The School of Sciences, UCLan Cyprus cordially invites you to a seminar titled:
“Towards the Personalization of CAPTCHA Mechanisms based on Individual Differences in Cognitive Processing”
with a talk from Professor George Samaras of the University of Cyprus.
One of the most
important security concerns on the World Wide Web today is to protect Web-based
systems and services against automated software agents whose purpose is to
degrade the quality of a provided service. Examples include among others the automated
creation of fake email accounts that are used later on for spam, generation of
massive scale advertising, manipulation of online voting systems, access of
private information, generation of hyperlinks in forums to improve their
Web-sites’ search engine ranking, dictionary attacks of passwords, etc.
A “Completely Automated Public Turing test
to tell Computers and Humans Apart” (CAPTCHA) is a widely used security
mechanism to protect Web applications, interfaces, and services from such malicious
software by verifying that the entity interacting with a system is actually a
human being, and not a machine. A typical example of a CAPTCHA mechanism requires
from a legitimate user to type letters or digits based on a distorted image
that appears on the screen. Such a challenge is based on the assumption that a
distorted text-based image can be easily recognized by the human brain but
present significant difficulty for an optical character or image recognition
system.
CAPTCHA challenges over the World Wide Web
are performed primarily with the use of text-recognition CAPTCHA. The reCAPTCHA
project, which is currently the most popular and widely used CAPTCHA online,
estimates that over 200 million reCAPTCHAs are completed daily, and it takes an
average of 10 seconds to complete one. In addition, major Web service providers
such as Google, Facebook, Microsoft and many others utilize text-recognition
CAPTCHA to protect their premises against automated attacks.
Current text-based CAPTCHA implementations
suffer from an important drawback; making the characters of the CAPTCHA hard to
be recognized by computer systems, also increases the difficulty for humans,
and thus decreases usability of interaction. The problem is further exacerbated
by the improvement of current character recognition systems that are more
capable of breaking CAPTCHA mechanisms, and as a consequence, the characters’
distortion and complexity is increased making it even more difficult to be
recognized by humans.
Within this realm, given that individuals share
different characteristics, needs and preferences, supporting usability of
CAPTCHA systems via adaptation and personalization technologies may improve the
system’s usability and user experience by providing users with adaptive and
personalized CAPTCHA challenges according to their unique characteristics. Given
that current text-based CAPTCHA implementations require from individuals to
recognize specific characters among irrelevant, noisy information, and process
this information on a cognitive level, we suggest that individual differences
in cognitive processing should be taken into consideration in the design of
current text-recognition CAPTCHA mechanisms.
To this end, the purpose of our research is
to investigate the effect of specific individual characteristics of users, targeting
cognitive processes (i.e., speed of processing, controlled attention and
working memory capacity), toward efficiency and effectiveness of different
variations of text-recognition CAPTCHA challenges in terms of complexity (i.e.,
low, medium and high level of complexity illustrating respectively, 5, 7 or 9
characters with increased character distortion and noise). We will report our
results of a three-month user study that investigates the effect of
individuals’ different cognitive processing abilities, targeting on speed of
processing, controlled attention and working memory capacity toward efficiency
and effectiveness with regard to different levels of complexity in
text-recognition CAPTCHA tasks. A total of 107 users interacted with CAPTCHA
challenges for a two month period indicating that the usability of CAPTCHA
mechanisms may be supported by personalization techniques based on individual
differences in cognitive processing.
Dr. Samaras is a Professor at the University of Cyprus
at the Computer Science department. He received a PhD in computer science from
Rensselaer Polytechnic Institute, USA, in 1989. He was previously at IBM
Research Triangle Park, USA and taught at the University of North Carolina at
Chapel Hill (adjunct Assistant Professor, 1990-93). He served as the lead
architect of IBM's distributed commit architecture (1990-94) and co-authored
the final publication of the architecture (IBM Book, SC31-8134-00, September
1994). He also served on IBM's internal international standards committees for
issues related to distributed transaction processing (OSI/TP, XOPEN, OMG). His
research interest includes personalization and location-based management,
collaborative infrastructures, mobile and pervasive computing, sensor networks,
mobile and wireless models and interfaces, context based services,
personalization for the wireless environment, transaction processing, commit
protocols and resource recovery, Web computing and information retrieval. He
has been involved or is currently participating, as coordinator or partner, in
more than 35 national and international (IST) funded projects. related to
eLearning, eHealth, Ambient Assisted Living, mobile, pervasive and wireless
computing.
He has published more than 220 articles in conferences
and journals and co-authored two books and a number of book chapters. His work
on utilizing mobile agents for Web database access has received the best paper
award of the 1999 IEEE International Conference on Data Engineering (ICDE99), the
work on personalizing eLearning utilizing cognitive characteristics received
the best student paper award at of the 5th International Conference on Adaptive
Hypermedia and Adaptive Web-based Systems (AH 2008) and recently the work on the
usability of CAPTCHA mechanisms by personalization techniques on individual
differences in cognitive processing received the best paper award of the 2013
SouthCHI International Conference on Human Factors in Computing & Informatics.
He also has a number of patents relating to transaction processing technology.
Dr. Samaras is on the Editorial Board of the International Journal of Internet
Technology and Secured Transactions and has served as guest editor for a number
of journals and on program committees of most top conferences and workshops on
Database and Mobile Computing. In
2005, he was the General Chair of the International Conference on Mobile Data
Management (MDM). In 2007 he co-chaired the technical program of the ACM Workshop
on Data Engineering for Mobile and Wireless Access (MobiDE) and in 2010 he was
the General Chair of the 9th Hellenic Data Management Symposium (HDMS 2010). He
is regularly (since December 1998) invited by the European Commission to serve
as an external project evaluator and auditor for the ESPRIT and IST Program
(FP5, FP6 and FP7) in areas related to internet, RFID and sensor networks,
mobile and wireless computing, Grid computing and Internet of Things. At the
European level he has reviewed and audited more than 120 multimillion-EURO
proposals and projects.
When: Wednesday, December 11th, 2013 at 14:00-15:00
Where: Room CY108 (First floor), UCLan Cyprus, Pyla CY-7080
