Q

What were some spam filtering techniques used before the introduction of ML techniques for spam detection?

Home, - Use of machine learning in cyber security

Question: The past decade has seen the gradual growth in the use of machine learning in cyber security to protect users from cyber-attacks. As they go about their day, users are quietly protected by several security features on their devices. When you use Gmail, Google automatically scans every email to determine if it is spam. Comment on the use of machine learning algorithms for spam detection.

What were some spam filtering techniques used before the introduction of ML techniques for spam detection? Justify the statement, "Spam detection is perhaps the classic example of pattern recognition"

Answer:
Machine learning algorithms for Spam detection

Email spam filtering is done by using the various algorithms of the spam filtering.
Deep learning algorithm
Naïve bayes algorithm
Support vector machine
K nearest neighbor algorithm

Google is using the various machine learning algorithms for the detection of the spam including the logistic regression and the neural network for the classification of the email in the gmail application. The users of the gmail are protected from image spam by using OCR. Spam messages are detected by Yahoo email by using the various machine learning algorithms. Outlook is also using machine learning algorithms for spam filters.

Spam filtering techniques

1) Adaptive spam filtering technique: The spam are grouped into the different classes and in this spams are detected and filtered in this. The comparisons are made between incoming email in this.

2) Previous likeness based spam filtering technique: Memory based machine leaning methods ar used in this approach in order to detect the spam. The incoming emails are compared with the stored examples in order to detect the spam.

3) Heuristic based spam filtering technique: Huge number of patterns are accessed in this approach by using heuristics. The regular expressions are accessed and scoring of message is done based on the pattern.

In the pattern recognition, the patterns are recognized with the use of the machine learning algorithm. In the same way, patterns are recognized in the data of the email. The data is analyzed in order to find the patterns in the data.


Leave a comment


Captcha

Related :-