Internet content filtering using isotonic separation on content category ratings

ABSTRACT

The World Wide Web has enabled anybody with a low cost Internet connection to access vast information repositories. Some of these repositories contain information (e.g., hate speech and pornography) that is considered objectionable, especially for children to view. Several efforts---legal and technical---are underway to protect children and the generic public from accessing this type of content. We propose a technical approach utilizing a recently proposed technique called isotonic separation for filtering with content metadata if they satisfy monotone conditions. We illustrate this approach using a category rating method of PICS. In essence, we formulate the Internet content filtering problem as a classification problem on content metadata and report on experiments we conducted with the isotonic separation technique.

REFERENCES

• 1: Ahuja, R. K., Magnanti, T. L., and Orlin, J. B. 1993. Network Flows. Prentice-Hall, Englewood Cliffs, NJ.
• 2: Balkin, J. M., Noveck, B. S., and Roosevelt, K. 1999. Filtering the Internet: A best practice model. Working paper, Yale Information Society Project, Yale Law School.
• 3: Barlow, R. E., Bartholemew, D. J., Bremner, J. M., and Brunk, H. D. 1972. Statistical Inference under Order Restrictions. John Wiley & Sons, New York.
• 4: Brickley, D. and Guha, R. V. 2004. RDF vocabulary description language 1.0: RDF schema. Tech. Rep. REC-rdf-schema-20040210, World Wide Web Consortium.
• 5: Brickley, D. and Swick, R. R. 2000. PICS rating vocabularies in XML/RDF. Tech. Rep. REC-rdf-pics-20000327, World Wide Web Consortium.
• 6: Brown, D. E., Corruble, V., and Pittard, C. L. 1993. A comparison of decision tree classifiers with backpropagation neural networks for multimodal classification problems. Pattern Recognition 26, 6, 953--961.
• 7: Jonathan P. Caulkins , Wenxuan Ding , George Duncan , Ramayya Krishnan , Eric Nyberg, A method for managing access to web pages: filtering by statistical classification (FSC) applied to text, Decision Support Systems, v.42 n.1, p.144-161, October 2006 [doi>10.1016/j.dss.2004.11.015]
• 8: Chandrasekaran, R., Ryu, Y. U., and Jacob, V. 1998. Breast cancer diagnosis using an isotonic separation approach. Working paper, Department of Information Systems and Operations Management, School of Management, The University of Texas at Dallas.
• 9: Chandrasekaran, R., Ryu, Y. U., Jacob, V., and Hong, S. 2005. Isotonic separation. INFORMS J. Comput. 17, 4, 462--474.
• 10: Peter Clark , Robin Boswell, Rule Induction with CN2: Some Recent Improvements, Proceedings of the European Working Session on Machine Learning, p.151-163, March 06-08, 1991
• 11: Peter Clark , Tim Niblett, The CN2 Induction Algorithm, Machine Learning, v.3 n.4, p.261-283, March 1989 [doi>10.1023/A:1022641700528]
• 12: Consumers Union. 2005. Filtering software: Better, but still fallible. Consumer Reports 70, 6 (June), 36--38.
• 13: Lorrie Faith Cranor , Lawrence Lessig, Web Privacy with P3p, O'Reilly & Associates, Inc., Sebastopol, CA, 2002
• 14: Cranor, L. F., Langheinrich, M., Marchiori, M., Presler-Marshall, M., and Reagle, J. 2002. The platform for privacy preferences 1.0 (P3P1.0) specification. Tech. Rep. REC-P3P-20020416, World Wide Web Consortium.
• 15: Evans, C., Feather, C. D. W., Hopmann, A., Presler-Marshall, M., and Resnick, P. 1997. PICSRules 1.1 specification. Tech. Rep. REC-PICSRules-971229, World Wide Web Consortium.
• 16: Simson Garfinkel , Gene Spafford, Web security & commerce, O'Reilly & Associates, Inc., Sebastopol, CA, 1997
• 17: Melody Y. Kiang, A comparative assessment of classification methods, Decision Support Systems, v.35 n.4, p.441-454, July 2003 [doi>10.1016/S0167-9236(02)00110-0]
• 18: Krauskopf, T., Miller, J., Resnick, P., and Treese, W. 1996. PICS label distribution label syntax and communication protocols, version 1.1. Tech. Rep. REC-PICS-labels-961031, World Wide Web Consortium.
• 19: Manola, F. and Miller, E. 2004. RDF primer. Tech. Rep. REC-rdf-primer-20040210, World Wide Web Consortium.
• 20: McBride, B., Wenning, R., and Cranor, L. 2002. An RDF schema for P3P. Tech. Rep. NOTE-p3p-rdfschema-20020125, World Wide Web Consortium.
• 21: Miller, J., Resnick, P., and Singer, D. 1996. Rating services and rating systems (and their machine readable descriptions), version 1.1. Tech. Rep. REC-PICS-services-961031, World Wide Web Consortium.
• 22: Murthy, S. K., Kasif, S., and Salzberg, S. 1994. A system for induction of oblique decision trees. J. Art. Intell. Res. 2, 1, 1--32.
• 23: Katta G. Murty, Network programming, Prentice-Hall, Inc., Upper Saddle River, NJ, 1992
• 24: J. R. Quinlan, Induction of Decision Trees, Machine Learning, v.1 n.1, p.81-106 [doi>10.1023/A:1022643204877]
• 25: J. Ross Quinlan, C4.5: programs for machine learning, Morgan Kaufmann Publishers Inc., San Francisco, CA, 1993
• 26: Young U. Ryu , R. Chandrasekaran , Varghese Jacob, Prognosis Using an Isotonic Prediction Technique, Management Science, v.50 n.6, p.777-785, June 2004 [doi>10.1287/mnsc.1030.0137]
• 27: Ryu, Y. U. and Yue, W. T. 2005. Firm bankruptcy prediction: Experimental comparison of isotonic separation and other classification approaches. IEEE Trans. Syst., Man, and Cybernetics, Part A: Systems and Humans 35, 5, 727--737.
• 28: James Ze Wang , Jia Li , Gio Wiederhold , Oscar Firschein, Classifying Objectionable Websites Based on Image Content, Proceedings of the 5th International Workshop on Interactive Distributed Multimedia Systems and Telecommunication Services, p.113-124, September 08-11, 1998
• 29: Wang, J. Z., Li, J., Wiederhold, G., and Firschein, O. 1998a. System for screening objectionable images. Comput. Comm. 21, 15, 1355--1360.

INDEX TERMS

Primary Classification:
  H. Information Systems
  H.3 INFORMATION STORAGE AND RETRIEVAL
      H.3.3 Information Search and Retrieval
          Subjects: Information filtering

Additional Classification:
  H. Information Systems
  H.3 INFORMATION STORAGE AND RETRIEVAL
      H.3.4 Systems and Software

General Terms:
Design, Performance

Keywords:
Internet content filtering, PICS, isotonic separation