Room 6C/6E Data Mining Social Media Networks For Terrorist Events Indicators

Friday, October 12, 2012: 8:00 PM
6C/6E (WSCC)
Nathaniel Gonzalez, BS , Department of Electrical & Computer Engineering and Computer Science, Polytechnic University of Puerto Rico, Hato Rey, PR
Alfredo Cruz , Department of Electrical & Computer Engineering and Computer Science, Polytechnic University of Puerto Rico, San Juan, PR
Jeffrey Duffany , Department of Electrical & Computer Engineering and Computer Science, Polytechnic University of Puerto Rico, San Juan, PR
Jose Ortiz , Department of Computer Sciences, Uninversity of Puerto Rico, San Juan, PR
Social media has given us the opportunity to tap into the collective conscience of the Internet and to use that knowledge to enhance our national security. Our goal is to establish a significant relationship between social media and terrorist events and to show that an automated system can be created to monitored it. Statistical analysis and data mining techniques were performed on 62 million user created documents and the events listed on the global terrorism database. Five countries were selected for further data analysis those being: Afghanistan, Colombia, India, Iraq, and Pakistan. We’ve found a significant relationship between fatalities, injuries and attacks and specific terms used on the documents. The existence of the relationship varied depending on the country that was under scrutiny. This research shows an effective way of retrieving and analyzing relevant terrorist events information out of incremental and varied social data sets. Information gathered should be used to develop strategies to diffuse or prevent future terrorist attacks.