Showing papers in "Informatica (slovenia) in 2002"

Combining World Wide Web and Wireless Security

Abstract: In current electronic commerce systems, customers have an on-line interaction with merchants via a browser on their personal computer. Also payment is done electronically via the Internet, mostly with a credit card. In parallel to this, e-services via wireless-only systems are emerging. This paper identifies security and functionality weaknesses in both of these current approaches. The paper discusses why and how general-purpose mobile devices could be used as an extension to PC based systems, to provide more security and functionality. General-purpose mobile devices are shown to be an alternative to costly special-purpose hardware. This combined approach has in many cases more interesting properties than when using mobile devices only. As an example of the combined approach, a GSM based electronic payment system is proposed and investigated. The system enables users to order goods through the World Wide Web and pay by using their mobile phone.

Trojan Horse Attacks on Software for Electronic Signatures

Abstract: Electronic signatures are introduced by more and more countries as legally binding means for signing electronic documents with the primary hope of boosting e-commerce and e-government. Given that the underlying cryptographic methods are sufficiently strong, attacks by Trojan horse programs on electronic signatures are becoming increasingly popular. Most of the current systems either employ costly or inflexible – yet still inadequate – defence mechanisms or simply ignore the threat. A signatory has to trust the manufacturer of the software that it will work in the intended way. In the past, Trojan horse programs have shown to be of growing concern for end-user computers. Software for electronic signatures must provide protection against Trojan horses attacking the legally relevant signing process. In a survey of commercial of the shelf signature software programs we found severe vulnerabilities that can easily be exploited by an attacker. In this work we propose a secure electronic paper as a countermeasure. It is a collection of preventive and restorative methods that provides, in parallel to traditional signatures on paper, a high degree of protection of the system against untrustworthy programs. We focus our attention on Microsoft Windows NT and Windows 98, two operating systems most likely to be found on the customers' computers. The resulting system is an assembly of a small number of inexpensive building blocks that offers reliable protection against Trojan horse programs attempting to forge electronic signatures.

Artificial Neural Networks Based Systems for Recognition of Genomic Signals and Regions: A Review

Abstract: In this review we present a number of important applications in computational genomics of a class of intelligent systems, namely artificial neural networks (ANNs). We present the current state-of-the-art solutions used in recognition of different genomic signals and regions. All systems to be commented are based fully or in part on the ANNs. We included systems that recognize different aspects of a/ transcriptional control information related to promoters, TATA-box regulatory region, and polyA signal, b/ those that relate to translation process comprising recognition of the translation initiation site, coding cDNA/EST fragments, reading frame-shift errors and their correction, and c/ splice-sites recognition. The review includes some of the most efficient systems for the indicated recognition problems in bioinformatics and aims to be an initial guide for those interested in these challenging problems.

Developing software across time zones: an exploratory empirical study

Abstract: Market forces and an increasingly reliable world-wide communications network have made geographically distributed software engineering a reality. Global software development enables businesses to respond more easily and more quickly to global market opportunities and to improve product and service quality. One of the many potential benefits of global development is a reduction in development time through the adoption of an ‘around the clock’ working practice.

Critical note on the role of the quantum mechanical "collapse" in quantum modeling of consciousness.

Abstract: We give a brief account on the existing strategies in quantum-mechanical approaching the problem of consciousness. To a list of distinguished approaches we add the next plausible notion: when treated quantum- mechanically, consciousness should be modeled as an open quantum system. This notion is tightly connected to the von Neumann's "collapse" ("wave packet reduction"). Here we strongly emphasize that the problem of the "collapse" cannot be resolved within the quantum mechanics of open quantum systems (or: decoherence theory). Some clues in this regard are briefly outlined.

Compiling and Using the IJS-ELAN Parallel Corpus.

Abstract: A new and distinct cultivar of chrysanthemum plant named 'Remix', bearing medium to large sized purple and white blooms with a yellow-green center and performance of 19-22 days.

Informon - An Emergent Conscious Component.

Abstract: This article deals with a conscious entity being the building block of conscious systems. Such an informational entity is called informon, with its local and global function. The idea of informon as a conscious unit roots in the property of a sufficient complexity and learning capabilities, realized by different sorts of informational interpretation, formalism, general and metaphysicalistic decomposition concerning informon. Emotional and cognitive informons show the complexity, intention, and capability of consciously, sub-consciously, and self-consciously autonomous entities informing within a conscious system.

Using Image Segmentation as a Basis for Categorization.

Abstract: Image categorization is the problem of classifying images into one or more of several possible categories or classes, which are defined in advance. Classifiers can be trained using machine learning algorithms, but existing machine learning algorithms cannot work with images directly. We consider a representation based on texture segmentation and a similarity measure which has been used successfully in the related area of image retrieval. A generalized kernel for use with the support vector machine (SVM) algorithm can be built from such a similarity measure. We compare this approach with a more straightforward representation based on auto-

Multisecret Sharing Immune Against Cheating.

Abstract: Cheating in multisecret sharing is considered. Multisecret sharing is defined by a mapping F : GF (p) −→ GF (p) that provides a generic model. In this model, we propose nonlinear multisecret sharing that is immune against cheaters. Two cheating strategies are considered. In the first one, all cheaters always submit their invalid shares and they collectively know their own valid shares. In the second one, some cheaters may submit their valid shares while again sharing their knowledge about their valid shares. The combiner (or recovery algorithm) interacts with shareholders by collecting shares from them and distributing the recovered secrets back to active participants. Two different scenarios are considered when the combiner recreates all secrets (this is simultaneous recovery) or gradually (so called sequential recovery). Probabilities of successful cheating are derived and constructions for cheating immune multisecret sharing are given.

Cryptanalysis of some Hash Functions based on Block Ciphers and Codes

Abstract: At PKC 2000, Inoue and Sakurai proposed some methods to design hash functions from block ciphers and codes (block codes and convolutional codes). They claimed that their hash functions are secure: encryptions are necessary to find a collision, where and are the minimal distance of the code and the block size of block cipher, respectively. However, we show in this paper that a collision could be found with about encryptions, where is a small number.