International Journal Papers

Permanent URI for this collectionhttp://dl.cerist.dz/handle/CERIST/17

Browse

End date: 2009Has files: Yes

Search Results

Now showing 1 - 10 of 27
  • Thumbnail Image
    Item
    A distributed mutual exclusion algorithm over multi-routing protocol for mobile ad hoc networks
    (Taylor et Francis, 2008-04-15) Derhab, Abdelouahid; Badache, Nadjib
    In this paper, we propose a new architecture to solve the problem of mutual exclusion in mobile ad hoc networks (MANET). The architecture is composed of two layers: (i) a middleware layer that contains a token-based distributed mutual exclusion algorithm (DMEA) and (ii) a network layer that includes two routing forwarding strategies: one to route request messages and the other to route the token message. We also propose a request ordering policy that ensures the usual mutual exclusion properties and reduces the number of hops traversed per critical section (CS) access. The paper also addresses the problem of network partitioning and unreachable nodes. The proposed mutual exclusion algorithm is further enhanced to provide fault tolerance by preventing the loss of the token and generating a new token if the token loss event occurs. The performance complexity as well as the experimental results show that the proposed algorithm experiences low number of hops per CS access.
  • Thumbnail Image
    Item
    Building Digital Collections in the Algerian Context: Assets and Prospects
    (Karen I. MacDonald, 2002-06) Bakelli, Yahia
    When we analyze the access behavior of Algerian users of information and documentation, we see a certain number of dysfunctions leading to a situation of imbalance between user demand on the one hand and library and document center access on the other. The delivery mechanisms of the information suppliers are slow and less e⁄cient, as we can deduce from surveys and studies, such as those of R. Carlier (1993); A. Geronimi (1993); Issolah (1992); Akrour (1999); Belhabib (1999); and Dahmane (2000).l To compensate for this inadequacy in the local information context, the CERIST Research Center on Scienti¢c and Technical Information (www.cerist.dz) is producing and maintaining several bibliographic catalogs and databases, including the following: The automated National Bibliography and National Union Catalogs (CAP, CAT, FNT), as well as the national databases of ‘‘Algerianna’’, ‘‘Algerie Science Information (ASI)’’ and ‘‘Algerian Scienti¢c Abstracts (ASA)’’. These catalogs and databases are available via the Internet through the www.dctd.cerist.dz Website. An example of a search session (in the National Bibliography) is given in Figs. 1^3: However, we are not sure if these bibliographic ¢les and catalogs are used and exploited by local users in an optimum way, especially in light of the problem of Internet connectivity. Also, we should consider if the contents of these bibliographic tools are su⁄cient to meet the new users’ needs. During the same search session there is indeed a tendency to require not only bibliographic information but the possibility of getting the full content of identi¢ed documents. Also, we must ponder the following question: Which add-value are these bibliographic tools assigning Mots Clés:
  • Thumbnail Image
    Item
    Balancing the tradeoffs between scalability and availability in mobile ad hoc networks with a flat hashing-based location service
    (2008-06) Derhab, Abdelouahid; Badache, Nadjib
    In this paper, we propose FSLS (Flat-based Some-for-some Location Service), a new location service for ad hoc mobile networks. The location service is based on the hash-based sets system that can offer a high location information availability. The network area is divided into non-overlapping zones. A node identifier is mapped to a set of home zones, each of which contains a unique location server, which makes FSLS works as a some-for-some approach. Using cross-layer design, the service can tolerate server mobility and server failures, and last for a long time period. We analyze FSLS and six other existing location services. The theoretical analysis as well as simulation results show that FSLS offers a good trade-off between location availability and scalability. It comes second after a quorum-based location service in terms of location availability and it is the closest competitor to a hierarchical location service in terms of scalability.
  • Thumbnail Image
    Item
    Self-stabilizing algorithm for high service availability in spite of concurrent topology changes in ad hoc mobile networks
    (Elsevier, 2008-06) Derhab, Abdelouahid; Badache, Nadjib
    Mobile nodes in ad hoc networks move freely and run out of battery power so quickly, which leads to frequent network partitioning. Network partitioning considerably reduces service availability when the server node is not in the same partition as the client nodes. In order to provide a continuous service availability for all mobile nodes, we propose a self-stabilizing algorithm that can tolerate multiple concurrent topological changes and can incur a cost of one server per long-lived connected component. By using (1) the time interval-based computations concept that distinguishes between disjoint and concurrent computations, and (2) Markov chain model, the proposed algorithm can within a finite time converge to a legitimate state even if topological changes occur during the convergence time. Our simulation results show that the algorithm can ensure very high service availability, and each node has a strong path to the server of its network component over 98% of the time.
  • Thumbnail Image
    Item
    Filtrage automatique de courriels Une approche adaptative et multiniveau
    (Springer-Verlag, 2005-12-01) Nouali, Omar; Blache, Philippe
    Cet article propose un système de courriers électroniques paramétrable avec plusieurs niveaux de filtrage: un filtrage simple basé sur l’information contenue dans l’entête du courriel; un filtrage booléen basé sur l’existence ou non de mots clés dans le corps du courriel; un filtrage vectoriel basé sur le poids de contribution des mots clés du courriel; un filtrage approfondi basé sur les propriétés linguistiques caractérisant la structure et le contenu du courriel. Nous proposons une solution adaptative qui offre au système la possibilité d’apprendre à partir de données, de modifier ses connaissances et de s’adapter à l’évolution des intérêts de l’utilisateur et à la variation de la nature des courriels dans le temps. De plus, nous utilisons un réseau lexical permettant d’améliorer la représentation du courriel en prenant en considération l’aspect sémantique.
  • Thumbnail Image
    Item
    Classification de courriers électroniques : Une approche par apprentissage basée sur des modèles linguistiques
    (Lavoisier, Cachan cedex FRANCE, 2005) Nouali, Omar; Blache, Philippe
    Nous proposons une double amélioration des systèmes de filtrage de courriels existants. D’une part, en utilisant une méthode d’apprentissage automatique permettant à un système de filtrage d’élaborer des profils utilisateur. D’autre part, nous utilisons un ensemble de connaissances linguistiques sous forme de modèles réduits issues de modèles linguistiques de textes. Dans ce contexte, nous cherchons à évaluer si l’utilisation de connaissances et de traitements linguistiques peut améliorer les performances d’un système de filtrage. En effet, nous utilisons, au-delà des caractéristiques lexicales, un ensemble d’indicateurs sur le message portant sur la structure et le contenu. Ces connaissances sont indépendantes du domaine d’application et la fiabilité repose sur l’opération d’apprentissage. Pour tenter de statuer sur la faisabilité de notre approche et d’évaluer son efficacité, nous l’avons expérimenté sur un corpus de 1 200 messages. Nous présentons les résultats d’un ensemble d’expériences d’évaluation
  • Thumbnail Image
    Item
    Automatic Classification and Filtering of Electronic Information: Knowledge-Based Filtering Approach
    (Zarqa Private University, Jordan, 2004) Nouali, Omar; Blache, Philippe
    In this paper we propose an artificial intelligent approach focusing on information filtering problem. First, we give an overview of the information filtering process and a survey of different models of textual information filtering. Second, we present our E-mail filtering tool. It consists of an expert system in charge of driving the filtering process in cooperation with a knowledge-based model. Neural networks are used to model all system knowledge. The system is based on machine learning techniques to continuously learn and improve its knowledge all along its life cycle. This email filtering tool assists the user in managing, selecting, classify and discarding non-desirable messages in a professional or non-professional context. The modular structure makes it portable and easy to adapt to other filtering applications such as web browsing. The performance of the system is discussed.
  • Thumbnail Image
    Item
    A Semantic vector space and features-based approach for automatic information filtering
    (Elsevier, 2004) Nouali, Omar; Blache, Philippe
    With advances in communication technology, the amount of electronic information available to the users will become increasingly important. Users are facing increasing difficulties in searching and extracting relevant and useful information. Obviously, there is a strong demand for building automatic tools that capture, filter, control and disseminate the information that will most likely match a user's interest. In this paper we propose two kinds of knowledge to improve the efficiency of information filtering process. A features-based model for representing, evaluating and classifying texts. A semantic vector space to complement the features-based model on taking into account the semantic aspect. We used a neural network to model the user's interests (profiles) and a set of genetic algorithms for the learning process to improve filtering quality. To show the efficacy of such knowledge to deal with information filtering problem, particularly we present an intelligent and dynamic email filtering tool. It assists the user in managing, selecting, classifying and discarding non-desirable messages in a professional or non-professional context. The modular structure makes it portable and easy to adapt to other filtering applications such as the web browsing. We illustrate and discuss the system performance by experimental evaluation results
  • Thumbnail Image
    Item
    Vectorial Information Structuring for Documents Filtering and Diffusion
    (Zarqa Private University, 2008) Nouali, Omar; Krinah, Abdelghani
    Information retrieval tries to identify relevant documents for an information need. The problems that an IR system should deal with include document indexing (which tries to extract important content from a document), user needs analysis (similar to document indexing but applied to a query), and their internal representation which makes them suitable for being explicitly manipulated by the corresponding algorithms (i.e., matching the query with the documents). This paper describes a vectorial approach for information organization, and its application to search/retrieval systems from a vast amount of textual data
  • Thumbnail Image
    Item
    Long-term preservation of electronic theses and dissertations in Algeria
    (De Gruyter, Berlin, ALLEMAGNE, 2003) Bakelli, Yahia; Benrahmoun, Sabrina
    In accordance with a decree issued by the Ministry of Higher Education and Scientific Research in Algeria in August 2000, an electronic copy of every Master's and PhD thesis defended in all academic institutions must be deposited at the CERIST Research Centre. Deposit is a condition for getting the diploma. CERIST is then entrusted with the mission to build a database of Algerian theses and to update the national inventory of current theses and research. However a serious problem of archiving and preserving these Electronic Theses and Dissertations (ETDs) has emerged. From December 2001 to November 2002 a great number of ETDs has been deposited and constitutes a set of more than 1000 floppy disks and 100 CD-ROMs. What guarantees that these digital materials deposited by students are preserved and safeguarded? What guarantees that the content of these materials are preserved and accessible at any time regardless of machine, operating system and software. This paper explores the problem of the long-term conservation and preservation of electronic theses in the Algerian context, and shows how international recognised standards and techniques for setting up and organising the local ETD's archives may be applied.