Research Reports

Permanent URI for this collectionhttp://dl.cerist.dz/handle/CERIST/34

Browse

Author: search.filters.author.Nouali, OmarEnd date: 2009

Search Results

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    Content-Oriented XML Retrieval with partial aggregation of relevance factors
    (CERIST, 2009-12) Bal, Kamal; Nouali, Omar
    In this paper, we are interested in content oriented XML information retrieval whose aim is to retrieve not a list of relevant documents, but only fragments of document (XML element) relevant to the user information need. Retrieved XML elements must not only contain relevant information but also be at good level of granularity. Retrieved elements must cover as well as the information need and be focused on this need (does not speak other non relevant topics). The majority of approaches developed in this context consider some relevance factors and aggregate theme globally to have a unique value called rsv (retrieval status value) for each element. We present in this paper an approach of XML retrieval where factor’s aggregation is done not on relevance factor’s values level but on relevance factor’s preference relations which models each factor.
  • Thumbnail Image
    Item
    Filtrage collaboratif par le web sémantique : Framework de similarités sémantiques
    (CERIST, 2009-09) Nouali, Omar
    L’adoption des systèmes de filtrage et de recommandation est assez importante aujourd’hui. Ils ont une importance majeure sur le web d’aujourd’hui en général et le e-business en particulier. Cependant ces systèmes souffrent de problématiques liées au démarrage à froid (au nombre peu important d’évaluations, au nouvel utilisateur, nouvelle ressource..). Aujourd’hui, le challenge est l’amélioration des pratiques et méthodes utilisées pour rendre ces systèmes plus précis, interactifs, adaptés aux contextes et performants. Dans cet article, nous présentons une approche qui utilise l’infrastructure «web sémantique» dans le but d’améliorer la qualité des systèmes de recommandation en termes de précision et de couverture. Nous proposons un Framework qui permet le calcul des similarités sémantiques entre entités (utilisateurs et ressources) de façon plus facile et plus précise. Pour la validation, nous avons mené un ensemble d’expériences pour évaluer les performances de notre approche de filtrage et du framework proposé et développé.
  • Thumbnail Image
    Item
    Using semantic Web to reduce the colt-sart
    (CERIST, 2009-06) Nouali, Omar; Belloui, Amokrane
    Collaborative filtering systems suffer from the cold-start problems (evaluation matrix, new user/new resource problem...). In this paper, we show that using semantic information describing users and resources can reduce the problems and lead to a better precision, coverage and quality for the recommendation engine. Semantic web is the infrastructure used for managing such semantic descriptions. We also present here the results of a set of evaluation experiments.
  • Thumbnail Image
    Item
    A basic platform of collaborative filtring
    (CERIST, 2008) Nouali, Omar; Kirat, Sabah; Meziani, Hadjer
    With the explosive growth of the quantity of new information the development of information systems to target the best answers provided to users, so they are closer to their expectations and personal taste, has become an unavoidable necessity. The collaborative filtering systems are among these information systems with particular characteristics that make the difference. The term refers to collaborative filtering techniques using the familiar tastes of a group of users to predict the unknown preference of a new user. This article describes a basic platform of collaborative filtering, which allows users to discover interesting documents, through automation of the natural process of recommendation, it allows them to express their opinion about the relevance of documents, according to their tastes and documents’ quality they perceive; it offers the opportunity to benefit from the evaluations on documents of other users, with similar profile, have found interesting. All these benefits are provided to users by the principle of collaboration, in return for an individual effort: evaluating documents.