Digital Library

cab1

 
Title:      TOWARD EFFICIENT DETECTION OF CHILD PORNOGRAPHY IN THE NETWORK INFRASTRUCTURE
Author(s):      Asaf Shupo , Miguel Vargas Martin , Luis Rueda , Anasuya Bulkan , Yongming Chen , Patrick C.k. Hung
ISBN:      ISSN: 1646-3692
Editors:      Pedro IsaĆ­as and Marcin Paprzycki
Year:      2006
Edition:      V I, 2
Keywords:      Computer forensics, packet classification, P2P networks
Type:      Journal Paper
First Page:      15
Last Page:      31
Language:      English
Cover:      no-img_eng.gif          
Full Contents:      click to dowload Download
Paper Abstract:      Child pornography is an increasingly visible problem in society today. Methods currently employed to combat it may be considered primitive and inefficient, and legal and technical issues can exacerbate the problem significantly. We propose a network-based detection system that uses a stochastic weak estimator coupled with a linear classifier, which is appropriate in this context due to the non-stationarity of the input data. Our experiments show that the system is capable of distinguishing child pornography images from non-child pornography images even when the obscene image is reduced to only 20% of its representation. This method for identifying offensive material is potentially attractive to law enforcement and can be accomplished with acceptable overhead. We believe our approach, with minor adaptations, is of independent interest for use in a number of network applications which benefit from packet classification beyond detecting child pornography. These include security applications such as detecting malicious packets, and network anomalies consisting of dangerous traffic fluctuations, abusive use of certain services, and distributed denial-of-service attacks.
   

Social Media Links

Search

Login