Users in a campus need information about relevant individuals, buildings, events and available resources. In this paper, we propose a system to perform situation-aware adaptive recommendation of information to assist mobile users in a campus environment. The idea is to show information about the most relevant buildings and particular individuals situated nearby the user, taking into account the user situation, i.e., a snapshot of his/her context in a given instant, including his/her current activity, position and profile. This recommendation must be adapted, depending on the dynamic evolution of the user situation. Thus, the system is pervasive in the sense that it performs the most suitable recommendation at the right moment in the right place. A prototype of the recommender is being implemented.
Mobile applications, that cover context-aware user modeling, are becoming increasingly prevalent. In this vein, information about the users are needed for the systems in order to provide them relevant services. This information enables the systems to figure out the users and their interests. For these reasons, different applications, in several areas, organize the user's properties, preferences and interests based on a structure, called a user model, which hold all relevant user-related information. In this respect, we propose a context-based user model in mobile Peer-to-Peer (P2P) environment. This model is based on aggregating different user information like past interests and the associated context. These past interests represent information about peers from which results were obtained and which were achieved from similar queries as well as from the user context. The basic idea of our proposal is to guess correlations between past requests, past peers from which results were obtained, associated user location and user interests. The generated correlations are based upon Formal Concept Analysis. We study, the exploitation of the proposed user model in results merging task in Peer-to-Peer Information Retrieval (P2PIR).
Peer-to-Peer Information Retrieval System (P2PIR) is a system that can query multiple information retrieval systems and merges ranked results list into a single result of documents. Classical methods are generally based on linear combination schemes. A major shortcoming of the classical methods is that there is no defined way to study dependencies and interactions existing among relevance criteria. In this paper, we propose a result merging method based on the Choquet Integral, called Choquet-Based Merging (CBM). The experimental results obtained on the test collection provided by TREC Contextual Suggestion track shows the effectiveness of our proposal.
This paper presents an extension of the AdaptWeb® web-learning environment in order to provide context adapted learning scenarios. This extension will enable intelligent behavior in building complex learning content on demand and customized for an intended audience depending on its current activity. For a specific scenario, ontologies may be employed to represent the knowledge the system has on the domain to be taught, on the situation where the study is being performed and on the learner profile
L'Internet des objets est une extension de l'Internet qui ouvre de nombreuses opportunites de construction de nouveaux services a haute valeur ajoutee dans de nombreux domaines. Cependant, les concepteurs de ces services doivent prendre en main des systemes logiciels de plus en plus complexes. Le projet INCOME etudie la brique logicielle de gestion de contexte qui traite et achemine les informations en provenance de l'Internet des objets entre producteurs et consommateurs d'informations fortement decouples. Il cible deux defis importants poses par l'Internet des objets : le traitement d'un flot continu d'informations et l'acces ouvert aux informations. INCOME propose des processus, des langages et des outils logiciels generiques pour la conception, la mise en œuvre, le deploiement et l'execution de gestionnaires de contexte capables non seulement de transformer et d'acheminer, mais en plus d'interpreter, de qualifier, de proteger et de filtrer les informations de contexte. Les propositions sont mises en œuvre dans un demonstrateur pour le guidage porte a porte des usagers de transports multimodaux dans la ville intelligente.
