Littérature | EPFL Graph Search

vignette|Fragonard, La Liseuse. La littérature est l’ensemble des œuvres écrites ou orales auxquelles on reconnaît une valeur esthétique ; c'est un art exprimant un idéal de beauté. Grâce aux productions littéraires, elle permet de manifester des émotions et de révéler aux lecteurs ou aux auditeurs ce qu'une personne a dans le cœur. La littérature vise à éduquer, à communiquer des pensées, à influencer et même à séduire. La littérature constitue un héritage patrimonial et peut concourir à la préservation du patrimoine d'un pays, lorsqu'elle en souligne les valeurs, la culture et la civilisation. Le mot littérature, issu du latin litteratura dérivé de littera (la lettre), apparaît souvent au début du avec un sens technique de « chose écrite » puis évolue à la fin du Moyen Âge vers le sens de « savoir tiré des livres ». Aux il prend son sens principal actuel, à savoir l'ensemble des œuvres écrites ou orales comportant une dimension esthétique (ex. : « C'est avec les beaux sentiments

Immobilization of single-walled carbon nanotubes by dielectrophoresis using a microfluidic channel

Paul Louis Mosser

The initial goal of this project was to determine by experimenting on a chip that where already design the ideal parameters for the dielectrophoresis of single walled carbon nanotubes. The epidemic blocking the access to the lab, the subject evolved to focus more on the literature study to compare the different parameters used in the literature for the dielectrophoresis and try to come up with a proposition of parameters to be tested on the chip by someone having access to the laboratory. The dielectrophoresis being the subject of a lot of studies as it is believed to be one of the best solution to build devices with carbon nanotubes, the literature on the subject is huge. It was impossible to take into account all the articles on the subject so this report is based only on the more relevant ones. The motivation behind this project is to find the best parameters possible for the dieletrophoresis in order to immobilized specific nanotubes for specific applications. The first specific application is the immobilization of nanotubes filled with metal-oxide precursors. This could enable the growth of an oxide where the nanotubes are immobilized enabling the fabrication of really small gas sensor based on metal-oxide. The second application presented is the immobilization of DNA functionalized nanotubes which can be used as bioreceptors enabling the detection of specific biomarkers. This could facilitate the fabrication of biosensors. This review show the most important parameters of the dielectrophoresis and how their modification impact the end results. By analyzing the results in the literature and based on the little experimentation made on the platform that was developed for this project, a set of parameters enabling the immobilization of one to a few single walled carbon nanotubes were determined.

2020

Rethinking Location Privacy for Unknown Mobility Behaviors

Carmela González Troncoso, Simón Oya Diez

Location Privacy-Preserving Mechanisms (LPPMs) in the literature largely consider that users' data available for training wholly characterizes their mobility patterns. Thus, they hardwire this information in their designs and evaluate their privacy properties with these same data. In this paper, we aim to understand the impact of this decision on the level of privacy these LPPMs may offer in real life when the users' mobility data may be different from the data used in the design phase. Our results show that, in many cases, training data does not capture users' behavior accurately and, thus, the level of privacy provided by the LPPM is often overestimated. To address this gap between theory and practice, we propose to use blank-slate models for LPPM design. Contrary to the hardwired approach, that assumes known users' behavior, blank-slate models learn the users' behavior from the queries to the service provider. We leverage this blank-slate approach to develop a new family of LPPMs, that we call Profile Estimation-Based LPPMs. Using real data, we empirically show that our proposal outperforms optimal state-of-the-art mechanisms designed on sporadic hardwired models. On non-sporadic location privacy scenarios, our method is only better if the usage of the location privacy service is not continuous. It is our hope that eliminating the need to bootstrap the mechanisms with training data and ensuring that the mechanisms are lightweight and easy to compute help fostering the integration of location privacy protections in deployed systems.

IEEE COMPUTER SOC2019

Compatibility modeling to support opportunity analysis in new product development

Julie Hohenegger

The general context of this Ph.D. dissertation is the new product development (NPD) which is the process for transforming an opportunity that is a business or technology gap in the current situation into a product available for sale. This Ph.D. dissertation focuses on the opportunity analysis in NPD. Therefore, a framework to support opportunity analyses is developed. The framework is structured in two parts: the information search and the modeling of compatibility. Opportunity analyses require a systematic approach to obtain all the information related to the identified opportunity, as well as an approach to determine whether or not this opportunity is compatible with respect to the different attributes/criteria representing the product development context. The state of the art review reveals that such approaches are not available in the published literature. Therefore, the objective of this research is to develop a comprehensive framework comprising an information search approach and a compatibility modeling approach to respond to these needs. The early phases of NPD are characterized by uncertainty that corresponds to the difference between the amount of information required to perform a task and the available amount of information. Uncertainty may have serious consequences on NPD, resulting in the worst case in a project failure. Therefore, information is crucial for the product development process. The information search approach developed in this Ph.D. dissertation allows the reduction of this uncertainty. This approach comprises four levels. The first level comprises the activities involved in NPD, such as market analysis and manufacturing. The second level proposes a classification of the different types of information required during the product development process: product-, process-, resource- and context-related. The third level represents the different sources where each type of information can be found. Two main classes are identified: the first one comprises the company departments such as marketing, and manufacturing, while the second class contains the sources from the external environment of the company, such as suppliers, customers, competitors, research communities, and governmental and non governmental organizations. The fourth level contains different methods of collecting information. Two main classes are distinguished: meetings- and documents-related. The links between the different levels are established on the basis of a thorough review of the related literature. Once the NPD activities associated with the identified opportunity are determined, the related information types, the sources where they can be found and the methods for their collections can be easily deduced from the information search approach. The compatibility modeling approach aims at evaluating the compatibility relationships for pairs of alternatives. Two types of compatibility can be distinguished: hard compatibility and soft compatibility. Hard compatibility corresponds to the notion of matching (i.e. the possibility that two alternatives coexist in a specific product development context) and is evaluated with respect to hard attributes/criteria related to the basic laws of nature or to the technical possibilities. Soft compatibility corresponds to the notion of non-degradation (i.e. the shift from a situation where the two alternatives are considered individually to a situation where they are considered together does not induce a degradation in a specific product development context) and is evaluated with respect to soft attributes/criteria such as development cost, and lead time. The core concept of the compatibility modeling approach is the compatibility structure. The crisp compatibility structure is a quadruplet of crisp relations M, C, I, and J, called mutual compatibility, strict incompatibility, mutual incompatibility and mutual incomparability respectively that satisfy a certain number of conditions. The relations M, C, C-1 and I correspond to four different situations: no degradation – no degradation, no degradation – degradation, degradation – no degradation, and degradation – degradation respectively. J corresponds to the situation where the alternatives cannot be compared using one of the relations M, C, C-1 or I. For each attribute/criterion, partial compatibility structures without incomparability are constructed from the impact that an alternative exerts on another and vice-versa. Then, relations characterizing the partial structures are aggregated into a global relation that takes into account all the attributes/criteria. By their nature, crisp compatibility structures cannot express degrees of mutual compatibility, strict incompatibility, mutual incompatibility or mutual incomparability, and are consequently too constraining in practice. Therefore, the crisp compatibility structures are meticulously extended to the case of fuzzy binary relations. A case study from the Swiss company Laurastar illustrates the framework that supports the opportunity analysis. This case study investigates the possibility to collaborate with new suppliers located in Bucharest and Hong-Kong. The compatibility between alternatives related to the supply chain activity is evaluated with respect to the various attributes/criteria.

EPFL2007

Compatibility modeling to support opportunity analysis in new product development

Julie Hohenegger

EPFL2007