Literature Survey: S. No. 1. Authors and Year Ming-Syan Chen and ChiYao Tseng – 2009 2. Haiying Shen and Ze Li – 2013 Leveraging Social Networks for Effective Spam Filtering In this work an effective spam filtering model is proposed named as SOAP. In SOAP, each node connects to its social friends; that is, nodes form a distributed overlay by directly using social network links as overlay links. Each node uses SOAP to collect information and check spam autonomously in a distributed manner. 3. Sunil B. Rathod and Tareek M. Pattewar – 2015 Content Based Spam Detection in Email using Bayesian Classifier In this work a Bayesian approach for classifying Spam and legitimate mails is discussed using supervised learning across features extracted. Title Incremental SVM Model for Spam Detection on Dynamic Email Social Networks Concept/ Theoretical Model In this work an incremental support vector machine (SVM) model MailNET is proposed for spam detection on dynamic email social networks Relevant Findings Drawbacks Several features of each user in the network are extracted to distinguish spam nodes from normal users The SVM learning approach is applied to train an SVM model based on extracted features SOAP exploits the social relationships among email correspondents and their (dis)interests to detect spam adaptively and automatically It integrates four components: social closeness-based spam filtering, social interest-based spam filtering, adaptive trust management, and friend notification The emphasized approach can detect the spam emails with an accuracy of 96.6% It is based on the training of the model on a specific dataset which is updated on a regular basis. The proposed model does not have the capability to retrain on addition of new mails. It has to be performed manually. Due to integration of all these components in the Bayesian filter, the complexity of the model increases drastically and expensive hardware is required for the implementation. The discussed approach is entirely based on the content filtering that is keyword matching, which can be fooled easily. 4. Shubhi Shrivastava and Anju R – 2017 Spam Mail Detection through Data Mining Techniques 5. Shubhangi Suryawanshi, Anurag Goswami and Pramod Patil – 2019 Email Spam Detection : An Empirical Comparative Study of Different ML and Ensemble Classifiers 6. Sanaa Kaddoura, Omar Alfandi and Nadia Dahmani – 2020 A Spam Email Detection Mechanism for English Language Text Emails Using Deep Learning Approach 7. Nikhil Govil, Kunal Agarwal, Ashi Bansal and Astha Varshney – 2020 A Machine Learning based Spam Detection Mechanism In this paper an approach The classifier models to spam classification perform better with using Bernoulli’s and Bernoulli’s continuous probability probability distribution is presented. distribution than Naive Bayes and with continuous Decision Trees algorithm probability were used to build the distribution classifier and the effect The performance of of overfitting on the classifier models performance and varies and depends accuracy is analyzed and on several factors implemented. such as the probability distribution used, dataset and the problem involved In this work different The performance of classification algorithms the classifier like Naive Bayes, SVM, depends on the KNN, Bagging and number of features Boosting (Adaboost), and size of the and Ensemble Classifiers dataset. are compared with a Ensemble Classifier voting mechanism on gives promising email dataset from results than the other Kaggle and UCI classifiers and the repository. speed of the testing is also better In this work, a spam Our FFNN model Email detection has been studied to mechanism based test their on FFNN is introduced performances to which focuses on segregate Emails as classifying link-less spam or ham emails using machine experiments on the learning approach, deep Enron dataset neural networks On the basis of the F-1 score our proposed model is better than the BERT model at classifying the link-less emails. In this work, Naïve The proposed Bayes classifier algorithm generates algorithm is used that dictionary and gives you the features and trains probabilistic index of that them through and helps identify machine learning for whether the mail is spam In this approach, traditional algorithm like naïve bayes and decision tree algorithm are used which could be replaced by random forest or fuzzy algorithm and better results could be obtained. In this work testing was performed on email dataset without taking in the evolving patterns in the emails which may affect the accuracy of a classifier In this work, deep neural network approach is implemented making a complex task to understand and implement. In this work, traditional algorithm like , Naïve Bayes classifier is used. A modern or not as per the shown results 8. Nandhini.S and Dr.Jeen Marseline.K.S – 2020 Performance Evaluation of Machine Learning Algorithms for Email Spam Detection 9. Sefat E Rahman and Shofi Ullah – 2020 Email Spam Detection using Bidirectional Long Short Term Memory with Convolutional Neural Network 10. Fahima Hossain, Mohammed Nasir Uddin and Rajib Kumar Halder – 2021 Analysis of Optimized Machine Learning and Deep Learning Techniques for Spam Detection In this work, the performance of five important machine learning classification algorithms viz. Logistic Regression, Decision Tree, Naïve Bayes, KNN and SVM are evaluated in order to train and build an effective machine learning model for email spam detection In this work a new model is proposed for detecting spam messages based on the sentiment analysis of the textual data of the email body. The model consists of three different networks namely WordEmbeddings, CNN and Bi-LSTM. effective results The random tree outperforms other classification algorithms in terms of all performance metrics KNN produces the same result, it took more time to build the model than random tree The recall, precision and f-score is used for comparing and evaluating the performance of our proposed approach The proposed model outperforms not only to some popular machine learning classifiers but also to state of the art approaches In this work, the Machine Learning proposed model is algorithms perform implemented in both better than deep machine learning and learning algorithms deep learning to establish for tabular dataset a comparative to identify spam eanalysis. Multinomial mails Naïve Bayes (MNB), Random Forest (RF), K-Nearest Neighbor (KNN), Gradient Boosting (GB) and Recurrent Neural Network (RNN), Gradient Descent (GD), Artificial Neural Network (ANN) were implemented for machine learning and deep learning respectively. algorithm like random forest could be used to get better results. The parameters used to tarin the model were not tuned optimally. They could be tuned perfectly when this process is automated. In this work the proposed model was only tested on the predefined datasets and the model may not come up with the anomality’s in the real time data and may not be as efficient. In this work, deep learning algorithm were not parameterized appropriately which leads to the poor performance.