Machine Learning-Based Advertisement Banner Identification Technique for Effective Piracy Website Detection Process

In the contemporary world, digital content that is subject to copyright is facing significant challenges against the act of copyright infringement. Billions of dollars are lost annually because of this illegal act. The current most effective trend to tackle this problem is believed to be blocking those websites, particularly through affiliated government bodies. To do so, an effective detection mechanism is a necessary first step. Some researchers have used various approaches to analyze the possible common features of suspected piracy websites. For instance, most of these websites serve online advertisement, which is considered as their main source of revenue. In addition, these advertisements have some common attributes that make them unique as compared to advertisements posted on normal or legitimate websites. They usually encompass keywords such as click-words (words that redirect to install malicious software) and frequently used words in illegal gambling, illegal sexual acts, and so on. This makes them ideal to be used as one of the key features in the process of successfully detecting websites involved in the act of copyright infringement. Research has been conducted to identify advertisements served on suspected piracy websites. However, these studies use a static approach that relies mainly on manual scanning for the aforementioned keywords. This brings with it some limitations, particularly in coping with the dynamic and ever-changing behavior of advertisements posted on these websites. Therefore, we propose a technique that can continuously fine-tune itself and is intelligent enough to effectively identify advertisement (Ad) banners extracted from suspected piracy websites. We have done this by leveraging the power of machine learning algorithms, particularly the support vector machine with the word2vec word-embedding model. After applying the proposed technique to 1015 Ad banners collected from 98 suspected piracy websites and 90 normal or legitimate websites, we were able to successfully identify Ad banners extracted from suspected piracy websites with an accuracy of 97%. We present this technique with the hope that it will be a useful tool for various effective piracy website detection approaches. To our knowledge, this is the first approach that uses machine learning to identify Ad banners served on suspected piracy websites.