Latest issue

International Review on Computers and Software - November 2008 - Papers

On the Usage of the Naïve Bayes Model in Unsupervised Word Sense Disambiguation
An Off Line System for the Arabic Handwritten Words Recognition
Ontology-based Personalization of Hypermedia Courses
Textured Images Classification Using Nearest Feature Line Method Added with Kurtosis and Skewness
An Efficient Image Retrieval Approach Based on Spatial Correlation of the Extrema Points of the IMFs
A New Method for the Feature Selection in Pattern Recognition Algorithms
Study of Specific Genetic Operators for Learning Kohonen Maps in Function of Initial Conditions
Evolutionary Algorithms In The Optimization Of Greenhouse Climate Model Parameters
Parallel MLTS Construction Preserving the Alpha Equivalence Reduction
Protein Secondary Structure Prediction: Application of RBF Networks
Comparative Performance of Embedded Coders at High Quality
A Method of Construction of Resilient Functions with Maximal Possible Nonlinearity
Real-Time Agent Computing Platform for Distributed Satellite Systems
Remote Control Architecture over Internet Based on Multi agents Systems
MoSAL: A Tile Logic Based Architecture Description Language for Mobile Software Applications

	IReCoS - November 2007	IReCoS - November 2006
IReCoS - September 2008	IReCoS - September 2007	IReCoS - September 2006
IReCoS - July 2008	IReCoS - July 2007	IReCoS - July 2006
IReCoS - May 2008	IReCoS - May 2007	================
IReCoS - March 2008	IReCoS - March 2007
IReCoS - January 2008	IReCoS - January 2007

International Review on Computers and Software - Papers

On the Usage of the Naïve Bayes Model in Unsupervised Word Sense Disambiguation
by F. T. Hristea

Abstract - The present survey paper refers to distributional approaches to unsupervised word sense disambiguation (WSD) that rely on monolingual corpora, with focus on the usage of the Naïve Bayes model in unsupervised WSD. The theoretical model is presented and its implementation is discussed. Special attention is paid to feature selection and parameter estimation, the two main issues of the model’s implementation. The paper refers to the main existing implementation, which has generated valuable software for sense disambiguation, while leaving the problem an open one.

Keywords: word sense disambiguation, unsupervised disambiguation, Bayesian classification, the EM algorithm.

An Off Line System for the Arabic Handwritten Words Recognition

by S. Ouchtati, L. Bennacer, M. Bedda

Abstract - In this work we present an off line system for the recognition of the Arabic handwritten words of the Algerian departments. The study is based mainly on the evaluation of neural network performances, trained with the gradient back propagation algorithm. The used parameters to form the input vector of the neural network are extracted on the binary images of the handwritten word by several methods: the Barr features applied on the whole binary image of the word, the parameters of distribution and the moments centered of the different projections. Let's note that these last two methods are applied on segments gotten after the division of the binary image of the word in six segments. The classification is achieved by an multi layers perceptron. A detailed experiments are carried and satisfactory recognition results are reported.

Keywords: Optical Characters Recognition, Neural networks, Barr features, Image processing, Pattern Recognition, Features extraction.

Ontology-based Personalization of Hypermedia Courses
by A. Yessad, C. Faron-Zucker, M.T. Laskri

Abstract - In the context of increasing need of life-long training, a major issue is the development of supports for personalized access to learning objects (resources, models and services) in order to reuse them in a particular training. Markup of learning objects with ontology-based metadata will allow reutilization and personalization. This paper presents OrPAF, an Adaptive Educational Hypermedia and web-based System (AEHS). The key feature of OrPAF is a learning ontology that formally specifies learning knowledge used to describe different classical learning models (domain model, pedagogical model and learner model). The learning ontology is semi-automatically constructed and permits to effectively adapt the system OrPAF to different training needs (a new domain/discipline, a new pedagogical strategy, new learner profiles, etc.). Based on its learning ontology, OrPAF generates an adaptive hypermedia course. The structure of the hypermedia is a graph of relevant domain learning concepts and the pages of the hypermedia are the relevant resources queried from external web repositories. Semantic annotations and a semantic relevance measure are proposed in order to identify the relevance of the external learning resources.

Keywords: Adaptive Educational Hypermedia, Semantic Web, Ontology, Metadata, Learning resource.

Textured Images Classification Using Nearest Feature Line Method Added with Kurtosis and Skewness
by T. Boudghene Stambouli, A. Ouamri, M. Keche

Abstract - In this study, we try to compare the performances of various textured images classification methods, all based on multi resolution decompositions. The images tested are extracted from the Brodatz album. For decomposition, we use the Separable (SWT) and NonSeparable Wavelet Transforms (NSWT). The sub images obtained allow extracting feature vectors. Two types of classes are made up and Nearest Feature Line Method is used to classify each sub image. The performances are evaluated by calculating the classification error rates and the retrieval efficiency in each case. Lastly, we try to identify the factors influencing the performances.

Keywords: Textures, Multi Resolution, Wavelet, Mallat Algorithm, Quincunx Subsampling, Classification, Nearest Feature Line (NFL).

An Efficient Image Retrieval Approach Based on Spatial Correlation of the Extrema Points of the IMFs

by A. Sabri, M. Karoud, H. Tairi, A.Aarab

Abstract - In this paper we propose a new content-based image retrieval approach based on spatial correlation of extrema points of the IMFs. This consists, first, in decomposing the image by the BEMD into its IMFs and, then, constructs the descriptor vector from the spatial correlation of the extrema points of each IMF. The choice of the BEMD is based on two major reasons: the first is that the obtained IMFs can be seen as a filter bank results like in wavelet decomposition, and the second is that the spatial information of the decomposed image is retained within the IMFs (each IMF contains information of a specific scale). The efficiency, the stability and the robustness of the proposed approach are tested on two different image databases constituted by images with different orientations and different scale transformation. The simulation results are very satisfactory compared with some existent shape descriptors like Gabor and Wavelet-Db2, and it shows the interest of these descriptors in handing image transformations such as image rotation and image scaling. This new approach allows a finer research by exploring the modes in the increase order.
Copyright © 2008 Praise Worthy Prize S.r.l. - All rights reserved

Keywords: Image retrieval, Descriptors vectors, BEMD, Intrinsic Mode Functions, Spatial correlation.

A New Method for the Feature Selection in Pattern Recognition Algorithms
by I. Esmaeil Zadeh, H. Nabovati, M. R. Asgari

Abstract - Most of the pattern recognition algorithms extract certain features of the input signal and compare them with some pre-saved patterns. Important criteria for such systems are the defining features which construct the feature vector and assign them proper weighting. Different methods for the optimization of the pattern recognition algorithms have been proposed in the literature, but most of them tend to fail to optimize the system when the number of the features increases beyond a certain value. The present paper presents a new method for the selection and weighting procedure of the features. The proposed method is applied to a voice recognition system and the simulation results are reported. The results showed about 28% improvement in the system performance in comparison with the not-optimized system.

Keywords: Feature vector, Genetic algorithm, Pattern recognition.

Study of Specific Genetic Operators for Learning Kohonen Maps in Function of Initial Conditions
by N. Arous, N. Ellouze

Abstract - In this paper, we propose a new hybridization method of self organizing maps (SOM) with genetic algorithms in order to design optimal quantified vectors of input speech signals. In particularly in the present work, we demonstrate that it is necessary to effectuate choice criteria of genetic operators concerning the way to evolve individuals. Based on the genetic algorithm, we determine which type of genetic operators is adequate to initial conditions and the objective to reach by guaranteeing the production of viable population converging towards similar performances. Experimental results show that the use of these operators depends on initial conditions and the objective to realize. The inter maps crossover allows decreasing the quantization error of a map. However, the intra map crossover allows decreasing the topological error of a map. Both recombination operators when applied successively aim at decreasing the distortion error of a map.

Keywords: Genetic algorithms, Genetic operators, Phoneme classification, Self-Organizing map, SOM variant.

Evolutionary Algorithms In The Optimization Of Greenhouse Climate Model Parameters

by A. Hasni, B. Draoui, T. Boulard, R. Taibi , A. Hazzab

Abstract - In the world, the greenhouse culture is getting more and more developed to meet the needs of a more competitive market that is conditioned by very strict norms of quality. Greenhouse production systems are getting highly sophisticated but extremely expensive. That is the reason why the greenhousers, who want to be competitive must optimize their investments thanks to strict production conditions. Nowadays, it is commonly admitted that the decisions related to the management of a greenhouse can be classified into different levels, starting from the on –line control, through the optimization of the environment, to the seasonal planning of the behaviour of the agriculture. The main purpose of our work is to optimize the physical sizes of a reduced model of a greenhouse. Thus, we use a digital simulation based genetic algorithm. Genetic algorithms are optimization algorithms based on techniques derived from genetics and natural development.

Keywords: Temperature, Pressure, Optimization, Geneticalgorithm, Greenhouse, Climate models.

Parallel MLTS Construction Preserving the Alpha Equivalence Reduction

by Z. El Abidine Bouneb, D. E. Saidouni

Abstract - In this paper we investigate an approach for the parallelization of the construction of the state space, which is modeled as Maximality Labeled Transition System, in fact , MLTS is a true concurrency model for process algebra it allow the design of critical system by refinement and also reduce the state space explosion problem, this last problem still limit the use of model checking for complex system even using MLTS, Many techniques have been developed to reduce the size of state space; the most important technique among them is the alpha equivalence reduction. This reduction helps improve the performance of Model checker based on the Maximality semantics by fusing many states respecting the alpha equivalence relation to one state. Distributed memory execution environment offers yet another choice. . The main contribution of the paper is to present a new algorithm combine the alpha equivalence reduction and parallel Maximality Labled Transition System construction , the main idea of this parallel algorithm is that the state graph is partitioned among the network nodes, each node executes an instance of the parallel generation algorithm which computes a reduced partial Maximality Labeled Ttransition System.

Keywords: Maximality semantics, model checking, LOTOS, state space construction, verification, distributed algorithm, true concurrency semantics, CTL logic.

Protein Secondary Structure Prediction: Application of RBF Networks
by B. Messabih, G. Khensous, N. Benamrane

Abstract - The Bioinformatics is a multidisciplinary research domain aiming the automatic treatment of the biologic information including the sequences (Multiple Sequence Alignment), the structures (protein’s secondary or tertiary structures), and the functions (protein structural motifs). One of the Bioinformatics challenges is the Protein Secondary Structure Prediction. The knowledge of the protein’s secondary structures is a subject in vogue which interests a large number of Biologists because it ensures the guess of the protein’s performance and evolution constituting therefore an interesting advance in the domain of drug design.
We present through this paper a method of prediction of different 2D structures of the protein from its amino acid sequences using Artificial Neural Networks and more precisely the Radial Basis Function Networks. To do so, we opted for the heuristic training technique; the Dynamic Decay Adjustment Algorithm which is applied on Rost and Sander data base containing 130 proteins. We could reach a Q3 score which is equal to 63.26 %, this score is better than the one gotten by Rost and Sander using Multi Layer Perceptron networks without introducing Multiple Sequences Alignments (Q3=61.7%).

Keywords: ANN, Chou and Fasman Classification, DDA Algorithm, MSA, Q3 Score, Rost and Sander Database.

Comparative Performance of Embedded Coders at High Quality
by D. Samai, N. Doghmane, M. Bedda

Abstract - The approach for still-image compression proposed by Said and Pearlman is designed initially for encoding the DWT coefficients. In this paper, the SPIHT scheme is used to encode the DCT coefficients. The DCT based scheme is computationally more efficient than the DWT based scheme. In order to improve the performances of our algorithm, we propose to increase the levels of resolution by increasing the DCT blocks size to 16×16, 32×32 and 64×64 instead of 8×8 as in standard JPEG. Experimented results show that the DCT based SPIHT provides high fidelity images, and it has been shown that these images are visually better than images reconstructed using other DCT-based coders for the same bit rate.

Keywords: Image coding, embedded coding, DWT, DCT, SPIHT.

A Method of Construction of Resilient Functions with Maximal Possible Nonlinearity
by A. Belmeguenaï, N. Doghmane

Abstract - A good design of Boolean function used in a stream cipher requires that the function satisfies certain criteria (balanced, high order correlation immunity, high algebraic degree and high nonlinearity) in order to resist different attacks. In this paper we give a method of construction of resilient functions with a maximal possible nonlinearity. We apply this construction to design more numerous resilient functions achieving the best possible trade-offs between resiliency order, algebraic degree and nonlinearity (that is, achieving Siegenthaler’s and Sarkar, al.’s bounds).

Keywords: Algebraic degree, Algebraic immunity, Boolean Functions, Nonlinearity, Resiliency, Stream Ciphers.

Real-Time Agent Computing Platform for Distributed Satellite Systems
by C. P. Bridges, T. Vladimirova

Abstract - The next generation of satellites for distributed satellite missions will exploit the latest computing and wireless technologies for intersatellite connectivity. These missions enable opportunities in multiple-point sensing, greater communications capabilities and spacecraft redundancy. Requirements for processing and network capabilities have risen dramatically to meet strict needs of the end user and overcome various space disturbances and perturbations once in orbit. One such problem lies with a ‘cluster’ of satellites that have been deployed from the same launcher where they will be close together so ad-hoc technologies allow satellite communication. This paper addresses the hardware and software requirements for distributed computing opportunities using intersatellite connectivity using Agent technologies. A complete review is also presented for Agents in real-time systems and their applications to show that they are still unsuited for mission critical networked and embedded systems. A system-on-a-chip design is proposed; including a general purpose processor core and a dedicated Java processing core to enable hard real-time Agent functionalities and software Agent applications. This will make the network resilient to various space perturbations and ensure mission longevity. Integration of these two non-heterogeneous cores into an FPGA technology demonstrator testbed is also outlined.

Keywords: Real-time, Agent Computing, Distributed Computing, Distributed Satellite System.

Remote Control Architecture over Internet Based on Multi agents Systems
by A. Sayouti, F. Qrichi Aniba and H. Medromi

Abstract - Internet is not only an information highway, but also a mean to remotely control mechanical systems, such as robotic devices. The Internet limits the quantity of information which can be transmitted and introduces delays which can make the remote control difficult or impossible. The solutions, proposed to face the limitations of the communication channel, are founded on the autonomy and the intelligence based on multi agents systems granted to the robot in order to interact with its environment and to collaborate with the remote user. In a first part of this paper, we are presenting the main interests of such a remote control and we are describing some existing applications. In a third part, we are proposing a system architecture to realize such remote control. In the fourth part, we are presenting our control architecture. An illustration of our architecture is given in an application of control of an autonomous mobile robot.

Keywords: Remote control, Internet, Multi agents systems, control architecture, robotic.

MoSAL: A Tile Logic Based Architecture Description Language for Mobile Software Applications
by C. Bouanaka, F. Belala

Abstract - Dynamicity and mobility modelling are complicating factors for distributed systems specification and development. Since, besides the usual computational evolution, the system is actually subject to dynamic reconfiguration changes. We propose a tile logic based architecture description language, MoSAL (Mobile Software Applications Language), as a common semantic framework for both system evolution dimensions. To maintain a separation of concerns between computation and dynamic reconfiguration changes, a layered model is presented: While core layer defines a component structure, the behavioural layer expresses its local possible evolutions over time. Essential primitives to manage components mobility are defined in the third layer via a set of tiles, an extended format of rewrite rules taking into account rewriting with side effects and rewriting synchronization. Interactions between behavioural layer and reconfiguration or mobility management one are ensured by the core layer via shared objects.

Keywords: Dynamic Software Architectures, Mobility, Tile Logic, Architecture Description Languages.