Nous considérons ici le problème où deux programmes différents situés sur deux grappes d'ordinateurs distantes, reliées par un réseau à haut débit, forment un couplage de code et échangent<br />régulièrement des données. Un tel échange s'effectue par une redistribution de données. Nous étudions comment effectuer une telle redistribution le plus efficacement possible en minimisant temps de communication et congestion du réseau.<br /><br />Nous utilisons pour ce faire, une modélisation du problème à l'aide de graphes bipartis. Le modèle choisi permet une prise en compte du délai d'initialisation des communications, des différentes bandes passantes et impose une limite d'une communication simultanée par interface réseau (modèle 1-port) et de k communications simultanées sur la dorsale.<br /><br />Nous effectuons une validation expérimentale du modèle puis l'utilisons pour développer deux algorithmes d'ordonnancement<br />des communications. Nous montrons que chacun d'entre eux<br />est un algorithme d'approximation garantissant un temps d'exécution dans le pire des cas 8/3 fois plus élevé que le temps optimal.<br />Nous concluons l'étude de ces algorithmes par une série d'expériences démontrant de bonnes performances en pratique.<br /><br /><br />Enfin, nous étendons le problème initial au cas de grappes hétérogènes :<br />ce cas imposant de sortir du modèle 1-port, nous montrons comment modifier nos algorithmes pour en tirer parti.<br />Nous étudions également le cas de redistributions exécutées en régime permanent sur un réseau d'une topologie plus complexe autorisant les communications locales.

La technologique utilisée dans les réseaux de télécommunication à commutation de paquets est en développement permanent. La tendance actuelle est orientée vers la mutualisation des services voix et données, de leurs infrastructures et des techniques associées (réseaux IP/MPLS, QoS dans Internet). Les méthodes classiques utilisées pour l'évaluation de performances des ces réseaux sont la modélisation analytique et la simulation événementielle. La simulation événementielle requiert des temps de calcul prohibitifs, tandis que les modèles issus des méthodes analytiques manquent parfois de précision. Le travail présenté dans cette thèse définit un cadre de modélisation appelé Simulation Hybride Distribuée qui combine de manière rigoureuse la théorie différentielle du trafic et la simulation événementielle. Dans une première technique, les ressources du réseau sont partitionnées en domaines dont certains sont simulés par événements, et les autres sont modélisés par des équations intégrées numériquement. La seconde technique proposée permet la circulation de trafics simulés par événements sur les ressources du réseau modélisées par des équations. La simulation hybride permet ainsi de prolonger la modélisation analytique au-delà de ses limites théoriques. Elle permet également de concevoir des modèles d'évaluation de performances très généraux et de développer des logiciels plus performants en temps de calculs. Un prototype de simulateur hybride a été conçu. Il permet de modéliser précisément des routeurs Internet Diffserv et MPLS, le protocole TCP ainsi que diverses sources de trafic multimédia (Audio, Vidéo). Pour réduire les temps de calcul, le simulateur peut être parallélisé. En ce sens, un nouveau noyau de communication de l'environnement de parallélisme LANDA a été conçu. Il offre une bibliothèque à la norme MPI et permet d'utiliser simultanément et efficacement (latence, bande passante) plusieurs media de communication haut-débit ainsi que la mémoire commune S MP.

To date, hundreds of millions of mini-QWERTY keyboard equipped devices (miniaturized versions of a full desktop keyboard) have been sold. Accordingly, a large percentage of text messages originate from fixed-key, mini-QWERTY keyboard enabled mobile phones. Over a series of three longitudinal studies I quantify how quickly and accurately individuals can input text on mini-QWERTY keyboards. I evaluate performance in ideal laboratory conditions as well as in a variety of mobile contexts. My first study establishes baseline performance measures; my second study investigates the impact of limited visibility on text input performance; and my third study investigates the impact of mobility (sitting, standing, and walking) on text input performance. After approximately five hours of practice, participants achieved expertise typing almost 60 words per minute at almost 95% accuracy. Upon completion of these studies, I examine the types of errors that people make when typing on mini-QWERTY keyboards. Having discovered a common pattern in errors, I develop and refine an algorithm to automatically detect and correct errors in mini-QWERTY keyboard enabled text input. I both validate the algorithm through the analysis of pre-recorded typing data and then empirically evaluate the impacts of automatic error correction on live mini-QWERTY keyboard text input. Validating the algorithm over various datasets, I demonstrate the potential to correct approximately 25% of the total errors and correct up to 3% of the total keystrokes. Evaluating automatic error detection and correction on live typing results in successfully correcting 61% of the targeted errors committed by participants while increasing typing rates by almost two words per minute without introducing noticeable distraction.

El estudio examinó la posible influencia del lenguaje y lugar de origen de un lector en la comprensión de mensajes de comunicación en salud y en la formación de actitud e intención hacia la realización de una conducta preventiva. Un instrumento de medición con escala de comprensión de lectura, actitud e intención fue suministrado a un grupo de puertorriqueños y dominicanos N=40) residentes en San Juan, Puerto Rico. La escala de comprensión tenía dos textos en diversos niveles de dificultad (bastante fácil y medio) de acuerdo con la Fórmula de Legibilidad para el idioma español de Francisco Szigriszt Pazos. Se encontró que no hubo diferencia estadísticamente significativa (p>.05) entre la muestra en el nivel de dificultad bastante fácil, no obstante en el nivel de dificultad medio sí hubo diferencia estadísticamente significativa (p<.05). Los análisis reflejaron que la comprensión y actitud en conjunto no predicen de manera significativa el grado de intención para la adopción de una conducta preventiva de salud. / This study examines the possible influence of a reader's place of origin on the comprehension of health communication messages and on the formation of attitudes and intentions of preventive behavior. A measurement instrument with a reading comprehension, attitude and intention scale was administered to a group of individuals of Puerto Rican and Dominican origin (N=40) living in San Juan, Puerto Rico.The reading comprehension scale used two texts with different levels of difficulty (relatively easy and intermediate) in accordance with the Readability Formula for Spanish of Francisco Szigriszt Pazos. There was no statistically significant difference (p>.05) among the samplefor the relatively easy level of the reading test; however, for the intermediate level, there was a statistically significant difference (p<.05). Analyses showed that comprehension and attitude together do not significantly predict the degree of intention to adopt healthy preventive behavior.

Att ta ett större ansvar inom miljömässiga, ekonomiska och sociala frågor blir en allt viktigare punkt på många företags agendor, och en allt betydelsefullare aspekt för konsumenter att ta hänsyn till när de överväger köp av produkter eller tjänster. Företagsansvar, internationellt benämnt som corporate social responsibility eller CSR, har också fått ett genombrott bland myndigheter och andra officiella organ såsom FN och EU, som jobbar för att underlätta för företag genom olika riktlinjer och redovisnings­standarder för CSR. Fristående intresseorganisationer har också en viktig del i CSR-frågan, både som samarbetspartner och som granskare. Många av företagens intressenter är intresserade just av statistik och resultat, men hur kommunicerar ett konsumentföretag till sina kunder att de satsar på CSR-frågor? Hur kan företag använda av olika typer av strategier, kanaler och budskap för att stärka sitt varumärke och nå ut med sitt ansvarstagande? Trots att arbetet med CSR ofta föregås av noggranna strategiska överväganden och diskussioner om hur CSR påverkar varumärket, är det inte alla företag som har en tydlig plan för hur de ska kommunicera CSR mot konsumenter. Denna uppsats syftar till att undersöka hur kommunikationen kring företags CSR-arbete sker, ser ut och uppfattas hos den intressentgrupp som företagets produkter eller tjänster riktar sig mot ­– konsu­menterna. I en form av nulägesanalys har fallstudier utförts på två svenska konsumentföretag, Arla och Löfbergs Lila, som båda har inledda CSR-åtaganden och har certifierade produkter i sitt sortiment. I fallstudierna ingick djupintervjuer med företagens CSR-ansvariga, kvalitativa innehålls­analyser av deras marknadsföringsmaterial och en enkätundersökning om konsumenternas kännedom och agerande inom området samt deras uppfattning av fallstudieföretagens ansvarstagande. Kopplat till en referens­ram innehållandes teorier om CSR som begrepp, CSR i dagens samhälle och hur CSR-orienterad kommunikation kan ske har frågor och modeller för dessa metoder tagits fram och analyserats. Analysen visar på att inget av företagen ännu har en tydlig strategi för sin CSR-kommunikation. Företagen har implementerat CSR-arbetet väl internt inom företaget, men kommunicerar bara i liten utsträckning externt mot konsumenter. Det finns en svårighet i att involvera konsumenter i CSR-frågan och låta denna dialog influera och ingå i CSR-kommunika­tionen. Båda företagen ger tillgång till mycket information kring sitt ansvarstagande via sina webbsidor, men webbsidornas information tycks inte nå fram till konsumenterna då de inte är en välintegrerad del av marknadsföringen. Primärt knyts CSR-värden till företagens pro­dukter, och konsumenter relaterar företagens ansvarstagande med deras certifierade produkter framför­allt. Företagen använder sig av olika typer av kommunikations­budskap, vilka i Arlas fall är för vaga och visionsfokuserade, och i Löfbergs Lilas fall har svårt att väcka känslor för varumärket och ansvarstagandet. Konsumenterna visar på ett intresse för att agera i ansvars­frågor samt en viss kunskap inom området, men när det kommer till hur de uppfattar företagens ansvars­tagande finns det en stor osäkerhet och låg kännedom om vad företagen faktiskt gör inom CSR-området. En stor del av respondenterna har inte fått reda på någon information om hur företagen tar ansvar, och det är ännu inte många bland konsumenterna som aktivt söker upp information eller deltar i en dialog med företagen. / To take greater responsibility in environmental, economic and social issues is an increasingly important item on many companies' agendas, and an increasingly important aspect for consumers to take into account when considering purchase of a product or a service. This field, internationally called corporate social responsibility or CSR, has also made a breakthrough among authorities and other official orga­nizations such as the UN and the EU, which are facilitating for companies engaging in CSR by provi­ding various guidelines and accounting standards for CSR. Non-governmental organizations are also an important part of the CSR issue, both as partners and as reviewers. Many of the companies' stakehol­ders are interested in statistics and results, but how do consumer businesses communicate to their customers that they are investing in CSR issues? How can companies use different kinds of strategies, channels and messages to strengthen their brand and make the public notice their responsibility? Although the work with CSR is often preceded by careful strategic considerations and discussions about how CSR affects the brand, not all companies have a clear plan how to communicate CSR to consumers. This thesis paper aims to examine how information concerning the companies' CSR initiatives is communicated, designed and perceived by the stakeholder group which the products or services are made for – the consumers. In the form of a situation analysis, case studies have been conducted on two Swedish consumer companies, Arla and Löfbergs Lila, both of which have initiated CSR commitments and have certified products in their range. The case studies included interviews with corporate CSR managers, qualitative content analysis of the companies' marketing materials and a survey of consumer knowledge and behavior in CSR issues as well as of the customers' view of the case study companies' responsibility. Linked to a theoretical framework including theories of CSR as a concept, CSR in today's society and how CSR-oriented communication can be designed, the issues and models for these methods have been developed and analyzed. The analysis shows that neither of the companies have a clear strategy for their CSR communication. The companies have implemented CSR well within the organisation, but communicate only to a small extent externally. There is a difficulty in involving consumers in CSR issues and making this dialogue influence and be part of the CSR communication. Both companies provide access to much information about their responsibility through their web pages, but this information does not appear to reach consumers as the web pages are not well-integrated in the marketing. Primarily, the CSR values are connected to the companies' products, and consumers in particular relate the companies' CSR involvement to their certified products. The companies use different types of communication messages, which in Arla's case are too vague and vision focused, and in Löfbergs Lila's case have difficulties to arouse fee­lings for the brand and the CSR issues. Consumers show an interest to act on CSR issues and have a certain knowledge in the field, but when it comes to how they perceive the companies' CSR initiatives, there is a great uncertainty and low knowledge of what companies actually do in the CSR field. A large proportion of the respondents has not found out any information about how the companies are taking responsibility, and it is at this time not many among the consumers who actively seek out information or participate in a dialogue regarding CSR issues with companies.

The overall aim of this thesis is to investigate the effects of coherence in the corporate branding process. Specifically, this thesis defines and develops the concept of corporate brand coherency (CBC) using four antecedents and four dimensions, and investigates the effects of incongruence based on the advertising antecedent by manipulating ad-brand congruence. Using the Elaboration Likelihood Model (ELM), the research tests whether two-sided messages aid the resolution of incongruence following central route processing and whether greater number of arguments could act as a peripheral cue. Accordingly, a conceptual model was developed to test the effects of ad-brand congruence, message sidedness and argument quantity on corporate credibility, attention and elaboration, attitude towards the ad and attitude towards the corporate brand. To empirically test this model an online experiment (using a 3x2x2 between-subjects factorial design) was conducted, where subjects were exposed to a modified print advertisement for a grocery store. A total of 528 responses were collected from New Zealand Internet users. Two- and three-way ANCOVAs as well as path analysis was used to analyse the hypothesised relationships. The results indicate support for CBC, as it was found that congruence led to positive evaluations and higher corporate credibility. Two-sided messages were not found to be effective in reducing the negative effect of incongruence. Instead, two-sided messages led to less favourable evaluations of incongruence. Regardless of the incongruence, the results show that two-sided messages did not lead to positive evaluations, while an increased number of arguments did lead to more favourable attitudes. Overall, the findings show the ELM is not applicable to explain consumers’ responses to incongruence, as incongruence was not found to affect message processing. Additionally, there was no support that message sidedness and argument quantity, respectively, acted as central and peripheral cues. Furthermore, findings indicate that incongruity in advertising can lead to positive evaluations when consumers have a negative prior brand attitude. The theoretical and managerial implications as well as directions for future research based on these findings are discussed.

Real-time information that informs transit riders about transit schedules, next bus or train arrivals, and service alerts, is becoming increasingly available, particularly through internet-enabled smartphone applications. However, the extent of communication technology usage amongst transit riders, specifically their access to mobile applications and alternative technologies that can provide real-time information, is largely unknown. Without this information, transit agencies are risking investing in an alternative technology that may not sufficiently supply real-time information to as many as possible riders. The purpose of this study is to identify the differences in individual technology accessibility and prioritize investing in real-time information application development that mirrors the unique characteristics of transit riders. This recognition and development will allow a wider availability of real-time information amongst transit riders. Paired with an investigation of cellular phone usage among transit riders and the general American population, an analysis of Saint Louis Metro’s Onboard Survey was performed. Cross tabulations and chi-squared tests were conducted to examine riders’ communication technology usage. Binary logit models were used to understand how, and whether, the ownership of smartphone applications is dependent on various demographic factors. These analyses identified specific demographic groups that would benefit from supplemental technology methods more conducive to their particular information accessibility. Results showed that communication technology usage has risen substantially in recent years but a portion of riders are still without access to smartphone applications. Specific demographic groups (e.g., riders over 40 years of age) were less likely to own smartphones, and these results indicate that computer-based websites and IVR are the best supplementary alternatives for those groups.

This study examines the conditions for successful online mentoring in order to develop writing skills in English in a workplace setting. Chapter 1 gives the background and context of the study. Problems to be addressed in the study and the aims, objectives, hypotheses and their rationale are presented. This is followed by testing procedures, research design, sources of data and research procedures. In Chapter 2, the literature review supports the hypotheses on the need for collaboration in materials development and delivery, mentoring relationships, motivation and computer and Internet efficacy. Chapter 3 presents the findings from the case study bringing into focus problems that would jeopardise a mentoring programme if training providers do not pay attention to the hypotheses. The findings are collated and the hypotheses are confirmed. Conditions for successful online mentoring are spelt out in Chapter 4. The study concludes that online mentoring works once the conditions are properly followed. / English Studies / M.A. (TESOL)

Internal marketing and internal communication are essential tools to align em-ployees’ mindsets with the necessary tasks which bring about employee satis-faction and organisational prosperity. Electronic newsletters serve as an internal marketing communication medium which can convey the necessary information to employees regarding the organisation’s goals and objectives. They also act as a medium to build relationships with employees and encourage improved service delivery and customer-oriented employee mindsets. Employees’ willingness to make use of electronic newsletters for the purpose of disseminating organisational information is an indication of employees’ acceptance of the organisational information and use thereof in performing organisational tasks successfully as the organisation intends. The purpose of the current study was to determine employees’ acceptance of electronic newsletters, as an internal marketing communication medium, in order to disseminate organisational information which contributes to the attainment of organisational goals, objectives and success. An empirical study was conducted to determine employees’ acceptance of elec-tronic newsletters by means of an adapted technology acceptance model, self-administered, e-mail survey disseminated to employees of a higher education institution. The study followed a quantitative research approach, utilising re-gression in the analysis of the data. Based on the research results, employees do accept the electronic newsletter for the dissemination of organisational information, albeit with suggestions on how it can be better utilised in future. / Business Management / M. Com (Business Management)

abstract: Assemblers and compilers provide feedback to a programmer in the form of error messages. These error messages become input to the debugging model of the programmer. For the programmer to fix an error, they should first locate the error in the program, understand what is causing that error, and finally resolve that error. Error messages play an important role in all three stages of fixing of errors. This thesis studies the effects of error messages in the context of teaching programming. Given an error message, this work investigates how it effects student’s way of 1) understanding the error, and 2) fixing the error. As part of the study, three error message types were developed – Default, Link and Example, to better understand the effects of error messages. The Default type provides an assembler-centric single line error message, the Link type provides a program-centric detailed error description with a hyperlink for more information, and the Example type provides a program centric detailed error description with a relevant example. All these error message types were developed for assembly language programming. A think aloud programming exercise was conducted as part of the study to capture the student programmer’s knowledge model. Different codes were developed to analyze the data collected as part of think aloud exercise. After transcribing, coding, and analyzing the data, it was found that the Link type of error message helped to fix the error in less time and with fewer steps. Among the three types, the Link type of error message also resulted in a significantly higher ratio of correct to incorrect steps taken by the programmer to fix the error. / Dissertation/Thesis / Masters Thesis Software Engineering 2017