to find the accuracy of the results for several k-Means techniques, like Fine KNN, Medium KNN, Coarse KNN and Cosine KNN. In addition to this, k-Means techniques are experimentally proven to provide better precision in classification compare to other algorithms.

Volume 12 | Issue 2

Pages: 1890-1906

DOI: 10.5373/JARDCS/V12I2/S20201233