Gestion des connaissances | EPFL Graph Search

La gestion des connaissances (en anglais knowledge management) est une démarche managériale pluridisciplinaire qui regroupe l'ensemble des initiatives, des méthodes et des techniques permettant de percevoir, identifier, analyser, organiser, mémoriser, partager les connaissances des membres d'une organisation – les savoirs créés par l'entreprise elle-même (marketing, recherche et développement) ou acquis de l'extérieur (intelligence économique) – en vue d'atteindre un objectif fixé. Gestion des connaissances appliquée aux groupes Nous sommes submergés d'informations. Les entreprises, les scientifiques ou même les particuliers sont maintenant obligés d'appliquer une stratégie dans le traitement et la transmission de l'information dans les activités de tous les jours : voter, travailler, chercher un emploi, gagner des marchés, etc. . D'après des praticiens et des académiciens tels que R. C. Tisseyre, Larry Prusak, J.Y. Prax , C. Desprès et D. Chauvel , il est nécessaire en

Building Capabilities for Service Innovation

Andreas Felix von Vangerow

The service sector is the most dominant sector in every developed economy today. The ability to develop and successfully launch new services is crucial for service organizations to stay competitive and to adapt to the ever-changing conditions of the environment. However, up until now, a mismatch remains between the importance of service innovation for society and the attention given to it in both academia and practice. Existing research has compared the innovative output of service firms with manufacturing firms and found that the difference is one of degree rather than of substance. Studies which focus on the process instead of the output of service innovation, however, are still relatively scarce. This dissertation aims to help close this research gap by examining three service innovation processes in different contexts. First, it offers a theoretical illumination of service innovation with the help of knowledge creation theory. This is beneficial for defining service innovation more precisely. The second study is concerned with the influence of emotional factors on the development of dynamic capabilities in new ventures. It shows that young ventures which are able to deliberately switch between emotional and analytical modes of operation have a higher likelihood of success. The study compares the development of two service-oriented ventures with two product-oriented ones that are part of an accelerator program at the School of Design Thinking at Stanford University. The third paper is concerned with product-service strategies, or Servitization, and the relation to vertical integration and the coordination of innovation in downstream markets. The study shows that the Servitization development, i.e. the vertical integration of manufacturing firms into its users' operation, is a response to the failure of architectural innovation in more flexible buyer-supplier relationships. The study finds four different strategies for the service business of manufacturing firms based on case studies in the mechanical engineering, electrical engineering, and chemical industries. Overall, I am able to show that firms need to be able to build three distinct capabilities to master their service innovation efforts. First, they need to build up a capacity for innovation to gain economies from partial repetition of solutions. Second, firms need to have skills in place to create and manage boundary objects, especially at the client interface. Finally, they need to have capabilities to build up complementarities which help protect revenues from their service innovation initiatives.

EPFL2015

End-to-End Task-Oriented Dialog Modeling With Semi-Structured Knowledge Management

Silin Gao

Current task-oriented dialog (TOD) systems mostly manage structured knowledge (e.g. databases and tables) to guide the goal-oriented conversations. However, they fall short of handling dialogs which also involve unstructured knowledge (e.g. reviews and documents). In this article, we formulate a task of modeling TOD grounded on a fusion of structured and unstructured knowledge. To address this task, we propose a TOD system with semi-structured knowledge management, SeKnow, which extends the belief state to manage knowledge with both structured and unstructured contents. Furthermore, we introduce two implementations of SeKnow based on a non-pretrained sequence-to-sequence model and a pretrained language model, respectively. Both implementations use the end-to-end manner to jointly optimize dialog modeling grounded on structured and unstructured knowledge. We conduct experiments on a modified version of MultiWOZ 2.1 dataset, Mod-MultiWOZ 2.1, where dialogs are processed to involve semi-structured knowledge. Experimental results show that SeKnow has strong performances in both end-to-end dialog and intermediate knowledge management, compared to existing TOD systems and their extensions with pipeline knowledge management schemes.

IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC2022

Knowledge-based decision support for improving the adaptive product development process

Predrag Spasojevic

The product development process (PDP) is a complex process encompassing many very diverse activities, and involving a fairly big number of actors, spread across different professions, teams and companies. In todayâs product development, more often than not these individual actors and groups of actors perform their duties almost isolated from the others, using their own methods and software tools to achieve their respective local goals within the overall development effort. Even though much of the same knowledge about the product is used in different phases and tasks of the PDP, in most cases it is not shared or structured in any formal manner, except locally in some cases. Instead, it is exchanged through meetings of the responsible people or other informal means. Current PDP support ICT solutions are focusing on specific parts of the process, but there is no integrated approach that includes product data and services required for the whole PDP, despite the knowledge commonalities shared by its parts and phases. Apart from that, information about the process itself (such as how the tasks are defined, planned and organized, which changes are made to a plan before and during the execution of the tasks, and based on which conditions are these decisions being made), is rarely kept or formally described, relying on the people and their memory. This lack of awareness about the process plans and execution among people responsible for different tasks within PDP leads to misunderstandings, collisions and delays arising much more often than necessary, and work being done in a suboptimal fashion. Having both the product and process knowledge captured, managed, shared and reused in a standardized manner would aid the process tremendously, by improving the collaboration between all the parties involved, making sure that the entire process and its individual tasks are executed at the right moment, with the correct resources, having minimal negative effect on the others, and with the best possible performance. This thesis presents an approach to reduce the impact of these issues, based on a PDP knowledge management and decision support software system. Ontologies, description logic and Semantic Web technologies are used to formalize the knowledge, enable reasoning and additional fact inferencing, as well as to make it easier to access by both machine and human actors in the PDP. As a first step, a general model of the PDP was created, based on detailed descriptions of product development processes in six companies from three different industry branches (aerospace, automotive and home appliances). In accordance with the model, a broad PDP ontology was designed and implemented, as part of a collaborative effort with other research institutes and companies. The ontology provides the formal structure for the knowledge encountered during the PDP by describing the relevant classes, properties and relations. In addition, the architecture of the ontology-centered knowledge base and the adjoining support software framework is described, showing relevant implementation details and features for supporting PDP planning and execution. They give insight into the benefits of applying the Semantic Web technologies in the product development domain, and how they provide the basis for improving the PDP. The benefits are reflected in shorter PDP project times and better results, which come from knowledge capture and reuse, improved collaboration, faster reaction and [...]

EPFL2015

Building Capabilities for Service Innovation

Andreas Felix von Vangerow

EPFL2015

Knowledge-based decision support for improving the adaptive product development process

Predrag Spasojevic

EPFL2015