Abstract This paper focuses our agent-based approach to dialogue management equipped with a deliberation mechanism further extensible with additional features. The aim of this study has been to create a manager which can exhibit complex behaviour from two points of view – the development and runtime ones. To support the fi rst case, the domain developer needs to provide a set... This paper focuses our agent-based approach to dialogue management equipped with a deliberation mechanism further extensible with additional features. The aim of this study has been to create a manager which can exhibit complex behaviour from two points of view – the development and runtime ones. To support the fi rst case, the domain developer needs to provide a set of plans which tell the agent how to cope with particular tasks. Then during the runtime, the agent processes these plans to manage the dialogu e in accordance with an optimization criterion specifi ed. This paper summarizes our effort and gives an overview about our architecture. Read More Read Less

Abstract In the context detection field, an important challenge is automatically detecting the user’s task, for providing contextualized and personalized user support. Several approaches have been proposed to perform task classification, all advocating the window title as the best discriminative feature. In this paper we present a new ontology-based task detection approach, and evaluate it against previous work. We show that... In the context detection field, an important challenge is automatically detecting the user’s task, for providing contextualized and personalized user support. Several approaches have been proposed to perform task classification, all advocating the window title as the best discriminative feature. In this paper we present a new ontology-based task detection approach, and evaluate it against previous work. We show that knowledge intensive tasks cannot be accurately classified using only the window title. We argue that our approach allows classifying such tasks better, by providing feature combinations that can adapt to the domain and the degree of freedom in task execution. Read More Read Less

Abstract In this article, we introduce an alternative distribution, namely Esscher transformed Laplace distribution good enough to model the fi le size distribution of Web servers. It exhibits asymmetry, peakedness and tail heaviness, which are common features of fi le size data. Esscher transformed Laplace distribution belongs to one parameter regular exponential family. We derive their representations and obtain explicit forms... In this article, we introduce an alternative distribution, namely Esscher transformed Laplace distribution good enough to model the fi le size distribution of Web servers. It exhibits asymmetry, peakedness and tail heaviness, which are common features of fi le size data. Esscher transformed Laplace distribution belongs to one parameter regular exponential family. We derive their representations and obtain explicit forms for their densities, distribution functions and quantile functions. Also the probability R = P(X > Y ) and an asymptotic confi dence interval for R, when X and Y are two independent but not identically distributed Esscher transformed Laplace variables are estimated and its application in the service performance of a Web server is discussed. Extensive simulation studies are carried out to study the performance of these estimators. Finally using real data, we established the goodness of fi t of the proposed distribution to the fi le size distribution of a Web server data. Read More Read Less

Abstract LTSC-128 is a high secure stream cipher based on the hardness of the Shortest Vector Problem in Lattice space. The cipher is based on vectors multiplication over finite field, where these vectors are represented by polynomials to enhance the per formance of keystream generation. The key size is 128 bits and there is no attack faster than exhaustive key search... LTSC-128 is a high secure stream cipher based on the hardness of the Shortest Vector Problem in Lattice space. The cipher is based on vectors multiplication over finite field, where these vectors are represented by polynomials to enhance the per formance of keystream generation. The key size is 128 bits and there is no attack faster than exhaustive key search has been identified. The experimental results showed that the encryption rate of LTSC-128 is about 50Mbit/second. LTSC-128 cipher has additional feature that its performance can be increased by applying multithreading technique on multi-core processors. The multithreaded version of LTSC-128 has achieved better encryption rate of 71Mbit/second on dual-core processor. Read More Read Less

Abstract Merging traditional ways of teaching with modern approaches as e-Learning, focused on online multimedia applications and technologies, developed in a collaborative manner, became a real necessity in order to enhance learning and educational management. In this paper we introduce our developed Education Management Tool (EDM), a visual, interactive and collaborative e-Learning solution, dedicated for optimizing the teaching, learning and testing... Merging traditional ways of teaching with modern approaches as e-Learning, focused on online multimedia applications and technologies, developed in a collaborative manner, became a real necessity in order to enhance learning and educational management. In this paper we introduce our developed Education Management Tool (EDM), a visual, interactive and collaborative e-Learning solution, dedicated for optimizing the teaching, learning and testing processes in different Computer Science topics but also in any other scientifi c fi eld. EDM allows geographically distributed students to contribute concurrently and collaboratively by working on the same application: designing, programming, debugging, testing and documenting. Besides that, EDM is free, easy to use, and helps to improve quality and performance in e-Learning, providing fl exibility, extensibility and interactivity. The EDM Learning module allows defi ning lessons that consists of text-descriptions and media fi les such as images, fl ash movies or video clips. We illustrate how third-party lessons can be integrated into EDM by presenting our interactive fl ash applications about digital logic circuits and algorithms, two fundamental topics of Computer Science. One of the latest development achievements consists in a Management module, with proper administration of marks, averages, lessons and student’s presence at courses and laboratories, which allows generating, printing and exporting reports. Another issue, extremely necessary during collaborative writing and programming, is versioning, a mechanism which allows recovering any of the last ten versions of lessons or source codes. Read More Read Less

Abstract Digital watermarking is an emerging technology for combating copyright piracy. In real applications, the transaction between a buyer and a seller is done through a specific watermarking protocol, which enables the seller to successfully identify a traitor from a pirated copy, while preventing the dishonest seller from framing an inno cent buyer. In this paper, we propose a new buyer-seller... Digital watermarking is an emerging technology for combating copyright piracy. In real applications, the transaction between a buyer and a seller is done through a specific watermarking protocol, which enables the seller to successfully identify a traitor from a pirated copy, while preventing the dishonest seller from framing an inno cent buyer. In this paper, we propose a new buyer-seller watermarking protocol. Compared to the previous works, it makes two improvements, which make it more practi cal. First, no third party is introduced and hence, the problem of the seller (or buyer) colludes with a third party to cheat the buyer (or seller) is solved. Second,a double watermark insertion is avoided, and then the final quality of the watermarked content is higher and the result of the watermark detection is more accurate. Read More Read Less