Université Constantine 2
Abdelhamid Mehri University Constantine2 Scholarlyworks RepositoryNot a member yet
128 research outputs found
Sort by
A Dual PSO-Adaptive Mean Shift for Preprocessing Optimization on Degraded Document Images
No full textIn the two past decades, solving complex search and optimization problems with bioinspired metaheuristic algorithms has received considerable attention among researchers. In this paper, the image preprocessing is considered as an optimization problem and the PSO (Particle Swarm Optimization) algorithm was been chosen to solve it in order to select the best parameters. The document image preprocessing step is the basis of all other steps in OCR (Optical Character Recognition) system, such as binarization, segmentation, skew correction, layout extraction, textual zones detection and OCR. Without preprocessing, the presence of degradation in the image significantly reduces the performance of these steps. The authors' contribution focuses on the preprocessing of type: smoothing and filtering document images using a new Adaptive Mean Shift algorithm based on the integral image. The local adaptation to the image quality accelerates the conventional smoothing avoiding the preprocessing of homogeneous zones. The authors' goal is to show how PSO algorithm can improve the results quality and the choice of parameters in pre-processing's methods of document images. Comparative studies as well as tests over the existing dataset have been reported to confirm the efficiency of the proposed approach.080
Towards A Generic Predictive-Based Plan Selection Approach for BDI Agents
No full textBelief-desire-intention (BDI) model is one of the most promising approaches dedicated to develop intentional agents. It provides developers with required robustness to deal with a wide variety of complex systems where flexibility and intelligence behavior of agents are needed. Existing BDI architectures are typically composed of a set of customizable components that are composed in a reasoning cycle. The plan selection process is one of the customizable components that has an important influence on the overall agent performance. It consists of selecting among a set of applicable plans the most appropriate one to achieve a given goal. In this paper we propose a generic predictive-based plan selection approach, that aims to select for a given goal, the applicable plan that has the highest potential of being executed successfully. Based on what we refer to as selection indices, it calculates for each plan a success probability that will be used later by the selection process to select the appropriate plan.113414
Institutional digital repositories in Algerian universities: Preparation of a mechanism for the construction and the implementation of the institutional repository of Abdelhamid Mehri University Constantine 2
No full textThis study is aimed at examining and identifying the digital institutional repositories in the Algerian universities which were counted by the OpenDOAR and estimated at 12 institutional repositories, by examining indicators related to the subjective distributions, information retrieval, Metadata description methods, additional services, software used, digital repositories deposit policies, and copyright management, as well as the establishment of a mechanism to build the institutional repositories to guide the university of Abdelhamid Mehri Constantine 2 in the development of institutional repository for the benefits of faculty authors, electronic theses and dissertation were the most resources available in the repositories under study by 91%, the digital repositories of the study were based on the open source software “Dspace” by 83%, and the study found lack of some policies in most of the repositories under the study.تأتي هذه الدراسة من أجل فحص وتحديد المستودعات الرقمية المؤسساتية الموجودة بالجامعات الجزائرية والتي تم إحصاؤها من طرف دليل المستودعات الرقمية ذات الوصول الحر OpenDOAR والمقدرة بـ 12 مستودعا رقميا مؤسساتيا وذلك من خلال فحص مؤشرات متعلقة بالتوزيع الموضوعي والزمني، طرق إسترجاع المعلومات، وطرق عرض بيانات الوصف لمصادر المعلومات بالمستودعات الرقمية، الخدمات الإضافية، البرمجيات المستخدمة في بناء المستودعات الرقمية، سياسات الإيداع بالمستودعات الرقمية، إدارة المحتوى وإدارة حقوق النشر. وكذا وضع ألية لبناء المستودعات المؤسساتية لتسترشد بها جامعة عبد الحميد مهري قسنطينة 2 في تطوير مستودعها المؤسساتي لفائدة أعضاء هيئة التدريس، وقد كانت الرسائل الجامعية الإلكترونية أكثر المصادر إتاحة بالمستودعات موضوع الدراسة بنسبة 91%، و اعتمدت المستودعات الرقمية موضوع الدراسة على برمجية Dspace المفتوحة المصدر بنسبة 83%، كما توصلت الدراسة الى انعدام بعض السياسات لدى أغلب المستودعات المؤسساتية موضوع الدراسة.181-213216
Toward an Efficient Ambient Guidance for Transport Applications
No full textWe address smart transportation systems whose computational process is based on autonomous and context-aware intelligent agents. As a use case, a pharmacy round tour application is presented, then an original guidance algorithm is proposed to handle the various intentions to be executed during a delivery mission. The efficiency of the algorithm relates to a combined approach involving planning, learning and physical path optimizations. In particular, the past experiences of actions are used to better select the plan which guides the driver contextually. Moreover, the move actions in the plan are concretely refined on shortest physical paths taking into account the spatio-temporal evolution of the driving context
Checking Behavioural Compatibility in Service Composition with Graph Transformation
No full textThe success of Service Oriented Architecture (SOA) largely depended on the success of automatic service composition. Dynamic service selection process should ensure full compatibility between the services involved in the composition. This compatibility must be both on static proprieties, called interface compatibility which can be easily proved and especially on behavioural compatibility that needs composability checking of basic services. In this paper, we propose (1) a formalism for modelling composite services using an extension of the Business Process (BP) modelling approach proposed by Benatallah et al. and (2) a formal verification approach of service composition. This approach uses the Graph Transformation (GT) methodology as a formal verification tool. It allows behavioural compatibility verification of two given services modelled by their BPs, used as the source graph in the GT operation. The idea consists of (1) trying to dynamically generate a graph grammar R (a set of transformation rules) whose application generates the composite service, if it exists, in this case (2) the next step consist in checking the deadlock free in the resulting composite service. To this end we propose an algorithm that we have implemented using the AGG, an algebraic graph transformation API environment under eclipse IDE.070
Individual Classification: an Ontological Fuzzy Based Approach
No full textRecently, serval reasoners for very expressive fuzzy Description Logics have been implemented. However, in some cases, applications do not require all the reasoner services and would benefit from the efficiency of just certain reasoning tasks. To this scope, we are interested in the individual fuzzy classification issue. In fact, decision-making applications for real world domain is often based on classifying new situations into fuzzy categories. Therefore, we propose Fuzzy Realizer to offer an effective classification even with imprecise/vague or incomplete knowledge so that appropriate decision can be made. Fuzzy Realizer is a Java prototype implementation for realizing fuzzy ontologies. It supports the well-known fuzzy description logic Z SHOIN (D). It allows (i) fuzzy concrete domains, (ii) modified and (iii) weighted concepts. It is able to (i) classify new individuals, even with incomplete descriptions, (ii) provide a more human-oriented classification by hiding the crisp boundaries between different fuzzy categories and (iii) to populate fuzzy ontologies which address an aspect of fuzzy ontologies evolution, a topic which is rarely discussed.410
Un nouveau formalisme pour la représentation et le raisonnement sur une ontologie hybride
No full textLes moteurs de recherche à base d’ontologie utilisent un raisonnement sémantiques, afin d’apporter à l’utilisateur les informations pertinentes dont il a besoin. En général, les ontologies utilisées dans la phase de raisonnement de ses moteurs sont présentées en OWL, mis vu les insuffisances de ce langage, il serait intéressant d’enrichir les logiques de description sur lesquelles il se base, avec d’autre paradigmes donnant ainsi naissance des ontologies hybrides. Ces paradigmes permettront d’augmenter l’expressivité d’OWL. Les moteurs d’inférence sur ces ontologies hybrides ne sont pas intégrés dans le classifieurs OWL. Le problème qui se pose alors, est l’absence des raisonneurs effectifs capable de faire interopérer les ontologies OWL et les autres paradigmes. Dans ce travail de recherche, Nous proposons SWRLx une extension du langage SWRL pour créer des règles intra-concept et nous proposons une approche hybride pour combiner le raisonnement sur le langage OWL et le raisonnement sur les règles SWRL dans un système d’interprétation des bilans de santé en médecine préventive. La méthode de raisonnement proposé permet de convertir les règles SWRLx du format OWL au format Jess. Nous nous basons sur les standards et les outils actuels du web sémantique en vue d’augmenter, en termes de qualité des résultats, les performances du système développéOntology-based search engines use a semantic reasoning, to provide to the user the relevant information they need it. In general, ontologies used in the reasoning phase of its engines are presented in OWL, set saw the shortcomings of this language, it would be interesting to enrich the description logics on which it is based, with other paradigms giving birth hybrid ontologies. These paradigms will increase the expressiveness of OWL. The inference engines on these hybrid ontologies are not integrated in the OWL classifiers. The problem is the lack of reasoning which able to interoperate OWL ontologies and other paradigms. In this research, we propose SWRLx (Extended Semantic Web Rule Language) as an extension to the SWRL rules. The set of rules obtained with SWRLx are posted to the Jess engine using rewrite metarules. The reason for this combination is that it allows the inference of new knowledge and storing it in the knowledge base. We propose a formalism for SWRLx along with its implementation through an adaptation of different object-oriented techniques. The Jess rule engine is used to transform these techniques to the Jess model. We include a demonstration that demonstrates the importance of this kind of reasoning. In order to verify our proposal, we use a case study inherent to interpretation of a preventive medical check-up
The developmental role of the stock market and the elements of Algiers stock market to achieve this role
No full textتؤدي أسواق الأوراق المالیة وظائف اقتصادیة متمیزة في تعبئة المدخرات وتوجیهها نحو الاستثمارات المناسبة والجاذبة، وكما هو متعارف علیه لا یخلو عمل أسواق الأوراق المالیة من مشاكل وصعوبات في عملها تعیقها عن تحقیق أهدافها خصوصا في ظل العولمة والأزمات المالیة التي تعیشها، وهذا ما تواجهه الأسواق المالیة العربیة في الوقت الحاضر، وقد تجسدت أهمیة البحث في أهمیة سوق الأوراق المالیة في الاقتصادیات المعاصرة لمساهمتها الفعالة في إنجاح الجهود الرامیة إلى تعزیز عملیة التنمیة الاقتصادیة وتحسین المناخ، ولهذا یجب على أسواق الأوراق المالیة العربیة وخاصة بورصة الجزائر الوفاء بمتطلبات قیام الأسواق المالیة الناجحة والكفؤة ومحاولة التوصل إلى التطورات الحاصلة في الأسواق الناجحة وهذا ما یهدف له البحثstock markets perform active and distinctive economic functions in mobilizing savings and orienting them towards suitable and appealing investments. It is already known that the operation of stock markets is not void of difficulties and obstacles that hinder it from achieving its objectives especially under the conditions of economic globalization and financial crises facing it, and this is the situation facing Arab financial markets at present. The importance of the research has emerged from the significance of financial markets in contemporary economies for its effective contribution in fulfilling efforts of enhancing the process of economic development and improving the investment climate. Thus; financial markets- Algeria stock market, in particular, should meet the requirements of establishing efficient financial markets and seeking to cope up with current developments in international financial markets; which is aimed to by the research.
Big data integration: A semantic mediation architecture using Summary
No full textThis paper presents our semantic mediation architecture for homogeneously retrieving data from big data. The architecture is split into three layers to find relevant answers of this data. In the proposed ontology-based approach, we use our domain ontology on alimentation risks field as well as a global schema of mediator and we apply the summarization process based on ontology for describes part of the data set. Thus, the auto evolution of ontology is based on this huge amount of data. The research of answer is related to relevant summaries which represent high-level semantics
Application de nouvelles approches d’optimisation pour la biométrie
No full textFace aux failles de sécurité au niveau des méthodes traditionnelles conçues pour l’établissement et la gestion de l’identité et qui font recourir aux nouvelles technologies pour des fins d’encerclement, la recherche ne peut rester indifférente aux conséquences adverses engendrées. La biométrie désigne une technologie qui ne permet pas seulement de faire autrement des taches existantes, comme l’authentification, mais aussi d’accomplir pour la première fois sur une grande échelle des taches d’identification. En effet, cette technologie tourne autour d’une reconnaissance d’un individu à base physio-biologique ou/et comportementale qui répond aux exigences des architectures et attendues des systèmes de sécurité dit systèmes biométriques, grâce aux facteurs de la non répudiation et la reconnaissance négative. Particulièrement, les systèmes biométriques à base de l’iris et l’empreinte digitale sont les plus répandus grâce à leurs performances et robustesses en termes de leurs taux de performances et crédits publiques élevés. Toutefois, il existe quelques conditions orientées environnement ou / et capteurs ou / et utilisateurs à interroger et qui violent l’aspect impeccable des attendus de ces systèmes. Le travail de thèse réalisé à travers de ce manuscrit présente deux approches solutions pour l’amélioration de deux processus critiques, notamment : le processus de segmentation de l’iris et le prétraitement des images des empreintes digitales. Par conséquent, une amélioration des taux de performances des systèmes biométriques à base de l’iris et de l’empreinte digitale est observée. La première approche consiste en une pré-segmentation de l'image de l'iris à base d’un seuillage à multi-niveaux qui sera injectée dans le processus basique de localisation de l’iris. Le seuillage à multi-niveaux est optimisé par des méthodes approchées, à savoir, l’optimisation par essaim de particules (PSO), la recherche coucous (CS), la colonie d’abeilles artificielle (ABC), et la colonie d’abeilles artificielle guidée par le meilleur (GABC). Dans la deuxième contribution, un prétraitement des images des empreintes digitales est couvert à travers une amélioration du contraste de ces images basée sur l’utilisation des méta-heuristiques pour une mise en correspondance et une définition de nouveaux niveaux de gris au niveau de ces images. Il s’agit de l’algorithme de la colonie d’abeilles artificielle (ABC), la recherche coucous (CS), et l’algorithme de chauves-souris (BA). L’examen expérimental à travers l’étude de différents seuils a permis la sélection des paramètres modélisant les solutions les plus optimales et les résultats obtenus apportent une amélioration remarquable au niveau des processus traités et de la performance totale des systèmes concernés.In front of the vulnerabilities of the traditional security methods particularly designed for establishing and managing the identity concept and which require new technologies for encircling endings, research cannot remain indifferent to the related negative consequences. Biometrics represents a technology which does not only perform differently some existing tasks, such as authentication, but also it allows the accomplishment of identification processes for the first time on a large scale. Indeed, this technology turns around a personal recognition of individuals based on their physio-biological and/or behavioral patterns taking into consideration the architectural requirements of the concerned security systems called biometric systems thanks to the non-repudiation and the negative recognition factors. Among these biometric systems, the iris and fingerprint-based recognition systems are particularly considered as the most widespread due to their performances and robustness in terms of their high-performance rates and public credibility. However, there exist some environmental and/or sensor-based and/or user-based conditions which violate the impeccable expectations of these systems potential. Through this manuscript, the work carried within this thesis represents two solution approaches to improve two critical basic processes, namely, the process of iris segmentation, and that of the preprocessing of fingerprint images. Hence and as a result, an improvement in the performance rates of the iris-based and fingerprint-based biometric systems is observed. The first approach consists of a pre-segmentation of the iris images based on a multilevel thresholding which is to be injected in the basic iris localization process. The multilevel thresholding is optimized using the meta-heuristics approximate methods, namely, the Particle Swarm Optimization (PSO), the Cuckoos Search (CS), the Artificial Bee Colony (ABC), and the G-Best guided Artificial Bee Colony (GABC). In the second contribution, a preprocessing of the fingerprint images is conducted through an enhancement of the contrast based on the use of meta-heuristics for defining new levels of gray in these images. We have used particularly, the Artificial Bee Colony algorithm (ABC), the Cuckoos Search (CS), and the Bat Algorithm (BA). The experimental tests and studies drawn about the different thresholds allow the optimal solution parameters, and the obtained results provide a significant improvement at the level of both the treated processes and the whole overall accuracies of the concerned systems