Ver Post

Dissertações MERC

25 Maio 2012, 10:04 - Maria João Silva Carvalho

Dissertação MERC: Sistema de distribuição de conteúdos para Digital Signage

Candidato: Nº 57713 Hugo Alexandre da Silva Pires 

Presidente: Professor Paulo Jorge Pires Ferreira

Orientador: Professora Teresa Maria Sá Ferreira Vazão Vasques

Co-Orientador: Professor Ricardo Jorge Feliciano Lopes Pereira

Vogais: Professor Paulo Rogério Barreiros D' Almeida Pereira

Data: 29/05/2012 -16:00h / Sala 1.38 IST-TagusPark.

Abstract: Following the actual continuous price drop on display devices with large dimensions and a relatively high range make them an attractive way to display and disseminate information. Content delivery also has an added value when it comes to advertising and content diffusion. These two concepts together turns content disclosure by digital signage in a attractive and complete concept. This thesis follows the content delivery by digital signage theme, creating a viable solution to develop and implement in some small and medium organizations. The main requirements for this project are flexibility, adaptability, low cost and the possibility of a future functionality extension. The solution developed is based in the concept of content capture through remote desktop sessions. This way, it is possible to unify them, and allow their diffusion to different kind of platforms in the same way. To complement this idea, a grouping of target audience is done, in order to display the right type of content to an audience that is more interested and available to retain that information. The last concept integrated in this solution is the concatenation of distinct types of contents in a single video stream, in order to allow different displaying modes to the audience.

Key Words: signage, content, remote desktop, delivery


Dissertação MERC: Energy efficient architectures for the current and future Internet


Candidato: Nº 57819 Tiago Filipe da Cruz Silva

Presidente: Professor Paulo Jorge Pires Ferreira

Orientador: Professor Artur Miguel do Amaral Arsénio

Vogais: Professora Teresa Maria Sá Ferreira Vazão Vasques

Data: 04/06/2012 -17:30h / Sala 1.38 IST-TagusPark.

Abstract: The energy consumption is a major concern nowadays not only because of environmental issues but also economical ones. The current Internet infrastructure wastes a lot of energy because the network elements are always working at their full capacity even with a low traffic demand. This energy waste of the Internet can be reduced by allowing some network elements to enter in energy saving modes. However, it may lead to the decrease of the network performance. This work consists on applying an energy saving model to the current Internet architecture and to the PSIRP architecture, taking into account the tradeoff between energy saving and network performance. This is mainly achieved by classifying the network elements according to their importance in the packet delivery process.

Key Words: Internet, Clean Slate Design, Energy consumption, Energy saving, Traffic engineering, Turning off network elements.


Dissertação MERC: PhoneSensing - Smartphone-based Application for Distributed Sensing in Human Networks

Candidato: Nº 55955 Nadir Amaral Turkman

Presidente: Professor Paulo Jorge Pires Ferreira

Orientador: Professor Rui Manuel Rodrigues Rocha

Vogais: Professor João Coelho Garcia

Data: 04/06/2012 -14:30h / Sala 1.38 IST-TagusPark.

Abstract: The evolution and proliferation of modern smart-phones has opened a door to a new type of Wireless Sensor Networks (WSN). While traditional WSNs consist of fixed sensor nodes, smart-phone based ones have a large mobility factor. Modern smart-phones also possess a respectable amount of sensors which allied with their mobility and processing power are able to produce a large-scale Opportunistic Sensor Network (OSN). The sensors also provide developers with the possibility to create applications that improve lifestyle by collecting useful contextual information from the user?s lifestyle. We propose to explore both the opportunistic and social potential by creating an application that uses an underlying OSN to improve the user?s day to day personal and social experience. To explore the opportunistic potential we propose a decentralized resource sharing protocol to apply to the underlying OSN. It will provide the possibility for mobile nodes to exchange resources and supply the necessary framework for the social sensing application.

Key Words: Smart-phone, Sensor, Sharing, P2P, Sensing, PhoneSensing


Dissertação MERC: Análise de Sequências em Tempo Real

Candidato: Nº 52878 Diana Mafalda Pereira Soares

Presidente: Professor Paulo Jorge Pires Ferreira

Orientador: Professor Diogo Manuel Ribeiro Ferreira

Vogais: Professor Andreas Miroslaus Wichert

Data: 29/05/2012 -14:30h / Sala 1.38 IST-TagusPark.

Abstract: Currently, many organizations use information systems to perform and to automate their business processes. These information systems record all events in an event log. Process Mining techniques allow to analyze event logs and see what process is being executed, who executes, its duration and if is in accordance with the plan. This analysis is performed based on information gathered over a given period of time. As a result, it may not reflect the current state of the organization. To discover patterns of sequences based on events recorded in the event logs clustering techniques can be applied. This clustering techniques aim to organize a set of objects into groups, each one contains objects with similar behavior. When applied to the event logs they organize their information, making possible to extract knowledge. It would be important for organizations to use these techniques in real time, providing information on what is happening, instead of using facts for the analysis based on past occurrences. This thesis proposes sequence clustering algorithms implemented for real time, as the events are being produced and recorded in the event log by an information system. These new algorithms also allow organizations to carry out all types of analysis techniques that Process Mining provides and with the most updated information instead the information gathered from previous times. The three main contributions of this work are the overview of the main clustering techniques, the use of these techniques to sequences, and finally to adapt the algorithms to real-time sequences.

Key Words: Event log, Sequence, Cluster, Algorithm, Process Mining, Real time.


Dissertação MERC: Decision Support Systems for Clinical Benchmarking: Design and Implementation of a Data Warehouse for Benchmarking in Clinical Rehabilitation

Candidato: Nº 54458 João Luís Dias Completo

Presidente: Professor Paulo Jorge Pires Ferreira

Orientador: Professora Maria Luísa Torres Ribeiro Marques da Silva Coheur

Co-Orientador: Rui António dos Santos Cruz

Vogais: Professora Helena Isabel de Jesus Galhardas

Data: 31/05/2012 -14:30h / Sala 2.10 IST-TagusPark.

Abstract: Clinical Benchmarking provides comparative analysis among healthcare institutions and healthcare services, in order to improve their quality and efficiency. This work describes the design and implementation of a Data Warehouse, following Kimball's Business Dimensional Lifecycle framework, to support a Decision Support System for Benchmarking in Clinical Rehabilitation. For the Data Warehouse development, the identification of the most important concepts as well as the data modeling process are the most challenging and important tasks, requiring deep knowledge on the subjects in matter. In order to accomplish that goal, an extended research was made in order to better understand the main concepts on Clinical Benchmarking and Clinical Rehabilitation, and to address the most suitable technologies regarding Data Warehouse architectures and On-line Analytical Processing mechanisms. Consequently, the developed Data Warehouse respects the specific characteristics related to Clinical Rehabilitation, addressed through a multi-dimensional modeling process, using different Data Marts for each of the major business processes. The evaluation of the solution using real clinical records demonstrates a clear orientation to the subject in matter as well as the capability to provide accurate current and historical analysis.

Key Words: Clinical Benchmarking, Data Warehouse, Decision Support Systems, Rehabilitation


Dissertação MERC: Implementação e monitorização de uma rede de sensores móveis para medição da qualidade do ar

Candidato: Nº 57697 Fábio Ferreira Gameiro

Presidente: Professor Paulo Jorge Pires Ferreira

Orientador: Professor João Pedro Castilho Pereira Santos Gomes

Vogais: Professor António Manuel Raminhos Cordeiro Grilo

Professor Paulo Jorge Coelho Ramalho Oliveira

Data: 31/05/2012 -16:00h / Sala 2.10 IST-TagusPark.

Abstract: The monitoring of atmospheric gases, especially polluting and endangering people?s health when found in high concentrations, has been something that has concerned public health officials and the citizens themselves more and more. The monitoring is currently, in its vast majority, made by stations of a fixed nature with few points of data collection. The use of mobile atmospheric monitoring stations integrated on a network of public buses creates the possibility of monitoring at various points of a city taking advantage of its mobility. One of the challenges that this type of monitoring arises is the communication between the atmospheric information gathering stations and the server that stores that information in a data base. The objectives of this thesis began by creating and implementing communications between mobile stations and a central server using technologies General Packet Radio Service (GPRS) and Wi-Fi in order to operate autonomously. In addition to developing communications, it was created a web platform in order to view the samples collected by the stations and also provide monitor and configure operations remotely. This document presents the state of the art in areas where this work was developed, especially in vehicular communications. It described the solutions created in order to meet the objectives of this thesis and also evaluated the functioning and performance of solutions. The evaluation made to the implemented work proved the correct operation of its features and a very satisfactory performance by them.

Key Words: Mobile Sensor Network, Vehicular Network, Machine To Machine Communication, Atmospheric Monitoring


Dissertação MERC: ePaga - sistema de pagamento electrónico

Candidato: Nº 55986 João António Frade Macedo de Almeida

Presidente: Professor Rui Jorge Morais Tomáz Valadas

Orientador: Professor Carlos Nuno da Cruz Ribeiro

Co-Orientador: Professor Paulo Jorge Pires Ferreira

Vogais: Professor Nuno Filipe Valentim Roma

Data: 04/06/2012 -16:00h / Sala 1.38 IST-TagusPark.

Abstract: The possibility of using mobile devices for payments promises greater speed, convenience and ubiquity when compared with current payment methods. Hence the large potential demonstrated by the mobile payments market. However, the lack of interoperability between payment systems, along with the absence of suitable technologies, has delayed the progress in the field of mobile payments. This project?s goal is to develop a support system for electronic payments, with a special focus on mobile payments, which takes advantage of multiple payment protocols installed on the same device. The application derived by this multiprotocol support achieves higher universality, efficiency and interoperability, when compared to a single protocol application. The proposed system is based on NFC (Near Field Communication) technology. This document also describes the prototype developed with the purpose of demonstrating the system?s behavior, as well as its evaluation.

Key Words: mobile computing, electronic payments, mobile payments, NFC.