Amara M.,SOIE Laboratory |
Ghedira K.,SOIE Laboratory |
Zidi K.,University of Tabuk |
Zidi S.,Qassim University
Proceedings of IEEE/ACS International Conference on Computer Systems and Applications, AICCSA | Year: 2016
Support Vector Machines (SVM) is a statistical classification approach which has been successfully applied to various types of problems. However, it has remained largely unexplored for Arabic recognition. SVMs are originally designed for binary classification problems. For multi-class problems, several methods used a combination of binary SVMs while some others solved the problem in one step. This paper introduces an evaluation of five SVM methods for the Arabic characters recognition problem; three are based on binary classifiers, and two considers all classes at once. The selected algorithms are compared in terms of training time, testing time and accuracy. Experiments conducted using the Arabic Printed Text Image Database-Multi-Font(APTID/MF ) showed that the 'one-against-one method' is the robust, fast and produces a very good score rate at a reasonable time. © 2015 IEEE.
Bouyahia Z.,SOIE Laboratory |
Bellalouna M.,SOIE Laboratory |
Ghedira K.,High Institute of Management
Computers and Industrial Engineering | Year: 2013
Abstract In a previous work (Bouyahia et al., 2010), we introduced and studied the probabilistic generalization of the classical total weighted completion time problem. We defined two a priori strategies devoted to the incorporated problem on parallel identical machines (UA and UB). The main drawback of the proposed a priori strategies is that the machine loads are not balanced after deleting the absent jobs from an a priori schedule. In this paper, we try to overcome this limitation by proposing an a priori strategy denoted as Uk which takes into account the load balance. First, we present the algorithm of the load balancing strategy and we explicit the objective function. Then, we study its complexity and its performance compared to the other strategies. Special care is also devoted to assess experimentally the performance of Uk. © 2012 Elsevier Ltd. All rights reserved.
Miled A.B.,SOIE Laboratory
Procedia Computer Science | Year: 2014
The work presented in this paper is related to knowledge management during the design process of mechanical products. More exactly, we are concentrating on the knowledge reuse process. We propose a knowledge reuse system (KRS) based on two approaches. Namely, an organizational approach and a modeling user approach. Both are ontology-based models. Our system takes into account, the actors' roles (by the organizational approach), their preferences (by the model user approach), and their collaboration throughout the product design process in order to help designers to get knowledge that is considered the most relevant in all the phases of the product design process. © 2014 The Authors. © 2014 Published by Elsevier B.V.
Monfort V.,University of Paris Pantheon Sorbonne |
Monfort V.,SOIE Laboratory |
Abbes H.,ISIM Pole Technologique de Sakiet Ezzit |
Abbes H.,MIRACL Laboratory |
And 2 more authors.
ICSOFT 2011 - Proceedings of the 6th International Conference on Software and Database Technologies | Year: 2011
Service-Oriented Architectures (SOA) are broadly used by companies to gain in flexibility. Web service is the fitted technical solution used to support SOA by providing interoperability and loose coupling. First research works allowed us developing genuine middleware based on a new paradigm called aspect based services to dynamically adapt services orchestration. This approach can also be used for SPARQL (Protocol and RDF Query Language) requests to dynamically access Web services ontologies. According to these pre requisites, this paper presents our approach, current and future research works.