Search

In today’s highly connected cyber-physical environments, users are becoming more and more concerned about their privacy and ask for more involvement in the control of their data. However, achieving effective involvement of users requires improving their privacy decision-making. This can be achieved by: (i) raising their awareness regarding the direct and indirect privacy risks they accept to take when sharing data with consumers; (ii) helping them in optimizing their privacy protection decisions to meet their privacy requirements while maximizing data utility. In this article, we address the second goal by proposing a user-centric multi-objective approach for context-aware privacy management in connected environments, denoted δ- Risk . Our approach features a new privacy risk quantification model to dynamically calculate and select the best protection strategies for the user based on her preferences and contexts. Computed strategies are optimal in that they seek to closely satisfy user requirements and preferences while maximizing data utility and minimizing the cost of protection. We implemented our proposed approach and evaluated its performance and effectiveness in various scenarios. The results show that δ- Risk delivers scalability and low-complexity in time and space. Besides, it handles privacy reasoning in real-time, making it able to support the user in various contexts, including ephemeral ones. It also provides the user with at least one best strategy per context.

Connected environments (e.g., smart buildings and cities) have attracted various research efforts over the last few years (e.g., data modeling, sensor evolution, network and data management). In this work, we focus on querying such connected environments. Although Event Query Languages (EQL) have been used to define/detect events in these environments, we address here some of their limitations regarding: (i) proposing various query types for defining/handling connected environment components; (ii) considering environment, sensor network, and application domain components; (iii) considering various data and datatypes (e.g., scalar, multimedia); (iv) proposing query operators that deal with spatial/temporal distributions; and (v) coping with the dynamicity of the environments. To do so, we propose an EQL specifically designed for connected environments. We describe its framework, language, syntax, and queries. Finally, we illustrate its usage on a web platform for connected environment management (OpenCEMS).

International audience; Advances in privacy-enhancing technologies, such as context-aware and personalized privacy models, have paved the way for successful management of the data utility-privacy trade-off. However, significantly lowering the level of data protection when balancing utility-privacy to meet the individual’s needs makes subsequent protected data more precise. This increases the adversary’s ability to reveal the real values of the previous correlated data that needed more protection, making existing privacy models vulnerable to inference attacks. To overcome this problem, we propose in this paper a stochastic gradient descent solution for privacy-preserving during protection transitions, denoted P-SGD. The goal of this solution is to minimize the precision gap between sequential data when downshifting the protection by the privacy model. P-SGD intervenes at the protection descent phase and performs an iterative process that measures data dependencies, and gradually reduces protection accordingly until the desired protection level is reached. It considers also possible changes in protection functions and studies their impact on the protection descent rate. We validated our proposal and evaluated its performance. The results show that P-SGD is fast, scalable, and maintains low computational and storage complexity.

International audience; Recent technological advances have fueled the rise of connected environments (e.g., smart buildings and cities). Event Query Languages (EQL) have been used to define (and later detect) events in these environments. However, existing languages are limited to the definition of event patterns. They share the following limitations: (i) lack of consideration of the environment, sensor network, and application domain in their queries; (ii) lack of provided query types for the definition/handling of components/component instances; (iii) lack of considered data and datatypes (e.g., scalar, multimedia) needed for the definition of specific events; and (iv) difficulty in coping with the dynamicity of the environments. To address the aforementioned limitations, we propose here an EQL specifically designed for connected environments, denoted EQL-CE. We describe its framework, detail the used language, syntax, and queries. Finally, we illustrate the usage of EQL-CE in a smart mall example.

Larroque, B.; Arnould P.; Luthon F.; Poncet, P.A.; Rahali, A., and Abadie, S., 2018. In-situ measurements of wave impact pressure on a composite breakwater: preliminary results. In: Shim, J.-S.; Chun, I., and Lim, H.S. (eds.), Proceedings from the International Coastal Symposium (ICS) 2018 (Busan, Republic of Korea). Journal of Coastal Research, Special Issue No. 85, pp. 1086–1090. Coconut Creek (Florida), ISSN 0749-0208.A detached horizontally composite breakwater protecting the Saint Jean de Luz bay (French Basque Coast, Bay of Biscay, France) has been equipped with two pressure sensors displayed vertically onto the vertical wall. The two sensors recorded wave impact pressure in ten minute burst each hour at 10 kHz. In parallel, incoming waves were also recorded about 1 km off the structure with a directional wave buoy giving access to spectral wave parameters as well as raw data. Finally water level, wind magnitude and direction were also acquired. The whole dataset covers winter 2015–2016 from January to March.The results show first the statistical distribution of the measured parameters confirming preceding similar studies in which significantly lower impact pressure values were obtained compared to physical experiments or numerical simulations. A linear multivariate model has then been adjusted showing the overwhelming influence of wave heights to explain maximum pressure variability, followed by water level. Nevertheless this result has still to be confirmed for data reduced to impulsive impacts. Finally, interesting events supposedly corresponding to, or approaching flip-through impact type, have been identified and need further investigations.

International audience; With the rapid expansion of smart cyber–physical systems and environments, users become more and more concerned about their privacy, and ask for more involvement in the protection of their data. However, users may not be necessarily aware of the direct and indirect privacy risks they take to properly protect their privacy. In this paper, we propose a context-aware semantic reasoning system, denoted as the Privacy Oracle, capable of providing users with a dynamic overview of the privacy risks taken as their context evolves. To do so, the system continuously models, according to a proposed Semantic User Environment Modeling (SUEM) ontology, the knowledge (received by the system) about the user of interest and his surrounding cyber–physical environment. In parallel, it performs continuous reasoning over modeled information, by relying on set of privacy rules, in order to dynamically infer the privacy risks taken by the user. To validate our approach, we developed a prototype based on the semantic web tools such as OWL API, SWRL API and the inference engine Pellet. We evaluated the system performance by considering multiple use cases. Our experimental results show that the Privacy Oracle can assist users by dynamically detecting their incurred privacy risks, and by tracking, in real-time, the evolution of those risks as user context changes.

International audience; Semantic web techniques (e.g., ontologies) have been recently adopted for sensor network modeling. However, existing works do not fully address these challenges: (i) representing different sensor types (e.g., mobile/static sensors) to enrich the network with different data and ensure better coverage; (ii) representing a variety of platforms (e.g., environments, devices) for sensor deployment, thus, integrating new components (e.g., mobile phones); (iii) representing the diverse data (scalar/multimedia) needed for various applications (e.g., event detection); and (iv) proposing a generic model to allow re-usability in various application domains. In this paper, we propose HSSN, an ontology that extends the Semantic Sensor Network (SSN) ontology which is already re-usable and considers various platforms. We extend the representation of sensors, sensed data, and deployment environments to cope with these challenges. We evaluate the consistency, accuracy, clarity, and performance of HSSN.

International audience; Recent advances in sensor technology and information processing have allowed connected environments to impact various application domains. In order to detect events in these environments, existing works rely on the sensed data. However, these works are not re-usable since they statically define the targeted events (i.e., the definitions are hard to modify when needed). Here, we present a generic framework for event detection composed of (i) a representation of the environment; (ii) an event detection mechanism; and (iii) an Event Query Language (EQL) for user/framework interaction. This paper focuses on detailing the EQL which allows the definition of the data model components, handles instances of each component, protects the security/privacy of data/users, and defines/detects events. We also propose a query optimizer in order to handle the dynamicity of the environment and spatial/temporal constraints. We finally illustrate the EQL and conclude the paper with some future works.

Morichon, D.; de Santiago, I.; Delpey, M.; Somdecoste, T.; Callens, A.; Liquet, B.; Liria, P., and Arnould, P., 2018. Assessment of flooding hazards at an engineered beach during extreme events: Biarritz, SW France. In: Shim, J.-S.; Chun, I., and Lim, H.S. (eds.), Proceedings from the International Coastal Symposium (ICS) 2018 (Busan, Republic of Korea). Journal of Coastal Research, Special Issue No. 85, pp. 801–805. Coconut Creek (Florida), ISSN 0749-0208.This study focuses on flooding hazards along the Grande Plage of Biarritz. This pocket engineered beach, located in the south west of France, was highly battered by a series of storms during the winter 2013–2014 and experienced great damages to buildings and infrastructures. The vulnerability of this type of beaches to storm impact is assessed using the XBeach surf beat model. For validation purpose, a dedicated spectral wave model is also used on the coastal area to force XBeach at 15 m depth. This coastal wave model is first validated against nearshore wave measurements provided by two sensors deployed during the winter 2017. Then the total water level elevations measured at the shore with a video monitoring system are used to assess XBeach results. The model is then used to run a series of extreme storm scenarios with different return periods (1, 10 and 50 years). Results show how the occurrence of flooding events increases with storm intensity. The response of the beach is alongshore non-uniform for all the storm scenarios. The section, where the dry beach width is smaller, is more exposed to flooding whereas vulnerability is reduced by wider dry beach areas. This study constitutes a milestone and outlines next steps toward the development of an operational warning system for the GPB.

International audience; In the context of social media, existent works offer social-event-based organization of multimedia objects (e.g., photos, videos) by mainly considering spatio-temporal data, while neglecting other user-related information (e.g., people, user interests). In this paper we propose an automated, extensible, and incremental Feature-centric Social Event Detection (F-SED) approach, based on Formal Concept Analysis (FCA), to organize shared multimedia objects on social media platforms and sharing applications. F-SED simultaneously considers various event features (e.g., temporal, geographical, social (user related)), and uses the latter to detect different feature-centric events (e.g., user-centric, location-centric). Our experimental results show that detection accuracy is improved when, besides spatio-temporal information, other features, such as social, are considered. We also show that the performance of our prototype is quasi-linear in most cases.

This paper presents the dead reckoning localisation technique of the robot ROMANE (“Robot Mobile Autonome de Navigation en Exterieur”) which is conceived to move outdoors on level roads without traffic. Vehicle localisation with incremental shaft encoders on the wheels is improved by integration of a magnetic compass into the robot system. Fusion of odometric and magnetometric data by Kalman filtering increases the accuracy of the estimated robot heading. Consequently the uncertainty of the vehicle position is reduced. To evaluate our localisation method we present some experimental results of trajectory estimations.

This article deals with a method of automatic adaptation of the scrolling time of the different actions proposed in a man/machine interface called "communicator". The EDITH project (for Digital Remote-action Environment for Disabled persons), elaborated at the University of Metz, was used as a test-bed to define this method. It is only after the clinical evaluation of the EDITH system with a patient suffering amyotrophic lateral sclerosis that we realized an optimization of the temporal management on the control sensor that enables to increase the performance of this type of interface.

Recent years have witnessed rapid progress in enabling technologies for data sensing, communication and mining, paving the way for the phenomenal growth of smart connected environments (e.g., smart buildings, cities, factories). These environments are currently providing interesting and useful applications that help users in their everyday tasks (e.g. increasing comfort, reducing energy consumption). However, such applications require to collect, exchange, store, and process large amount of fine-granular data that is often privacy-sensitive for their users (e.g., location, energy-consumption), as its analysis allows data consumers to reveal sensitive information about them, such as their health conditions and preferences.Consequently, involving users in the management of their privacy is nowadays receiving extensive attention. Nonetheless, various improvements are still required. For instance, how to raise user awareness of the privacy risks involved in their data sharing and/or imposed by their environments. Moreover, how to enable users to assess their situations and make optimal data utility-privacy decisions accordingly.In this thesis, we focus on six main challenges: (1) representing diverse user contexts with a high semantic expressiveness power; (2) performing a holistic (all-data-inclusive) context-based privacy risk reasoning; (3) achieving user-centric privacy management; (4) making optimal context-based privacy decisions; (5) coping with the inter-context data dependency; and (6) delivering scalability and efficiency in order to assist the user in a variety of situations.To address these challenges, we first present an ontology-based data model capable of representing various user contexts with high-level information coverage. Following that, we introduce a context-aware semantic reasoning approach for privacy risk inference that provides a dynamic/contextual overview of risks tailored to the user's expertise. Then, to enable optimal management of data utility-privacy trade-offs, we propose a user-centric multi-objective approach for context-aware privacy management that provides dynamic best data protection strategies to be implemented based on user situations and preferences. Finally, we propose a new stochastic gradient descent solution for privacy-preserving during protection transitions, which gives an additional layer of protection against data inference attacks.The aforementioned contributions are regrouped in one global generic and extensible framework for context-aware privacy management.; Ces dernières années, les technologies de détection, communication et gestion de données ont connu de progrès rapides, ouvrant la voie à la croissance phénoménale des environnements connectés intelligents (bâtiments, villes intelligentes). Ces environnements fournissent actuellement des applications intéressantes et utiles qui aident les utilisateurs dans leurs tâches quotidiennes (augmenter le confort, réduire la consommation d'énergie). Cependant, de telles applications nécessitent de collecter, échanger, stocker et traiter une grande quantité de données à granularité fine qui sont souvent sensibles pour leurs utilisateurs (localisation, consommation d'énergie), surtout que leurs analyses permettent aux consommateurs de données de révéler des informations sensibles (état de santé, préférences des utilisateurs).Par conséquent, l'implication des utilisateurs dans la gestion de leur vie privée reçoit aujourd’hui énormément d'attention. Néanmoins, diverses améliorations sont encore nécessaires. Par exemple, comment sensibiliser les utilisateurs aux risques liés à leur vie privée impliqués dans le partage de leurs données et/ou imposés par leurs environnements. De plus, comment permettre aux utilisateurs d'évaluer leurs situations et de prendre des décisions optimales en matière d'utilité et de confidentialité de données en conséquence.Dans cette thèse, nous nous concentrons sur six défis principaux: (1) représenter des contextes diversifiés de l'utilisateur avec une haute puissance d'expressivité sémantique ; (2) effectuer un raisonnement holistique (toutes données incluses) et contextuel sur les risques en matières de vie privée ; (3) parvenir à une gestion de vie privée centrée sur l'utilisateur ; (4) prendre des décisions contextuelles optimales liées à la protection de la vie privée ; (5) gérer la dépendance des données inter-contextuelles ; et (6) fournir une solution évolutive et efficace afin d'assister l'utilisateur dans diverses situations.Pour ce faire, nous présentons d'abord un modèle de données basé sur une ontologie capable de représenter divers contextes utilisateur avec une couverture d'informations de haut niveau. Ensuite, nous introduisons une approche de raisonnement sémantique qui fournit un aperçu dynamique/contextuel des risques en matière de vie privée, adapté à l'expertise de l'utilisateur. Ensuite, pour permettre une gestion contextuelle optimale des compromis entre l'utilité et la confidentialité des données, nous proposons une approche multi-objectifs, centrée sur l'utilisateur, qui fournit d’une manière dynamique les meilleures stratégies de protection de données à mettre en œuvre en fonction des situations et préférences des utilisateurs. Enfin, nous proposons une nouvelle solution de descente de gradient stochastique pour assurer la protection de la vie privée des utilisateurs lors des transitions des niveaux de protection associés à leurs données. Cette solution offre ainsi une couche supplémentaire de protection contre les attaques par inférence de données.Les contributions susmentionnées sont regroupées dans un framework global générique et extensible pour la gestion contextuelle de la protection de la vie privée.

The rising interest in smart connected environments (e.g., smart buildings, cities, factories) and theevolution of sensors, data management/communication technologies have paved the way forinteresting and useful applications that help users in their every day tasks (e.g. increasing comfort,reducing energy consumption). However, various improvements are still required. For instance, howto enhance the representation of such complex, dynamic, and heterogeneous environments.Moreover, how to facilitate the interaction between users and their connected environments, and howto provide tools for environment monitoring and management.In this thesis, we focus on four main challenges: (i) representing a diverse set of components andelements related to the environment and its sensor network; (ii) providing a query language thathandles user/connected environment interactions (e.g., environment definition, data management,event definition); (iii) coping with the dynamicity of the environment and its evolution over time; and(iv) proposing a generic event detection mechanism for improved environment monitoring.To do so, we first present an ontology-based data model that represents hybrid environments/sensornetworks. Thus covering diverse sensors (e.g., static, mobile), environments (e.g., infrastructures,devices), and data (e.g., scalar, multimedia). Then, we introduce a query language that one might usefor various tasks (e.g., defining the connected environment, information retrieval, event definition,data management). Furthermore, to keep up with the environment changes we provide a queryoptimizer that allows the submitted queries to cope with the dynamicity of the environment prior totheir execution. Finally, we propose an event detection core that takes event definitions as input anddetects the targeted events.We group the aforementioned modules in one global framework for event detection in connectedenvironments. Our proposal is generic, extensible, and could be used with different connectedenvironments such as buildings, cities. . .; L’intérêt croissant pour les environnements connectés (bâtiments, villes, usines intelligents) etl’évolution des réseaux de capteurs, technologies de gestion/communication de données ont ouvertla voie à des applications intéressantes et utiles qui aident les utilisateurs dans leurs tâchesquotidiennes (augmenter la productivité dans une usine, réduire la consommation d’énergie).Cependant, diverses améliorations sont encore nécessaires. Par exemple, comment améliorer lareprésentation de ces environnements complexes, dynamiques et hétérogènes. En outre, commentfaciliter l’interaction entre les utilisateurs et leurs environnements connectés et comment fournir desoutils de surveillance et de gestion de tels environnements.Dans cette thèse, nous nous concentrons sur quatre défis principaux: (i) représenter un ensemblediversifié de composants et d’éléments liés à l’environnement et à son réseau de capteurs; (ii) fournirun langage de requête qui gère les interactions utilisateur/environnement connecté (pour la définitionde l’environnement, la gestion de données, la définition d’événements); (iii) faire face à la dynamiquede l’environnement et à son évolution dans le temps; et (iv) proposer un mécanisme générique dedétection d’événements pour mieux surveiller l’environnement.Pour ce faire, nous présentons d’abord un modèle de données basé sur une ontologie qui représentedes environnements et réseaux de capteurs hybrides. Couvrant ainsi divers capteurs (statique, mobile),environnements (infrastructures, équipements) et données (scalaires, multimédia). Ensuite, nousintroduisons un langage de requête que l’on pourrait utiliser pour diverses tâches (définirl’environnement connecté, la recherche d’informations, la définition d’événements, la gestion dedonnées). De plus, afin de suivre les changements d’environnement, nous fournissons un optimiseurde requêtes qui permet aux requêtes soumises de gérer la dynamique de l’environnement avant leurexécution. Enfin, nous proposons un noyau de détection d’événement qui prend en entrée lesdéfinitions d’événement et détecte les événements ciblés.Nous regroupons les modules susmentionnés dans un framework global pour la détectiond’événements dans des environnements connectés. Notre proposition est générique, extensible, etpourrait être utilisée avec différents environnements connectés tels que des bâtiments, des villes. . .