Social Graph Based Suspicious Chat Log Identification Using Apriori Algorithm and Support Vector Machine
Journal Title: International Journal for Research in Applied Science and Engineering Technology (IJRASET) - Year 2017, Vol 5, Issue 6
Abstract
With the increasing use of Instant Chat Messengers to share information, suspicious activities has also increased. There are many sources to share the information but instant chat messengers and social networking websites are the quick and easy means to share anything. Sometimes, even news stories are initially broken up on social media sites and further on chat messengers instead of any news channel and newspaper etc. Due to these technology advancements, some people are misusing these instant chat messengers to share suspicious activities and make plans to do something suspicious. This kind of chat is mainly available in textual format. In this paper, a social graph based concept is used to identify suspicious terms, chat sessions and users. Here, users are considered as nodes and the relation of user with any chat log is considered as edge of the graph. In this process, Support Vector Machine is used to identify & classify the suspicious key terms. Apriori algorithm is used for the social graph generation. Suspiciousness of any chat group can be predicted based on the support & confidence level of Apriori algorithm. Experiment results are evaluated in order to identify suspicious key terms, key users and key sessions. Also the weightage of key terms, key user score and normalized score has been evaluated. The declaration of user as suspicious or authentic is based on weightage that is evaluated using decision tree approach.
Authors and Affiliations
Amit Verma, Sonali Gupta, Rahul Butail
Keywords
Related Articles
- EP ID EP24460
- DOI -
- Views 333
- Downloads 11
Transmission of Numeric Data and Voice Using Light Fidelity (LIFI) Technology
LIFI is a field of research which has to cover several miles in each and every phase to prove its efficacy as a potent technology that holds the strength to challenge the wide ranging technology today. Be it the radio s...
Analysis of Web Log Server Files of E-Commerce Websites to Study Customer Behavior Pattern
web usage mining refers to the automatic discovery and analysis of patterns in click stream and related data collected or generated as a result of user interactions with web resources on one or more websites. It consist...
Suggesting Pesticides for Farmers using Data Mining
Agriculture plays a vital role in India’s economy. Virtually fifty eight % of the rural people rely upon the Agriculture. One amongst the challenges that Indian farmers encounter nowadays is how to use the pesticides in...
Crowbar Devices Used for Over Voltage Protection
One of the most common reason for the damage of electrical equipment in your power system is due to voltage Transient. There are various methods for overvoltage protection. In this paper, Protection from overvoltage is...
A Note on an Upper Bound for (n, d)
In this correspondence, we have to obtain an upper bound for the value of (n,d), we have related to the bounds on the number of code words in a linear code C of length n. In particular we have given the exact inequality...