Variable Length Character N-Gram Approach for Online Writeprint Identification

The Internet´s numerous benefits have always been coupled with shortcomings due to the abuses of online anonymity. Writeprint identification is a technique to identify individuals based on textual identity cues people leave behind online messages. Character n-gram is one of the most effective approaches to identify writeprint according to previous research. In this study, we propose a variable length character n-gram based writeprint identification framework to address the identity tracing problem, integrating a genetic algorithm (GA) based feature selection component to solve the definition problem of n. To examine the approach, experiments are conducted on a test bed encompassing hundreds of reviews posted by 20 Amazon customers. The experimental results show the proposed approach is effective, obtaining a considerable improvement in identification accuracy and a heavy reduction of feature dimensionality.