Cybercrime: To Detect Suspected User’s Chat Using Text Mining

Abstract

In this fast growing era of modern technology, it has become important need for people to communicate with each other. Various mediums like social media are used for communication business, organizational details, etc. Due to the increase in technology, there are chances of performing the crimes in newer ways. Using these Social Networking Sites (SNS), many criminal activities have the ability to extract the information. Such activities are spamming, cyber prediction, cyber threatening, killing, blackmailing, phishing, etc. The suspicious messages can be transferred via different SNS, mobile phones, or other sources. Most of the crime-related activities on information on a web are in text format, which is tedious task to trace those criminal activities. Detecting and exploring the crime and identifying the criminals are involved in the analyzing “crime process.” Criminology is a field of using text mining techniques that describe the complexity of relationship between crime datasets. Text mining technique is an effective way to detect and predict criminal activities. Text mining is the process of extracting interesting information or knowledge or patterns from the unstructured text that are from different sources. Since this uses text mining algorithm to continuously check for suspicious words even if they are in short forms or phrases, then they find social graph of users. In this framework, n-gram technique with SCHITS (Hyperlink-Induced Topic Search) algorithm is used to find suspected message, user, sessions, etc. Information can be extracted using social graph-based text mining, and also the suspected user’s profile has been found.