Analysis of Machine Learning Algorithms for Email Classification Using NLP

EasyChair Preprint 4107

Abstract

In the exponentially growing world, people are using email across all areas of industries including Educational field. Therefore, it is very much important to differentiate between legitimate and spam email. In this paper, we have preprocessed emails using natural language processing and applied several machine-learning algorithms to analyze their performance on email classification. The performance observed here is accuracy and F1 score. The result shows that ANN outperforms the other algorithms. The ANN best accuracy is 98.80% and F1 score is 0.977778.

Keyphrases: Emails, Natural Language Processing, machine learning, spam classification

@booklet{EasyChair:4107,
  author    = {Mithun Das and Hardik Patel and Subrata Samanta},
  title     = {Analysis of Machine Learning Algorithms for Email Classification Using NLP},
  howpublished = {EasyChair Preprint 4107},
  year      = {EasyChair, 2020}}