Informatics Publishing Limited

A. Pandian ¹, J. Venkatasubramanian ¹, K. Balaji ²

Affiliations
1 Dept of MCA, SRM University, Chennai, IN
2 SRM University, Chennai, IN

Abstract

Authentication is the first line of defense against compromising confidentiality and integrity. Though traditional login/password based schemes are easy to implement, they have been subjected to several attacks. As alternative, token and biometric based authentication systems were introduced. However, they have not improved substantially to justify the investment. Thus, a variation to the login/password scheme, viz. graphical scheme was introduced. But it also suffered due to shoulder-surfing and screen dump attacks. In this paper, we introduce a framework of our proposed (IPAS) Implicit Password Authentication System, which is immune to the common attacks suffered by other authentication schemes. IPAS is similar to the Pass Point scheme with some finer differences. In every "what youknow type" authentication scheme we are aware of, the server requests the user to reproduce the fact given to the server at the time of registration. This is also true in graphical passwords such as Pass Point.

Keywords

Security System, Authentication Scheme, IPAS.

Full Text

     

A. Pandian ¹, J. Venkatasubramanian ¹, S. E. Chandiran ²

Affiliations
1 Dept of MCA, SRM University, Chennai, IN
2 SRM University, Chennai, IN

Abstract

Existing sequence mining algorithms mostly focus on mining for subsequences. However, a large class of applications, such as biological DNA and protein motif mining, require efficient mining of “approximate” patterns that are contiguous. The few existing algorithms that can be applied to find such contiguous approximate pattern mining have drawbacks like poor scalability, lack of guarantees in finding the pattern, and difficulty in adapting to other applications. In this paper, we present a new algorithm called FLAME (FLexible and Accurate Motif Detector). FLAME is a flexible suffix tree based algorithm that can be used to find frequent patterns with a variety of definitions of motif (pattern) models. It is also accurate, as it always find the pattern if it exists. Using both real and synthetic datasets, we demonstrate that FLAME is fast, scalable, and outperforms existing algorithms on a variety of performance metrics. Using FLAME, it is now possible to mine datasets that would have been prohibitively difficult with existing tools.

Keywords

FLAME, Data Mining, Distributed Algorithms, Dataset, Decision Trees, Classification.

Full Text