High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

SMS Spam Filtering and Classification Based on PMI and Naive Bayes Algorithm


Devika Nair , Agnel Institute of Technology and Design; Diksha Poulenkar, Agnel Institute of Technology and Design; Achal Tari, Agnel Institute of Technology and Design; Prachi Savaikar, Agnel Institute of Technology and Design


SMS, PMI, Naïve Bayes, Spam, Ham, Information Retrieval, Co-Occurrence, Classification, Categorization


Short Message Service (SMS) is one of the most widely used media of communication due to its cheapness and convenient usage. Since all the SMS are Push -Type messages and there is no flow control over the number of messages received which leads to generation of huge amounts of SMS on the device. As the growth of the subscriber and messaging volumes continue, the spam messages become more frequent leading to degradation of mobile network performance. This has resulted in increase in demand for efficient spam solutions. The main focus of this project is to filter the spam SMS using point-wise mutual information (PMI) to detect the word co-occurrences and to classify the spam / ham messages by using Naïve Bayes classifier. Then the ham messages are further categorized into different categories such as Bank, Festival, Entertainment, Shopping, Greetings and Sports.

Other Details

Paper ID: IJSRDV6I40862
Published in: Volume : 6, Issue : 4
Publication Date: 01/07/2018
Page(s): 1357-1359

Article Preview

Download Article