Conspiracy Detection by Real Time Email Analysis

Abstract

In this thesis, we have proposed a method to turn this psychological concept into a machine that can automatically detect the conspiracy among the employee by analyzing their email data in real time. Here we have proposed the design using vector based classification method for analyzing the text data. We have used TFIDF method to victimization and prioritize the frequency of conspiracy related word and concept. And also we used Logistic Regression, a prediction based classifier to classify the text sentiment. Supervised vector-based methods to sentiment can design rich lexical meanings. This method for machine learning is largely used in present days. Sentiment analysis for online text document has been a burgeoning field of text mining among researchers for the past few decade. Nevertheless and sentiment analysis on Email data, a ubiquity means of social network and communication, has been studied thoroughly. Email has become the most popular communication tools for official purpose. Almost every private company uses their own mail server for exchanging their official mail. So, it has a great significance in terms of business and communication. In the other hand conspiracy is a social concept that has also a great importance and impact over the working place. It is a pure psychological concept. It influences in the progress of any working place.