Дисертації з теми "Confidentialité des clients"

La thèse se compose de trois essais sur "Modèles d'intelligence artificielle pour la gestion de la relation client". En particulier, les essais sont construits sur de nouveaux algorithmes d'apprentissage en profondeur connus sous le nom de réseaux de neurones à mémoire longue et courte durée (LSTM). Le premier essai montre que les réseaux de neurones LSTM, qui reposent exclusivement sur des données brutes en entrée, peuvent remplacer les modèles basés sur l'ingénierie des fonctionnalités pour prédire le comportement des clients dans le contexte du marketing direct. Dans le deuxième essai, je développe un modèle de réseaux de neurones à double LSTM pour les prédictions de la valeur à vie du client (CLV) qui intègre la politique adaptative de l'entreprise et capture l'attrition latente. Je montre que l'intégration de la politique marketing adaptative de l'entreprise dans les modèles d'évaluation des clients conduit à des trajectoires divergentes de rentabilité future, mettant en évidence des bimodalités et des modes à la queue de la distribution CLV. La présence de bimodalités et de modes à la queue capturée par mon modèle a des implications pour une compréhension plus approfondie du potentiel du client (c'est-à-dire le meilleur scénario par rapport à l'attente CLV traditionnelle) et pour affiner la hiérarchisation des clients. Dans le troisième essai, j'exploite la capacité des états de mémoire des réseaux de neurones LSTM à résumer les informations utiles des données de transaction passées pour développer une nouvelle approche qui répond aux préoccupations de confidentialité des clients dans la gestion de la relation client (CRM)
The dissertation consists of three essays on “Artificial Intelligence Models for Customer Relationship Management”. In particular, the essays are built on novel deep learning algorithms known as long-short term memory (LSTM) neural networks. The first essay shows that LSTM neural networks, which rely exclusively on raw data as input, can replace feature-engineering based models for predicting customer behavior in the direct marketing context. In the second essay, I develop a dual-LSTM neural networks model for customer lifetime value (CLV) predictions that incorporates the firm’s adaptive policy as well as captures latent attrition. I show that integrating the firm’s adaptive marketing policy into customer valuation models leads to diverging paths of future profitability, highlighting bimodalities and modes at the tail of the CLV distribution. The presence of bimodalities and modes at the tail captured by my model has implications for a deeper understanding of customer potential (i.e., best-case scenario vs. traditional CLV expectation) and for fine-tuning customer prioritization. In the third essay, I leverage the ability of LSTM neural networks’ memory states to summarize useful information from the past transaction data to develop a novel approach that addresses customer privacy concerns in customer relationship management (CRM)

Internet a profondément changé la manière dont les entreprises interagissent avec leurs clients. Le développement d’une relation client en ligne a mené les organisations à repenser l’idée qu’elles se faisaient de cette relation, ainsi que leurs pratiques en termes de marketing relationnel. Cette recherche vise à mieux comprendre la manière dont les utilisateurs d’espaces clients de sites web évaluent la qualité de la relation client en ligne, et inclut une dimension importante dans l’évaluation de la qualité de la relation en ligne, l’appropriation de l’espace client.Une étude qualitative a été menée auprès de 15 personnes, afin d’étudier les variables permettant de construire le modèle de recherche et de proposer une échelle de mesure de l’appropriation des espaces clients de sites web. Par la suite, un modèle de recherche a été créé et un questionnaire en ligne a été administré. Un échantillon final se composant de 534 personnes, utilisatrices d’espaces clients de sites web dans le secteur bancaire et dans le secteur de la téléphonie mobile, a permis de tester les hypothèses de recherche. Il en résulte que l’appropriation de l’espace client du site web joue un rôle médiateur entre deux variables - la facilité d’utilisation perçue et l’utilité perçue - et la qualité de la relation client en ligne, que ce soit pour les utilisateurs d’espaces clients de sites web dans le secteur bancaire ou dans le secteur de la téléphonie mobile. De même, la sécurité/confidentialité perçue de l’espace client du site web exerce une influence positive sur la qualité de la relation client en ligne. L’hypothèse postulant que l’interactivité perçue de l’espace client du site web exerce une influence positive sur la qualité de la relation client en ligne est partiellement validée. En revanche, la présence sociale de l’espace client du site web n’exerce pas une influence positive sur la qualité de la relation client en ligne
The Internet has dramatically changed the way companies interact with their customers. Because of the importance of e-customer relationship management, companies have to reconsider their strategies in terms of relationship marketing. The purpose of this research is to investigate the way users of personal websites’ accounts evaluate e-relationship quality within two sectors: the banking sector and the mobile phone sector. This research deals with an important concept: appropriation.A qualitative study has been conducted in order to build a research model and to create a measurement scale to study the appropriation of a personal websites’ account. Therefore, an online survey (N=534) was conducted to test the hypothesis. The findings suggest that the appropriation of a personal websites’ account is a mediating variable between perceived ease of use, perceived usability and relationship quality, both in the banking sector and in the mobile phone sector. In the same way, privacy has a positive influence on e-relationship quality. Moreover, perceived interactivity has a positive influence on e-relationship quality, but that particular hypothesis is partially validated. Finally, social presence does not have a positive influence on e-relationship quality

The present study investigated how the issues of perceived and desired confidentiality are related to the working alliance between adolescent clients and their counselors. Fifty-one students between the ages of 14 and 18 years were recruited through two school boards in Canadian cities. Results indicated that adolescents preferred greater levels of confidentiality than they thought they would actually get in hypothetical situations, but preferred significantly less in actual situations. In addition, the level of confidentiality adolescents preferred in both hypothetical and actual situations did not impact the working alliance. The level of confidentiality adolescents thought they would get in hypothetical situations was a significant predictor of the working alliance. In actual situations, however, the level of confidentiality did not impact the working alliance. Theoretical and practical implications for counselors and other researchers, limitations of this study, and future research directions are discussed.

Cloud storage services allow users to store and access data in a secure and flexible manner. In recent years, cloud storage services have seen rapid growth in popularity as well as in technological progress and hundreds of millions of users use these services to store thousands of petabytes of data. Additionally, the synchronization of data that is essential for these types of services stands for a significant amount of the total internet traffic. In this thesis, seven cloud storage applications were tested under controlled experiments during the synchronization process to determine feature support and measure performance metrics. Special focus was put on comparing applications that perform client side encryption of user data to applicationsthat do not. The results show a great variation in feature support and performance between the different applications and that client side encryption introduces some limitations to other features but that it does not necessarily impact performance negatively. The results provide insights and enhances the understanding of the advantages and disadvantages that come with certain design choices of cloud storage applications. These insights will help future technological development of cloud storage services.

La fraude au fournisseur (Supplier Impersonation Fraud, SIF) est un type de fraude se produisant dans un contexte Business-to-Business (B2B), où des entreprises et des commerces interagissent entre eux, plutôt qu'avec le consommateur. Une fraude au fournisseur est effectuée lorsqu'une entreprise (fournisseur) proposant des biens ou des services à une autre entreprise (client) a son identité usurpée par un fraudeur. Dans cette thèse, nous proposons, d'utiliser les techniques et outils récents en matière d'apprentissage machine (Machine Learning) afin de résoudre à ces différents points, en élaborant des systèmes de détection de fraudes se basant sur l'analyse de données. Deux systèmes de détection de fraude basés sur l'analyse de données sont proposés: ProbaSIF et GraphSIF. Ces deux systèmes se composent d'abord d'une phase d'entraînement où les transactions historiques sont utilisées pour calculer un modèle de données, puis d'une phase de test où la légitimité de chaque transaction considérée est déterminée. ProbaSIF est un système de détection de fraudes au fournisseur qui se base sur un modèle bayésien (Dirichlet-Multinomial). ProbaSIF utilise la probabilité d'un compte en banque à être utilisé dans une transaction future d'une entreprise pour déterminer sa fiabilité. GraphSIF, le second système de détection de fraude au fournisseur que nous proposons, a pour but d'analyser les propriétés relationnelles créées par l'échange de transactions entre une entreprise et ses fournisseurs. À cette fin, une séquence de différents graphes compilant tous les liens créés entre l'entreprise, ses fournisseurs, et les comptes en banque utilisés pour payer ces fournisseurs, appelés séquence de comportement, est générée. Une transaction est catégorisée en l'ajoutant au graphe le plus récent de la séquence et en analysant les motifs formés, et en les comparant à ceux précédemment trouvés dans la séquence de comportement.Ces deux systèmes sont comparés avec un jeu de données réelles afin d’examiner leurs performances
Supplier Impersonation Fraud (SIF) is a kind of fraud occuring in a Business-To-Business context (B2B), where a fraudster impersonates a supplier in order to trigger an illegitimate payment from a company. Most of the exisiting systems focus solely on a single, "intra-company" approach in order to detect such kind of fraud. However, the companies are part of an ecosystem where multiple agents interacts, and such interaction hav yet to be integrated as a part of the existing detection techniques. In this thesis we propose to use state-of-the-art techniques in Machine Learning in order to build a detection system for such frauds, based on the elaboration of a model using historical transactions from both the targeted companies and the relevant other companies in the ecosystem (contextual data). We perform detection of anomalous transactions when significant change in the payment behavior of a company is detected. Two ML-based systems are proposed in this work: ProbaSIF and GraphSIF. ProbaSIF uses a probabilistic approach (urn model) in order to asert the probability of occurrence of the account used in the transaction in order to assert its legitimacy. We use this approach to assert the differences yielded by the integration of contextual data to the analysis. GraphSIF uses a graph-based approach to model the interaction between client and supplier companies as graphs, and then uses these graph as training data in a Self-Organizing Map-Clustering model. The distance between a new transaction and the center of the cluster is used to detect changes in the behavior of a client company. These two systems are compared with a real-life fraud detection system in order to assert their performance

Thesis (Ph. D.)--West Virginia University, 2007.
Title from document title page. Document formatted into pages; contains vi, 149 p. Includes abstract. Includes bibliographical references (p. 117-125).

Magister Legum - LLM
“Threats to the independence of the legal profession have become a preoccupation for bar leaders, regulators and academics, driven by the dual pressures of globalization and the changing business structure of the profession.”1 Money laundering is a transnational economic crime that has plagued the world economy for many decades. It is a crime that eluded the attention of most world leaders. Hence, it is this elusiveness and this non-interest in money-laundering as a serious economic crime, that afforded many individuals such as former dictators and military leaders in developing countries the chance to avoid prosecution for depleting the economic resources of their particular state.2

Today’s massive usage of smartphones has put a high demand on all application developers in the matter of security. For us to be able to keep using all existing and new applications, a process that removes significant security vulnerabilities is essential. To remove these vulnerabilities, the applications have to be tested. In this thesis, we identify six methods for functional and security testing of client-server applications running Android and Python Flask. Regarding functional testing, we implement Espresso testing and RESTful API testing. In regards to the security testing of the system, we do not only implement fuzz testing, sniffing, reverse engineering and SQL injection testing on a system developed by a student group in a parallel project, but also discover a significant security vulnerability that directly affects the integrity and reliability of this system. Out of the six identified testing techniques, reverse engineering exposed the vulnerability. In conjunction with this, we verified that the system’s functionality works as it is supposed to.

During the last part of the 20th century many companies has started to build their organizations on knowledge and information instead of manufacturing of products.  Today it is easier to communicate and spread information. At the same time the numbers of untraditional employments like temporary employments and workers from staffing agencies are increasing. Many organizations have higher employee turnover now than before. This allows more people to have access to confidential information, whose spreading might cause great damage for the employer. This essay aims to look into the protection of confidential information for the client against offenses from staffing agency workers that are working within the clients’ organization. Protection of confidential information is included in the duty of loyalty in terms of confidentiality, protection of trade secrets and competition clauses. Duty of loyalty lies within the employment contract between the agency employer and the staffing agency. Therefore the agency employer has no duty to be loyal against the client. For the agency worker to be able to perform the required work duties he or she might be provided with confidential information. Spreading of this information could cause damage for the client. Can the rules of justice that regulates duty of loyalty be applied on the relationship between the client and the agency worker?  What are the consequences of the rules that exist today? This essay is written with a juridical method and with sociology of law perspective. Other parts of the duty of loyalty is not included as well as the duty of loyalty for the employer. The conclusion is that the staffing agency is responsible for indemnity towards the client if an agency employer reviles confidential information about the clients’ organization.  Some rules of justice can be interpreted as an expression that duty of loyalty lays between the client and the agency worker. For example, the agency worker is bound by confidentiality in terms of the conditions at the clients’ organization.  Suggestions of changes in the regulations about protection of confidential information will force the agency worker to pay indemnity to the client if he or she reviles confidential information.   In staffing agencies there are twice as many immigrants than in the regular labor market. The rules of duty of loyalty can contribute to a holdback for immigrants that are working in staffing agencies. Immigrants will be at bigger risk of standing outside the general labor market than people born in Sweden because of static discrimination.
Företag har under senare delen av 1900-talet gått från traditionell varuproduktion till att allt mer bygga sin verksamhet på information och kunskap. Den tekniska utvecklingen har gjort det lättare att överföra och sprida kunskap i samhället vilket ökar sårbarheten hos företagen. Icke- traditionella anställningar såsom tidsbegränsade anställningar och inhyrd arbetskraft ökar och företag har en högre personalomsättning nu än tidigare. Fler personer får tillgång till företagsspecifik information och spridning av den skulle kunna skada arbetsgivaren. I min uppsats undersöker jag skyddet av hemlig information för arbetsgivare som hyr in arbetskraft från ett personaluthyrningsföretag. Skydd för hemlig information ingår i den så kallade lojalitetsplikten i form av tystnadsplikt, skydd för företagshemlighet och konkurrensklausuler. Lojalitetplikt föreligger i anställningsavtalet mellan uthyrd arbetstagare och uthyrningsföretaget men arbetstagarens arbetsprestation utförs hos kunden. Kan regleringar kring lojalitetsplikt tillämpas på förhållandet mellan kund och inhyrd arbetstagare? Vad blir konsekvenserna av de regler som finns idag? Uppsatsen är skriven med juridisk metod som har sin utgångspunkt i rättskälleläran samt med rättssociologisk angreppssätt. Andra delar av lojalitetsplikt såsom upplysningsskyldighet, bisysslor eller kritikrätt behandlas inte i uppsatsen, ej heller lojalitetsplikt för arbetsgivaren. Slutsatsen är att uthyrningsföretaget står för den uthyrde arbetstagarens handlingar mot kundföretaget eftersom det föreligger lojalitetsförpliktelser i uthyrningsavtalet mellan kund och uthyrare. Uthyraren står för exempelvis skadestånd om en arbetstagare röjer eller utnyttjar en företagshemlighet hos kunden. Vissa regleringar ger uttryck för att det föreligger lojalitetsplikt mellan kund och inhyrd arbetstagare. Exempelvis är den uthyrda arbetstagaren bunden av tystnadsplikt vad det gäller förhållanden hos kunden. I ett förslag från regeringen på kommande ändringar i lagstiftningen kommer den inhyrde arbetstagaren bli direkt skadeståndsansvarig vid röjande eller utnyttjande av företagshemlighet hos kunden. I uthyrningsbranschen finns dubbelt så stor andel utlandsfödda arbetstagare som på arbetsmarknaden generellt. På grund av hur regleringen kring lojalitetsplikt ser ut idag kan den bidra till inlåsningseffekter för utlandsfödda arbetstagare som arbetar i uthyrningsbranschen. Skilda villkor för svensk- och utlandsfödda såsom statisk diskriminering på arbetsmarknaden gör att utlandsfödda arbetstagare löper större risk än svenskfödda arbetstagare att få stå utanför den traditionella arbetsmarknaden.

Un service du nuage peut employer des techniques de sécurités pour assurer la sécurité de l’information. Ces techniques protègent une donnée personnelle en la rendant inintelligible pour toutes personnes autres que l’utilisateur du service. En contrepartie, certaines fonctionnalités ne peuvent plus être implémentées. Par exemple, la technique du chiffrement symétrique rend les données inintelligibles, mais empêche le calcul sur ces données.Cette thèse avance qu’un service du nuage doit composer les techniques pour assurer la sécurité de l’information sans perdre de fonctionnalités. Elle se base sur l’étude de la composition de trois techniques qui sont le chiffrement, la fragmentation verticale et les calculs côté client. Cette étude montre que la composition sécurise sans perdre de fonctionnalités, mais complexifie l’écriture du service. La thèse propose alors un nouveau langage pour l’écriture de services du nuage qui assurent la sécurité des données personnelles par compositions des techniques de sécurité.Ce langage est muni de lois algébriques pour dériver,systématiquement, un service local sans protection vers son équivalent sécurisé du nuage. Le langage est implémenté en Idris et profite de son système de type expressif pour vérifier la composition correcte des techniques de cryptographie. Dans le même temps, un encodage traduit le langage en ProVerif, un vérificateur de modèle pour l’analyse automatique des propriétés de sécurité sur les protocoles cryptographiques. Cette traduction vérifie alors la sécurité des données personnelles dans le service
A cloud service can use security techniques to ensure information privacy. These techniques protect privacy by converting the client’s personal data into unintelligible text. But they can also cause the loss of some functionalities of the service. For instance, a symmetric-key cipher protects privacy by converting readable personal data into unreadable one. However, this causes the loss of computational functionalities on this data.This thesis claims that a cloud service has to compose security techniques to ensure information privacy without the loss of functionalities. This claim is based on the study of the composition of three techniques: symmetric cipher, vertical data fragmentation and client-side computation. This study shows that the composition makes the service privacy preserving, but makes its formulation overwhelming. In response, the thesis offers a new language for the writing of cloud services that enforces information privacy using the composition of security techniques. This language comes with a set of algebraic laws to systematically transform a local service without protection into its cloud equivalent protected by composition. An Idris implementation harnesses the Idris expressive type system to ensure the correct composition of security techniques. Furthermore, an encoding translates the language intoProVerif, a model checker for automated reasoning about the security properties found in cryptographic protocols. This translation checks that the service preserves the privacy of its client

Magister Legum - LLM (Criminal Justice and Procedure)
The combating of money laundering has become crucial since it has escalated from a domestic to an international problem and vice versa. South Africa, in an attempt to combat money laundering has enacted anti-money laundering (AML) and anti-terrorism legislation. The legislation consists of the Prevention of Organised Crime Act 121 of 1998 (POCA); the Financial Intelligence Centre Act 38 of 2001 (FICA); and the Protection of Constitutional Democracy against Terrorist and Related Activities Act 33 of 2004 (POCDATARA). In essence POCA and POCDATARA criminalise money laundering and terrorist financing. FICA requires certain professions, such as the legal profession to maintain specific controls. Such controls include AML measures, combating the financing of terrorism (CFT) and also require the keeping of confidential client records, the filing of suspicious transaction reports (STR's) and certain cash transaction reports (CTR's).

The aim of this thesis is to specify certain legal relationship arising between attorney and his client. This legal relationship characterises, to certain extent, role of attorney in the legal system of the Czech Republic and that is why it deserves attention. The main legal source for this thesis is Act No. 85/1996 Coll., on the Legal Profession, as subsequently amended. This act is interpreted in following chapters in coherence with other acts, subordinate legislation and professional rules. After brief introduction into the issue I focus on particular causes of the formation of attorney-client legal relationship. The origin itself is determining for specification of mutual rights of both subjects. Besides talking about the formation, I will also mention the rejection and termination of the legal relationship. The next chapter deals with the ethics in advocacy, which we should keep in mind while interpreting particular rights and obligations. Ethics is characteristic sign of attorney's profession. Chapters 5 and 6 are the actual core of this thesis, when interpreting individual rights, that constitute the actual attorney-client legal relationship. I systematically divided those rights and obligations according to the subject, which has the duty to ensure those obligations. In Chapter 5 I focus on...

Resumé Theme of this thesis is legal relationship attorney - client and their mutual rights and obligations. I focused on the current legal regulation contained in the Act No. 85/1996 Coll., advocacy law, and related legal regulations, primarily in professional regulation published by Czech bar association. My objective was to try to give all- embracing explanation of this theme even though I realised this theme and all its aspects is too extensive to be explained completely. This thesis is divided into six parts, which are further systematically divied into chapters. First part gives basic historical review of development of legal regulation of advocacy since ancient Rome. Second part gives definitions of the most important terms of this thesis - attorney and client, with evident orientation on legal regulation of conditions which must fulfil a person in orded to become an attorney. Third part deals with possible methods of constitution of legal relationship between attorney and client (to conclude a contract or to be appointed by court or to be designated by Czech bar assosiation) and related attorney's obligation to refuse to plead client in certain situations as well as termination of this relationship. Fourth part is dedicated to individual rights and obligations o attorneys and clients. Each one is...

The thesis is focused on the internal relationship between the lawyer and the client. Although this topic was already processed many times, I believe that, in particular, in the context of the current case law and the amendment of certain provisions, this work may contribute to the better understanding of this topic. The thesis is systematically divided into six chapters. In the first chapter, there is outlined the nature of this relationship and stressed out the importance of mutual trust. To further break down the individual rights and obligations in detail, it is necessary to outline the formation of the relationship between solicitor and client first. It is important to note that the specific obligations imposed on lawyer legislation even before the formation of the legal relationship. Under the conditions described in the first chapter, the lawyer has a duty to refuse to provide legal services to the client, which can be considered as the pre-contractual obligation. Attorney has an obligation, although there is still no relationship between them. Considering the entire work, the second and third chapters are the most important ones. In the second chapter I push the reciprocal rights and obligations in the context of their legal arrangements, both legal and underpinned, and the professional and...