Bilal Chebaro

A session based personalized search using an ontological user profile

Within the information overload on the web and the diversity of the user interests, it is increasingly difficult for search engines to satisfy the user information needs. Personalized search tackles this problem by considering the user profile during the search. This paper describes a personalized search approach involving a semantic graph-based user profile issued from ontology. User profile refers to the user interest in a specific search session defined as a sequence of related queries. It is built using a score propagation that activates a set of semantically related concepts and maintained in the same search session using a graph-based merging scheme. We also define a session boundary recognition mechanism based on tracking changes in the dominant concepts held by the user profile relatively to a new submitted query using the Kendall rank correlation measure. Then, personalization is achieved by re-ranking the search results of related queries using the user profile. Our experimental evaluation is carried out using the HARD 2003 TREC collection and shows that our approach is effective.

Learning implicit user interests using ontology and search history for personalization

The key for providing a robust context for personalized information retrieval is to build a library which gathers the long term and the short term users interests and then using it in the retrieval process in order to deliver results that better meet the users information needs. In this paper, we present an enhanced approach for learning a semantic representation of the underlying users interests using the search history and a predefined ontology. The basic idea is to learn the users interests by collecting evidence from his search history and represent them conceptually using the concept hierarchy of the ontology. We also involve a dynamic method which tracks changes of the short term users interests using a correlation metric measure in order to learn and maintain the users interests.

Mining for video production invariants to measure style similarity

This article focuses on video document comparison using audiovisual production invariants (API). API are characterized by invariant segments obtained on a set of low‐level features. We propose an algorithm to detect production invariants throughout a collection of audiovisual documents. The algorithm runs on low‐level features, considered as time series, and extracts invariant segments using a one‐dimensional morphological envelop comparison. Then, based on the extracted results, we define a style similarity measure between two video documents. A derivative pseudo distance is also proposed.

Proposition d'un système de RI personnalisé à base de sessions intégrant un profil utilisateur sémantique

L’objectif de la recherche d’information (RI) personnalisee est de repondre mieux aux besoins en informations de l’utilisateur tout en integrant son profil dans la chaine d’acces a l’information. Les principaux defis en RI personnalisee concernent la modelisation du profil utilisateur et son exploitation dans le processus de recherche. Ce papier presente une conception et une evaluation d’un systeme de RI personnalise integrant un profil utilisateur semantique. Le profil utilisateur est represente selon un graphe de concepts issu d’une ontologie de reference, l’ODP. Il est construit le long de requetes correlees et est utilise dans le reordonnancement des resultats. Nous avons evalue notre systeme sur deux collections TREC differentes et avons montre une amelioration significative de la RI personnalisee par rapport a la RI classique.

Style similarity measure for video documents comparison

We define a style similarity measure for video documents based on the localization of common elements and on the temporal order in which they appear in each document. Common elements for a couple of compared videos are segments presenting similar behaviors on a subset of low or mid level features extracted for the comparison process. We propose a method to compare two video documents and to extract those similar elements using dynamic programming and one-dimensional morphological operations. The similarity measure is applied on TV-news broadcast to illustrate its behavior.

Bayesian Trust Scheme: A Decentralized Safety Message Trust Method in Multi-Hop V2V Networks

Abstract Vehicular Ad hoc Networks (VANETs) have an important role in improving road safety especially when no infrastructure is available. Inter vehicle communications (IVCs) provide decentralized communications where vehicles cooperate together to disseminate road traffic data relying on broadcasting reports and warning messages. Trusting the data of every warning message must be accomplished during dissemination as its information is public. Many researches have concentrated on securing the system entities by adding authenticity to each vehicle or aggregating digital signature. These traditional security schemes at some point needed a central management. In this paper, we introduce an approach to trust the information of the warning message before disseminating it through multi-hop V2V communications. This approach is a decentralized scheme that relies on evaluating random environment variables and their conditional dependencies using Bayesian Network (BN). Our contribution is depending on two-stage decentralize data trusting scheme that the warning message passed through before forwarding it further to avoid nodes from acting maliciously.

A Framework for Disseminating Trusted Safety Message in V2V Communication

VANET (Vehicular Ad hoc Networks) has become recently one of the most interesting areas of researches in the world of wireless networks. This paper reviews important aspects in VANET as well as commonly encountered problems. Our goal is to present a model to disseminate a warning message properly between vehicles and make trust decision to ensure message integrity before rebroadcast it farther. Our model enhances Vehicle to vehicle communications by avoiding three of the VANETs problems, the broadcast storm problem, the hidden node problem, and incoming message collision at the recipient nodes. A comparison had been made with the efficient directional broadcast (EDB) protocol for message disseminating to evaluate our model results.

Toward using data trust model in VANETs

VANET exposed too many data threats due to vehicles short communication time given to cooperate and exchange public information. Using message trust schemes helps in preventing a malicious message from an authorized user to be disseminated between vehicles. In this paper, we introduce security vulnerabilities during exchange safety message between vehicles. Then we propose a suggested data trust algorithm based on hypotheses that help in building a data trusted scheme.

A Web Services Provider

In this paper, we define a generic tool ‘Generic Serv’ that offers a’ service providers’ platform which facilitates the programming tasks of web application development. The system’s architecture is generalized to propose three patterns for business applications’ development. The paper is divided into two major parts. In the first one, we expose the motivation for the definition of the service provider, where we emphasize the architecture of the system and the arguments to choose such architecture. In the second part, we define the patterns based on this generic service provider.