Title: Text Mining and Cybercrime
Speaker: April Kontostathis, Ursinus College
Date: Tuesday, February 15, 2011; 12:00-1:00pm
Location: DIMACS Center, CoRE Bldg, Room 431, Rutgers University, Busch Campus, Piscataway, NJ
This talk will describe progress on an ongoing project that integrates both communication theories and computer science algorithms to create a program that can detect the occurrence of internet predation and bullying in online social settings. Though there has been much work exploring social media in general, this particular aspect of online social interaction remains largely untapped.
The first phase of our study attempts to evaluate and categorize the strategies used by online sexual predators in their attempts to develop relationships with children using the Internet. Our software system, ChatCoder, is designed to decide which lines in the chat log contain predatory language. The labels we use are based on the communicative model for online predation that we have developed and are refining. We have previously developed phrase-matching and rule-based approaches to classify and label lines of chat logs, and have recently expanded these techniques to use machine learning algorithms to classify posts.
More recently we have begun to study the communicative patterns of online bullies. We expect the use of linear algebraic techniques, such as Latent Semantic Indexing (LSI) to help us learn to identify dangerous conversations. The eventual goal of the project is to develop and deploy open source software to monitor the use of social networking tools and sites by minors.
April Kontostathis is an Associate Professor of Mathematics and Computer Science at Ursinus College. She earned a Ph.D. in computer science from Lehigh University in 2004. She has written numerous articles that describe methods for determining the most critical values in the reduced dimensionality matrix and improving the efficiency and effectiveness of Latent Semantic Indexing (LSI). Currently, Dr. Kontostathis is the Principal Investigator, with Dr. Lynne Edwards, on a three-year $498,000 grant from the National Science Foundation to study and prevent cyber-bullying and cyber predation. She is also the Co-Principal Investigator, with Dr. Mohammed Yahdi, on a National Science Foundation Research Experiences for Undergraduate Sites grant.