Afin d'éviter au voyageur de consulter plusieurs sites web d'opérateurs de transport en commun pour planifier son déplacement, ce travail vise à concevoir un Système d'Information Coopératif de Mobilité (SICM) pour l'aide au déplacement multimodal. Il s'agit d'automatiser cette démarche de recherche et de composition d'itinéraires, pour fournir une information multimodale via un système intégrateur, en s'appuyant sur la théorie des systèmes multi agents (SMA) pour l'intégration et la médiation des systèmes d'information des différents opérateurs de transport.<br /><br />Pour produire l'information multimodale et multi opérateurs nécessaire à l'aide au déplacement, le SICM doit accéder aux différents systèmes d'information des opérateurs de transport et intégrer des résultats de recherche qui sont générés par les différents algorithmes des différents opérateurs. Dans cette approche, le SICM est un intergiciel (middleware) qui devient un client parmi d'autres usagers des systèmes d'information existants. Le SICM devient alors l'intermédiaire entre les différentes sources d'informations hétérogènes et distribuées d'une part et les clients d'autre part. Ce système doit être capable à la fois de trouver la bonne source d'information pour l'interroger selon les différentes requêtes des utilisateurs, et de regrouper les informations de manière cohérente pour répondre aux requêtes. Pour fournir un itinéraire composé mais surtout optimisé selon les critères de l'utilisateur, le recours à des algorithmes de plus courts chemins distribués « en ligne », et adaptés à des graphes dynamiques (dépendant du temps) a été retenu afin de réaliser ce moteur de recherche et de composition d'itinéraires multimodaux en ligne.

Studies show that instruction manuals are rarely read, and Volvo CE often finds the manual in the bookshelf at the office at customer visits. The instruction books are thus far from the users. How will information reach users? In this study I have worked with Volvo CE, Eskilstuna based on the question "How can information about how best driving economy be achieved and be designed to motivate drivers of wheel loaders to drive fuel efficiently?". Based on literature studies, interviews, analyzes and hearings, a prototype with accompanying concepts has been developed for Volvo CE for continued development. The prototype consists of a design where three factors that affect fuel efficient driving have been selected and constructed. The elements are text and image-based and then placed in an ECO OPERATOR program in the Volvo CE Co-pilot. Volvo CE already works with information via Co-pilot, which is a display located in the wheel loaders cab. Conclusions are that human centered design where the user is put in focus can be used to generate fuel efficient driving information adapted for wheel loaders.

<p>The availability of abundant multimodal data, including textual, visual, and sensor-based information, holds the potential to improve decision-making in diverse domains. Extracting data-driven decision-making information from heterogeneous and changing datasets in real-world data-centric applications requires achieving complementary functionalities of multimodal data integration, knowledge extraction and mining, situationally-aware data recommendation to different users, and uncertainty management in the open-world setting. To achieve a system that encompasses all of these functionalities, several challenges need to be effectively addressed: (1) How to represent and analyze heterogeneous source contents and application context for multimodal data recommendation? (2) How to predict and fulfill current and future needs as new information streams in without user intervention? (3) How to integrate disconnected data sources and learn relevant information to specific mission needs? (4) How to scale from processing petabytes of data to exabytes? (5) How to deal with uncertainties in open-world that stem from changes in data sources and user requirements?</p> <p><br></p> <p>This dissertation tackles these challenges by proposing novel frameworks, learning-based data integration and retrieval models, and algorithms to empower decision-makers to extract valuable insights from diverse multimodal data sources. The contributions of this dissertation can be summarized as follows: (1) We developed SKOD, a novel multimodal knowledge querying framework that overcomes the data representation, scalability, and data completeness issues while utilizing streaming brokers and RDBMS capabilities with entity-centric semantic features as an effective representation of content and context. Additionally, as part of the framework, a novel text attribute recognition model called HART was developed, which leveraged language models and syntactic properties of large unstructured texts. (2) In the SKOD framework, we incrementally proposed three different approaches for data integration of the disconnected sources from their semantic features to build a common knowledge base with the user information need: (i) EARS: A mediator approach using schema mapping of the semantic features and SQL joins was proposed to address scalability challenges in data integration; (ii) FemmIR: A data integration approach for more susceptible and flexible applications, that utilizes neural network-based graph matching techniques to learn coordinated graph representations of the data. It introduces a novel graph creation approach from the features and a novel similarity metric among data sources; (iii) WeSJem: This approach allows zero-shot similarity matching and data discovery by using contrastive learning<br> to embed data samples and query examples in a high-dimensional space using features as a novel source of supervision instead of relevance labels. (3) Finally, to manage uncertainties in multimodal data management for open-world environments, we characterized novelties in multimodal information retrieval based on data drift. Moreover, we proposed a novelty detection and adaptation technique as an augmentation to WeSJem.<br> </p> <p>The effectiveness of the proposed frameworks, models, and algorithms was demonstrated<br> through real-world system prototypes that solved open problems requiring large-scale human<br> endeavors and computational resources. Specifically, these prototypes assisted law enforcement officers in automating investigations and finding missing persons.<br> </p>

La politique des transports dans le monde et en Europe évolue vers une vision co-modale. Cette nouvelle politique n’oppose plus la voiture au transport public mais encourage une combinaison de tous les modes de transport en espérant ainsi assurer un développement rentable et durable.Nous focalisons notre étude sur le service transport de personnes qui s’inscrit au cœur des politiques co-modales en combinant tous les modes de transport en commun (métro, bus..) et promeut de nouveaux modes d’utilisation de la voiture particulière comme le covoiturage (partage d’un véhicule personnel) ou l’AutoPartage (voiture en libre-service).Toutefois, pour générer un itinéraire exploitant les services de plusieurs opérateurs de transport, il faut consulter plusieurs sites internet. Selon le déplacement à réaliser, cette tâche de planification complexe peut être très difficile à réaliser et ne garantit pas l’optimalité de l’itinéraire sélectionné.Nous nous sommes donc intéressés à la conception d’un système d’aide au déplacement capable de fournir une information voyageur (co-modale) en mettant en relation plusieurs opérateurs de transport (en commun et individuel). Le système en question doit être capable d’assister l’utilisateur dans la phase de planification par la constitution d’un carnet de voyage proposant plusieurs itinéraires multi-opérateurs. De plus, il assiste l’utilisateur en cas de perturbation en l’informant et en lui proposant des itinéraires de secours. Ce travail est basé sur des avancées technologiques qui facilitent l’optimisation dans un environnement distribué (Multi-agent - SOA) et rendent l’information accessible grâce à un grand nombre de médias (téléphone, PDA..) / Today's transport policy in the world, and more specifically in Europe, is moving towards a co-modal vision. This new policy does not oppose private to public transport but encourages a combination of all modes of transportation in the hope of assuring a lasting development.We focus our study on one special service: people transport which combines all modes of public transport (train, subway, tram, bus ..) and integrates new ways of car use such as carpooling (sharing a personal vehicle and the travel cost between drivers and passengers) and the EcoPartage (self-service cars in town).However, to generate a route using several transport operators, we have to check different Internet websites. This complex planning task can be very difficult and does not guarantee the optimality of the selected route.We are therefore interested in designing a support information system capable of providing comodal information linking several transport operators (public and individual).The system in question must be able to assist the user in the planning phase by offering a travelogue suggesting several possible routes; each route may include one or more transport operator (public or individual ones).In addition, it assists the user during the trip by informing him in case of disturbance and by proposing alternatives routes if necessary.This work is based on various technological developments in order to facilitate the information optimization in a distributed environment (Multi-agent - SOA) and make the information accessible to the user through a large number of media (telephone, mobile, PDA ...)