Papers
  • Łupkowski P., Urbański M., 2008, SemCAPTCHA. Telling Computers and Humans Apart by Means of Linguistic Competence and Positive Semantic Priming, In: L. Rutkowski, R. Tadeusiewicz, L. A. Zadeh, J. Zurada (Eds.), Computational Intelligence: Methods and Applications, Academic Publishing House EXIT, Warszawa, pp. 525-531. {BibTeX} /request a copy from authors/

Abstract. This paper describes results of an experiment carried out to test a novel construction of a CAPTCHA (Completely Automated Public Turing Test To Tell Computers and Humans Apart) based on linguistic competence and positive semantic priming. The results suggest that solution time of a certain type of linguistic tasks can form a basis for human/bot differentiation. Moreover, positive semantic priming improves solution time of this type of tasks, although it does not affect solution accuracy.

  • Łupkowski P., Urbański M., 2008, SemCAPTCHA — user-friendly alternative for OCR-based CAPTCHA systems, In: G. Demenko, K. Jassem, S. Szpakowicz (Eds.), Speech and Language Technology, Volume 11, Polish Phonetic Association, Poznań, pp. 278-289. {BibTeX} /request a copy from authors/

Abstract. In this paper we present a new CAPTCHA system (Completely Automated Public Turing Test To Tell Computers and Humans Apart). This proposal, SemCAPTCHA, is motivated by an increasing number of broken OCR-based CAPTCHA systems and it is based not only on text recognition but also on text understanding.

We describe SemCAPTCHA from both user's perspective and system's perspective and compare it to some currently popular CAPTCHAs. We also briefly describe an experiment carried out to test our CAPTCHA on human users.

  • Łupkowski P., 2009, The Question of OCR-based CAPTCHAs Safety, In: R. S. Choraś, A. Zabłudowski (Eds.), Image Processing & Communications. Challenges, Academic Publishing House EXIT, Warszawa, pp. 218-224. {BibTeX} /request a copy from author/

Abstract. In this paper we discuss the level of trust that we may still have for the OCR-based CAPTCHA systems. We refer to the recent attacks on well known CAPTCHA systems used by popular portals like Yahoo, Google and MSN. We also present the results of our attack on the CAPTCHA system used by the wp.pl internet portal. Using only very simple techniques we have achieved a high level of success against this CAPTCHA scheme. In the end of the paper we present our proposal of a CAPTCHA system which security relies not only on character recognition, but also on completing a semantic task.


This research was partially supported by AMU Faculty of Social Sciences
grant No. WSO/133/2006.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License