Spam Detection using NLP Techniques

Natural Language Processing is a vital field of research having applications in different subjects. Text Classification is a part of NLP where the text is converted into a machine-readable form by performing various methods. Tokenizing, part-of-speech tagging, stemming, chunking are some of the text classification methods. Implementing these methods on our data gives us a classified data on which we will train the model to detect spam and ham messages using Scikit-Learn Classifiers. We proposed a model to solve the issue of classifying messages as spam or ham by experimenting and analyzing the relative strengths of several machine learning algorithms such as K-Nearest Neighbors (KNN), Decision Tree Classifier, Random Forest Classifier, Logistic Regression, SGD Classifier, Multinomial Naive Bayes(NB), Support Vector Machine(SVM) to have a logical comparison of the performance measures of the methods we utilized in this research. The algorithm we proposed achieved an average accuracy of 98.49% with SVM model on ‘SMS Spam Collection’ dataset