A COLLABORATIVE AND EARLY DETECTION OF EMAIL SPAM USING MULTITASK LEARNING
Main Article Content
Article Sidebar
Abstract
Purpose:This paper includes a unique solution that attempts to use deep neural network, a machine learning technique which detects any pattern of recurrent words which may have been classified as spam.
Methodology:The algorithm used in this paper is Deep Neural Network. Neural Networks work quite similar to a human brain. They are very much capable of producing extraordinary amounts of output from limited input. This is because they are constantly learning and improving from each input provided. Data loss is not a problem for neural networks because they store all the data within themselves instead of a database. Like mentioned earlier, they are constantly learning and therefore can come up with solutions for real-time problems by comparing them with existing problems.
Findings: In the proposed system, two important techniques of neural network are Dropout and Activation. When detecting spam, especially in larger neural networks, a lot of generalization takes place. Dropout technique prevents this generalization error to its maximum extent. When generalization error is fixed, it becomes quite easy for the neural network to understand the rules of the English language and therefore figure out relationships between words like how CPU is important to a computer is similar to how Brain is important to Humans. This opens up a large field of play in various ways. When the neural network learns how to function like above, it can be used for much smaller datasets and so on, therefore improving its efficiency and its functionality to perform almost as much as a human brain. Basically, it understands that when two words are coming together next to each other a lot of times, they form some meaning and it is on the basis of this that the entire learning process of neural networks is based on which is similar to how humans think.
Originality/value: This work provides a conclusive proof on deep neural network being superior to other methods and techniques in terms of spam detection.The future work is to work on detecting botnets attacks on mobiles. Botnets are malicious machines that attack the user’s device.
How to Cite
Article Details
Deep Neural Network, Synthetic Minority Over-sampling technique, Term Space Partition Algorithm.
[2]. F. Wang, T. Xu, T. Tang, M. Zhou, and H. Wang, ‘‘Bilevel feature extraction-based text mining for fault diagnosis of railway systems’', IEEE Transactions on Intelligent Transportation Systems, vol. 18, no. 1 (2017), pp. 49-58.
[3]. Dr.AmmarAlmomani, Ahmad Al Nawasrah, Farid Meziane, Mohammed Azmi Al-Betar, “Botnet detection used fast-flux technique, based on adaptive dynamic evolving spiking neural network algorithm”, International Journal of Ad Hoc and Ubiquitous Computing, vol. 36, no. 1, (2021), pp. 50-65.
[4]. A.-Z. AlaM, H. Faris, M. A. Hassonah, “Evolving support vector machines using whale optimization algorithm for spam profiles detection on online social networks in different lingual contexts”, Knowledge-Based Systems, vol. 153, (2018), pp. 91-104.
[5]. WafaHerzallah, Hossam Faris , Omar Adwan ,“Feature engineering for detecting spammers on Twitter: Modelling and analysis”, Journal of Information Science, vol. 44, no. 2 (2018), pp. 230-247.
[6]. G. Jain, M. Sharma, and B. Agarwal, ‘‘Spam detection on social media using semantic convolutional neural network’’, International Journal of Knowledge Discovery in Bioinformatics (IJKDB), vol. 8, no. 1, (2018).
[7]. Mohammad Alauthman, “Botnet Spam E-Mail Detection Using Deep Recurrent Neural Network”, International Journal of Emerging Trends in Engineering Research, vol. 8, no. 5, (2020).
[8]. S. Saha, S. Das Gupta, and S. K. Das, ‘‘Spam mail detection using data mining: A comparative analysis’’, Smart Intelligent Computing and Applications, vol. 104, (2019), pp. 571-580.