In the realm of online security and privacy URLs malicious in nature post a major and grave risk. This is not the limit, most fundamental way to attack on online user are types of URLs.
Use of Machine Learning in Big Data
One of the primary sources of distribution of malware or viruses in the world of the internet is URL(s). This has led to increasing in the urge of URLs classification. For the sake of security of users from attacks, different cyber-security companies use methods like blacklisting the URLs as well as block these malicious URLs. But that may not be enough, because in the realm of online world there are millions of these types of URLs generated every single day and blacklisting these into a central database is a monotonous routine or method. This whole process also missed out on the freshly generated URLs.
Recommend:
Deep Convolutional Neural Network Techniques
How to solve this using machine learning?
Now to solve these types of problems machine learning plays a vital role by grabbing attention in recent past years this is very helpful to figure out the hidden patterns in the dataset of URLs. In spite of the fact that it shows a guarantee, it is by all accounts wasteful when the size of information is very huge. This prompts the presentation of enormous information innovations where there is a need of an hour to apply ML calculations in a disseminated situation. The main idea behind in this is that, first we compare performance of state of the art ML methods and techniques using distributed modern day ML algorithms of Spark MLlib. As our main model classifier we choose One-Class SVM to train our proposed classification model which gives the integrity related to URLs.
0 Comments