Skip to main content

Trinity College Dublin, The University of Dublin

Menu Search



Showcase 2015

Showcase of MSc Dissertations

Thursday 3rd September 2015 

O’Reilly Institute

4:00p.m. - 4:45 p.m.

  • Opening and outline of MSc programmes and school activities: Prof Lucy Hederman, Director of Postgraduate Teaching and Learning, School of Computer Science and Statistics
  • Dean of Graduate Studies, Prof Aideen Long
  • Keynote Speaker: Ken Perlin, Professor of Computer Science, NYU
  • Student "fast-forward" comprising one minute presentations of selected student posters

 

4:45 p.m. - 6 p.m.

  • Poster display, demonstrations and installations at which students, academics and industry interact 

    Keynote Speaker: Professor of Computer Science: Ken Perlin,  Media Research Lab in the NYU 

    Ken Perlin, a professor in the Department of Computer Science at New York University, directs the NYU Games For Learning Institute, and is a participating faculty member in the NYU Media and Games Network (MAGNET). He was also founding director of the Media Research Laboratory and director of the NYU Center for Advanced Technology. His research interests include graphics, animation, augmented and mixed reality, user interfaces, science education and multimedia.

    Inventing the Future

    "The question our lab at NYU is asking is "How might people in the future communicate with each other in every day life, as computation and display technologies continue to develop, to the point where computer-mediated interfaces are so ubiquitous and intuitive as to be imperceptible?" To address this, we are combining cutting edge virtual reality and motion capture technologies, to allow people to walk freely around, unconstrained by cords or cables, in physical space, interacting with other people and physical objects, just as they do in everyday life. Yet everything those people see is computer-mediated, thereby allowing them to share any reality they wish, and create and interact with virtual objects and creatures in the air, as though by magic. But today's magic will be tomorrow's reality"

This event provides an opportunity for business, industry, the public sector and academia to link with graduates who are showcasing their research on six MSc programmes in the School of Computer Science and Statistics: Interactive Entertainment Technology; Mobile and Ubiquitous Computing; Networks and Distributed Systems; Technology and Learning; Interactive Digital Media, and Health Informatics. It also offers an opportunity for industry to interact with academics and researchers identifying possible future collaborations.


Graduates on these programmes have gained experience with the latest tools and technologies and have been exposed to cutting-edge research being conducted within the School. The nature of the work undertaken equips these graduates with transferable skills relevant to careers in the global technological industry. As part of their MSc studies, students undertake an individual research project leading to a submission of a dissertation of publishable standard.This dissertation reflects an independent piece of exploratory research which significantly contributes to the advancement of technology and involves design, development and evaluation.

If you would like to attend, please register your interest to: events@scss.tcd.ie

Guide to selected posters, demonstrations and Installations

Please check for updates

Poster Display  (Foyer in O'Reilly Institute)

A Personalised Ontology and Rule-based Approach to Managing Message Overload

 

Student:              FengJiao LV          
Supervisor:         Prof Owen Conlan
Course:                MSc in Computer Science (Networks and Distributed Systems)

Large numbers of messages keep annoying, distracting people from current tasks and interests. Also, messages take up valuable resources of devices and deteriorate performance.
The thesis provides a novel approach in handling messages; Message assistant system is a test-oriented, light-weight, efficient and intelligent one, based on rule-based approach, integrating context, and user preferences. Moreover, this approach contributes to the state of the art by offering a distributed architecture.
In addition, the system was using iterative development and three-stage experimentation, which carried on series of experimental and analytical tests on performance, scalability and load balance. 
A key challenge of the system is how to deal with large numbers of messages. The thesis proposed a chaining-process, distributed design to make phased operational decision on uplifted messages. Another challenge is handling novel messages and novel senders. This research designed rules to automatically learn and adapt the results in run-time. 

A Real-time and Location-based advertising model


Student:              Jinwei Gu          
Supervisor:         Prof Siobhán Clarke
Course:                MSc in Computer Science (Networks and Distributed Systems)

Nowadays, advertisement is seen as a vital part of business marketing. Sellers are engaging in various kinds of approaches to advertise their stores or products as more and more advertising models are available. In the context of increasing competition, it is getting harder for shop owners to increase their markets and popularize their products. In many cases, shops do not get enough customer engagement even though they spend a lot of money and resources. 

In order to mitigate sellers' especially the small business's problems, this dissertation describes a real-time and location-based advertising model which aims to help small businesses to improve advertising effectiveness and save money and also compare it with current existing advertising models.
The research is conducted in the context of a project - Soosokan which is an information retrieving and publishing system. Data is collected from the experiments, questionnaires/ interviews to compare and evaluate the advertising model.

A Relationship Model for Believable Social Dynamics of Characters in Games                                               
Student:              Brendan O'Connor
Supervisor:         Prof Mads Haahr
Course:                MSc in Computer Science (Interactive Entertainment Technology)


This dissertation investigates believable social dynamics based on a relationship model between Non-Player Characters (NPCs) in games. The majority of interactions between characters in games is rigidly defined; with limited scope for character relationships that change as a result of interactions with each other or other factors. This limited ability hinders player immersion and believability in their interactions with the characters. This dissertation presents a model which can be easily applied to NPCs to provide them with relationship awareness.

The model supports characters that store their own "assumed knowledge" of other characters relationships, in addition to their own direct relationships with other characters in their environment. This allows the characters awareness of not just their own relationships to other characters (e.g. to a sibling or child character), but also awareness of the relationships between other characters (e.g. that two other characters may be enemies). This knowledge can be treated as an assumption; allowing characters to be provided with misleading information - the results of which are left up to extended implementations of the model. A messaging system is also implemented that allows the characters to communicate with one another directly as well as through their relationships (e.g. passing a message to all members of a group). The model has been implemented with a generic interface that allows other types of information to be tracked between NPCs as extensions to the basic relationships between characters. This provides for advanced implementations such as characters that "remember" previous relationship states with other NPCs.

The model is tested and evaluated within a prototype environment within the Unity game engine. The implementation showcases a model with promising potential where characters act upon, learn and communicate about relationships they are aware of.

 

An Analysis of a Relational Database and NoSQL Database on an eCommerce Platform

Student:              Xuejiao LIU
Supervisor:          Prof Siobhán Clarke
Course:                MSc in Computer Science (Networks and Distributed Systems)

Database Management System for online commerce platforms has to address demanding performance requirements, especially scalability and processing efficiency. On the other hand, from a business perspective, the usage and maintenance of their database must be cost-effective. These conditions demand that online commerce platforms must be built on a scalable and practical database.

To improve a platform’s ability to process data, this dissertation investigates different kinds of databases and how they help to mitigate issues related to large volumes of data. There are two major types of database: Relational Database and NoSQL Database. In this dissertation, the response time, error rate and throughput of different Database Management Systems are compared. Database process time complexity and space complexity are used to measure the efficiency. Furthermore, the dissertation comparatively analyzes different databases economic costs.
The research is based on a real business platform - Soosokan, which is a project conducted within the Trinity College innovation incubator, called Launchbox.

An approach to facilitate exploratory search through suggestion and reflection                                    

Student:              Yamin Xue
Supervisor:         Prof Owen Conlan
Course:                MSc in Computer Science (Networks and Distributed Systems)

An Implementation and Evaluation of a Co-rotational Finite Element Method on Mobile Architectures                                                                                                                                                                          
Student:              Giovanni Campo
Supervisor:        
Prof Michael Manzke
Course:                MSc in Computer Science (Interactive Entertainment Technology)

Finite element methods (FEM) have been an active area of research for physicalsimulations over the last 30 years. FEM is mainly used to simulate deformation and
fractures of solid objects. Its application is of particular interest in engineering and scientific fields, where accuracy is more important than plausibility. However, due
to its complexity, it is only suitable for offline simulations. Notwithstanding these limitations, FEM can be used for interactive applications. Earlier work has shown the
feasibility to run FEM in real-time contexts, on limited console hardware, using linear tensors. Instability problems, which arose from the use of linear approximations, were
successfully addressed using a co-rotational formulation.

This dissertation explores the viability of achieving a robust and real-time FEM
implementation on mobile architectures.A co-rotational FEM is fully implemented on both CPU and GPU hardware. Experiments are conducted to benchmark and to evaluate the efficiency of memory hierarchy on the Tegra architecture. The results are promising, showing interactive frame rates on both CPU and GPU implementations.

 

BLEProxiMesh : A Bluetooth Smart Mesh Network for Sensing Indoor Location

Student:              Swati Sehgal
Supervisor:         Prof Jonathan Dukes
Course:                MSc in Computer Science (Mobile and Ubiquitous Computing)

BLEProxiMesh is a low-cost, energy efficient sensor and access network for detecting the presence of people (or things) in a smart building. In particular, the sensor network facilitates the discovery and tracking of tags in a building, with a particular emphasis on security and emergency response applications.

The proposed system consists of Bluetooth Smart Tags and Beacons. Each Tag estimates its own position in the building based on “pings” received from nearby Beacons using Centroid localization algorithm and communicates its estimated position to the Beacon network through one or more nearby Beacons. This information is replicated throughout the Beacon network using the Trickle algorithm, which has been implemented using the GAP and GATT profiles of the Bluetooth Smart protocol stack. The system is designed to achieve a low-cost, energy-efficient, easy-to-deploy solution for data propagation and replication that is robust enough to be operational in emergency situations which not only involve power constraints but also power failure in the worst case. The system is dynamic and enables fast detection and propagation of changes in the network. The network is designed not only to provide a sensing platform but also to provide an access network for retrieving Tag information.

A working prototype has been implemented to validate the approach using a small-scale Beacon network. Future work will evaluate the performance of the system at a larger scale.

 

Blockchain  -  Decentralizing Critical Resources

Student:              Vineet Kumar Singh
Supervisor:         Prof Ciaran McGoldrick
Course:                MSc in Computer Science (Networks and Distributed Systems)

Bitcoin and the underpinning technology, blockchain, has gained a lot of  attention  from  industry  and academia. In this work we study the problem  of  P2P secure computation where privacy is maintained by the network rather than at end points. We use the blockchain as a tool for decentralization of critical services which are usually provided by a central authority. By modelling application logic through transactions, the requirements of trust are recast. Validation is effected by  consensus  through  leveraging  the hashing power of the network.  We  study  the  potential  application  of such a service in vehicular  networks  in  which  the  incremental  policy  updates  are provided  by  one  such  node  in the network and the complete network achieves consensus on the authenticity of the update independently. As a different example we also present a P2P file storage system in which blockchain  manages  privacy  and  access  control  of the data blocks stored  in  the  peer  to peer network. We show that blockchain can be used to build  decentralized services in the absence of trust, which previously required trust validated by a central authority.

 

Building Dynamic & Interactive Natural Game Worlds  
                                                                               
Student:              Jeremiah Dunn
Supervisor:         Prof Mads Haahr
Course:                MSc in Computer Science (Interactive Entertainment Technology)


Dynamic game worlds are game worlds which change over time according to a given rule set. Interactive game worlds allow for users to modify the state of the game world in some way. Combining these two concepts together can result in more immersive and believable game worlds. It can also lend itself to the creation of new game play mechanics.

This project will look at how natural processes can be modelled to create believable, macro-scale dynamic game worlds with high levels of user interaction. It will discuss the different components that make a natural dynamic world and explore how they can be modelled.

A prototype combining these systems together will be designed and then implemented. It will demonstrate how the different components of the dynamic world will work together. The prototype will be highly interactive, allowing the user to actively interact with these systems in real time and discus the potential applications of such a system.

 

ConPAN: A Concast service for 6LoWPAN

Student:              Bhakt Vatsal Trivedi 
Supervisor:         Prof Stefan Weber
Course:                MSc in Computer Science (Networks and Distributed Systems)

​Transport protocols were made to deliver files from source to destination without adapting to the type of data they were transferring. With the advancement of technology, particularly The Internet of Things, networks today are flooded with small packets of data. Since the IP and Transport networks do not interact, the routing of these small packets becomes an issue. Many Application oriented solutions  have been proposed, but these solutions depend on the type of data they are handling. The aim of this project is to analyze the effects of these small packets on the routers in the current architecture and then try to devise a solution to prevent the flooding of networks due to these packets. The protocol which we will try develop will provide an application independent solution to the problem.

Cooperative Localization Using Bluetooth Low Energy

Student:              Yujing Huang 
Supervisor:         Prof Jonathan Dukes
Course:                MSc in Computer Science (Mobile and Ubiquitous Computing)

Global Positioning System (GPS) is one of the most popular navigation systems which provides location and time information. The Global Positioning System (GPS) is embedded in most of the phones and tablets nowadays.  In some cases, however, localization without using Global Positioning System (GPS) would be more suitable. For example, young kids who cannot operate mobile phones and are lost.

Accurate and low cost localization without GPS remains an interesting area in research. Bluetooth Low Energy (BLE) tags have benefits of low cost and low energy consumption. If the BLE tag is surrounded by several GPS enabled devices, location data collected by devices can be used to localize the tag. Cloud servers are cost efficient and flexible. It is ideal to store location data so data can be accessed anytime anywhere. This project proposed a system to help tag owners to localize Bluetooth Low Energy enabled tags. BLE tags broadcast their existence periodically. GPS-enabled devices repeatedly scan for the presence of Bluetooth Low Energy (BLE) enabled tags. When a tag is detected, the location of the mobile device along with the proximity to the device is sent to the cloud. The tag owner can then send request to the cloud server to ask for location information

 

Depth Perception in Virtual 3D Environments

Student:              Xinwei  Xiong
Supervisor:         Prof John Dingliana
Course:                MSc in Computer Science (Interactive Entertainment Technology)


Depth perception plays significant roles in the entertainment industry. Different kinds of rendering techniques being used to increase the game's reality and helps the player to perceive the depth in video games. Depth perception is very important, especially in 3D and other games that need player react from the depth change.


This leads to questions about how these rendering techniques bring depth perception benefit into the game and which rendering techniques works better. In order to figure out this question, this dissertation was to design an experiment to test which rendering techniques, get better results in depth perception and how this rendering techniques help users perceive the depth in different scenarios.


There are 3 different kinds of rendering techniques being used in this project, which are depth of field, shadow mapping and stereoscopic rendering. It verified that these rendering techniques improve the depth perception in games.

Design Characteristics of Honeypots

Student:              YASSER Afifi          
Supervisor:         Prof Stefan Weber
Course:                MSc in Computer Science (Networks and Distributed Systems)

Digital portfolios revolution – Capturing Student voices for AFL

Student:              Colm O'Callaghan
Supervisor:         Prof Immaculada Arnedillo-Sánchez
Course:                MSc in Technology and Learning

Feedback, when used effectively and when structured appropriately can greatly impact upon how teachers teach and learners learn (Clark, 2012, Hattie and Timperley, 2007).  Within the framework of an Assessment for Learning culture, where students use their assessment data to direct, improve and ultimately regulate their own learning, technology has begun to play an important role.  This study aimed to establish a technology enhanced learning experience for first class students working in a learning support environment so that they could capture the content and power of the feedback they offered one another through dialogue and discussion. 

Dialogic feedback involves students engaging in purposeful conversations about the quality and value of their work.  It is informed by the creation of shared learning goals and continuous reflection.  Online portfolios are common among second and third level students.  Primary students, particularly students with emerging literacy skills can lack the ability to adequately write about their learning.  They rely on oral description (Dunphy, 2008) in order to express their reflections on learning.  The
study aimed to provide an online portfolio system, which would afford students the ability to capture their responses and reflections to their work and present it in an accessible manner while also serving as a valuable assessment tool.  An online portfolio platform was used to allow students capture their dialogic feedback.  It allowed them to upload work samples with accompanying comments and suggestions. 

This repository of work samples was used to direct and focus the students as they attempted to improve the quality of their learning and further their discursive and reflective capabilities.  Employing a mixed method case study, the research aimed to examine the attitudes and experiences of six first class students working in a learning support environment in a large urban primary school.  The findings suggest that students are eager to use technology to represent their learning and that dialogic feedback can be supported and encouraged through the use of an online portfolio system.  There is evidence to suggest that what students think and say about their work, can be a powerful tool in developing an assessment for learning culture (Black and Wiliam, 1998) in schools.  This culture of students and teachers sharing in the assessment process radically changes the power paradigm associated with assessment and serves to make students more autonomous and self regulating in their learning.

 

Estimating Passenger Flow and Occupancy on Board Public Transport Buses Through Mobile Participatory and Opportunistic Sensing

Student:              Stephen Brandon          
Supervisor:         Prof Siobhán Clarke
Course:                MSc in Computer Science (Networks and Distributed Systems)

Estimating passenger flow and occupancy on public transport buses usually involves dedicated hardware or conducting surveys, both of which can be expensive and time consuming. Bus occupancy data is useful for a variety of reasons. Bus operators can reduce running costs and environmental impact by understanding how to optimise the way they use their fleet. Passengers feel more comfortable and are more likely to use public transport when it is not congested.
Recent research has addressed various methods of collecting bus occupancy data. Much of this research has focused on hardware that needs to be installed on board the bus such as cameras and Wi-Fi Scanners. Crowdsensing applications have proved successful in other smart city projects including public transport applications such as estimating bus arrival time.

In this project we propose a novel solution to estimate passenger occupancy and flow on board buses. A mobile crowdsensing application capable of detecting other mobile devices through Bluetooth and Wi-Fi was created. Using a binary classification model we found that it is possible to differentiate if a bus is full or not based on the number of Wi-Fi users.

 

Expressive Information of Movement with Stylised Motion Cues                                                                                                                      

Student:              Yafei Qui
Supervisor:          Prof John Dingliana
Course:                MSc in Computer Science (Interactive Entertainment Technology)

Artists often use stylized motion cues to convey motion in static graphics such as paintings, comics and animations. Most of these techniques are commonly used for animations as well. Generally artists abstract the motion of an animation or a scene to visualize the movement. In real-time 3D computer graphics, these motion cues are no less essential and therefore the approaches are similar. One type of motion cue cannot express all the features of movement. Different motion cues are more effective at visualizing certain features.

This dissertation explores four commonly used methods to stylistically represent certain aspects of motion: motion blur, motion lines, stroboscopic effects and deformation. To implement those types of motion cues, we explored several 3D graphics techniques such as framebuffer objects, geometry shaders and cartoon-like squash and stretch. Four features of movement were mainly focused upon: velocity, acceleration, direction of movement and previous poses. We compared the expressiveness of each kind of motion cue for different movement features and investigated different ways of the combining them. Finally, we discuss the results and conclude with an outlook of the future development.

 


Game Design using Information-Centric Networking                                                                               

Student:              Chenghao Yu
Supervisor:         Prof Stefan Weber
Course:                MSc in Computer Science (Networks and Distributed Systems)


Gaze Direction in Interactive 3D Environment

Student:              Fan Li
Supervisor:         John Dingliana
Course:                MSc in Computer Science (Interactive Entertainment Technology)


The human visual system (HVS) has unique features for perceiving and processing visual information. By utilizing these characteristics we can enhance the difference between specific target elements and environment contexts when rendering a real-time photo-realistic 3D scene. Using various image processing techniques, the enhancement can be achieved by applying modulation to relatively inconspicuous regions in the scene so that the peripheral vision will detect the change and draw the viewer’s foveal vision to the modulated region.

However, in order to keep a viewer immersed within the virtual environment, it is better to implement the modulation in a subtle manner by using gaze directing cues which the viewer will not be consciously aware of. This is achieved by monitoring gaze position with an eye-tracker in an interactive application (we use the Eyetribe 2, a low-cost eye-tracker placed below the computer screen). Before the viewer’s foveal vision enters the modulated region, the modulation is automatically terminated.

This dissertation explores several different rendering techniques that might be useful for subtle gaze direction, analyzes the rendering results using established saliency metrics and seeks to combine the most useful techniques to subtly attract visual attention in real-time 3D scenes.


Interaction and Rendering of Virtual Humans in Augmented Reality                                                                   
Student:              Saloni Sharma
Supervisor:         John Dingliana
Course:               MSc in Computer Science (Interactive Entertainment Technology)


With the recent advancement in augmented reality techniques, there is an increas-ing demand to incorporate realistic behaviour of virtual objects in augmented reality.
In augmented reality, a virtual object is typically displayed on top of the real world scene even though a virtual object is blocked by a real one from the viewpoint of the
user. Virtual objects also exhibit ghost-like appearance compared to the surrounding real world objects. Despite continued research in augmented reality techniques, these
visibility and interaction problems have been handled separately in different research works. Moreover, there has been a lack of research work done for di erent mobile oper-
ating systems with respect to occlusion and interaction of virtual objects with physical objects. Therefore, the requirement arises to give a combined solution which could
incorporate these features and provide an application which could be compatible with different mobile operating systems.

To fulfil these requirements, the focus of this project is to deliver an application as a proof of concept. Prede ned position and size of proxy objects were given as
inputs into the framework. The implementation comprises a model which compares the depth values of real and virtual objects from camera and then prevents the display
of blocked portions of virtual objects. Moreover, it delivers a simulated environment for the virtual object to interact in a way to prevent itself from colliding with real
world objects.

Is it feasible to design a useful interface for COPD self-management?

Student:              Richard Eibrand
Supervisor:         Prof Gaye Stephens
Course:                MSc in Health Informatics

Chronic Obstructive Pulmonary Disease (COPD), is a degenerative disease of the lungs that is treatable but not curable. Physical activity such as walking can greatly assist management of the condition. This project aimed to identify if it was feasible to design a self-management interface using activity data and self-management principles for COPD patients. Results show that yes, it is feasible, and that such research warrants future work.

 

Is it feasible to have fine grained run time message monitoring of web service, with lesser performance impact?

Student:              Simperpreet Singh Anand
Supervisor:         Prof Stephen Barrett
Course:                MSc in Computer Science (Networks and Distributed Systems)

 

Modelling Environmental and Temporal Factors on Background Characters in Open World Games

Student:              Tony Cullen
Supervisor:         Prof Mads Haahr
Course:                MSc in Computer Science (Interactive Entertainment Technology)

The goal of this dissertation is to explore methods for increasing the believability of "background" characters in open world games. While the background characters do not influence the game's main storyline, they play a vital role in making the game world feel more believable to the player. These characters often follow scripted behaviours, carrying out the same tasks each day unless otherwise influenced by the player. Furthermore, the environmental and time-related cycles of the game world, such as day/night cycles and weather cycles, have no impact on the behaviours and actions of these characters.

This project proposes an emotional model which drives the behaviours of these background characters through both the tasks they carry out and the environmental and temporal factors of the game world. A State Manager is used to determine the next task for the non-player characters (NPCs) based on their current emotional state and the traits they possess. Traits, e.g. nocturnal or cold-blooded, are introduced to accommodate the different races found in open world games, such as humans, lizards, and orcs. Having the environmental and temporal factors influence the behaviours of these races differently adds a further element of believability to these characters. The model is tested in a game environment within the Unity game engine and shows potential, as the characters' actions are determined by their emotional state, as opposed to the scripted behaviours typically found in open world games. With further work, this model could have potential for use in open world games

 

Movement Tracking using Bluetooth Low Energy

Student:              Neil Savio Carvalho
Supervisor:         Prof Jonathan Dukes
Course:                MSc in Computer Science (Mobile and Ubiquitous Computing)

With the ever rising costs, non-uniformity in procedures and localised medical data in healthcare, eHealth and mHealth technologies strive to develop pervasive, low cost, uniform and ubiquitous health monitoring at the comfort of one's home. While the costs of rehabilitation can be about 60€ per session and about 10 sessions on an average with a physiotherapist , a motion capture system with therapeutic feedback would come at a one time small cost of about 50-60€ only. Motion sensing over Bluetooth Low Energy (MoBiLE) system aims to design, implement and evaluate a low cost, lightweight and energy efficient remote exercise monitoring system for physiotherapy. This monitoring system processes a real time data stream from motion sensors and then utilises this data to animate human motion using augmented reality. This system uses Bluetooth Low Energy (BLE) as low power communication protocol and implements a novel Power Efficient Data Adaptive (PEDA) algorithm to conserve battery life. This proposed solution allows the user to use the motion capture sensor with a coin cell battery which can last for hours.  Initial experiments comparing this algorithm with other raw data techniques show that this technique is able to conserve considerable power and produce comparable animation results.

 

Qualitative Analysis of Open-Source SDN Controllers

Student:              Andrei Bondkovskii
Supervisor:          Prof Stefan Weber
Course:                MSc in Computer Science (Networks and Distributed Systems)

The adaptation of software-defined networks (SDN) technologies promises significant benefits and cost reductions to operators of data networks. Telecommunication companies with expensive networks may become the biggest beneficiaries of SDN; however, in contrast to traditional routers, the development of SDN controllers is driven by open source projects with involvement of the industry. Two prevalent projects in SDN development are the OpenDaylight and the ONOS controller. These SDN controllers are advanced in their development - having gone through a number of releases - and have been described as being useful for a large number of use-cases. In this work, we compare and evaluate these controllers in our evaluation environment by configuring them for a representative use-case, port mirroring.

 

Rare Disease Challenge- Developing a Disease Specific Minimum Data Set in Hypermobility Type Ehlers Danlos Syndrome

Student:              Fiona Curran
Supervisor:         Prof Gaye Stephens
Course:                MSc in Health Informatics

Strategies for recognition and management of rare diseases have recently been developed at international and national level.  Information technology has a key role to play in the successful implementation of these strategies.  Developing disease specific rare disease essential datasets is however, inherently challenging. 
This research presents the methodology and the first iteration for development of a minimum dataset for a potentially disabling rare heritable connective tissue disorder - Ehlers Danlos Syndrome Hypermobility Type (EDS-HT).

Side Quest Generation using Interactive Storytelling for Open World Role Playing Games                        
Student:              Sarah Noonan
Supervisor:         Prof Mads Haahr
Course:                MSc in Computer Science (Interactive Entertainment Technology)

 

Smart Metering:  adopting CCN in IoT                                                                                                     

Student:              Haonian Li         
Supervisor:         Prof Stefan Weber
Course:                MSc in Computer Science (Networks and Distributed Systems)

 

Soosokan Over the Cloud

Student:              Tian Tian
Supervisor:         Prof Siobhán Clarke
Course:                MSc in Computer Science (Networks and Distributed Systems)

This dissertation analyses the benefits of using a cloud computing service by investigating cloud computing's architecture, technologies and service types for different business models and comparing cloud computing platforms offered by various companies. The dissertation also investigates business models and the status of cloud computing services, reviewing the differences between various cloud computing solutions.
This dissertation designs and develops a real mobile business application, setting up two kinds of servers for comparison: one that uses a traditional server and the other that deploys a server to the cloud. In the evaluation of the experiment, the dissertation mainly considers the service latency, reliability, deployment effort and cost saving.

 

Structural Health Monitoring of Wind Turbines using Unmanned Aerial Vehicles

Student:              Smita Shivaram
Supervisor:        Prof Ciaran McGoldrick
Course:                MSc in Computer Science (Mobile and Ubiquitous Computing)

Structural Health Monitoring (SHM) is a key technique in ensuring the health and safety of civil  and  mechanical  structures. The most commonly used  monitoring  methods  make use of  a combination of techniques such as vibration-based methods, acoustic emission detection,  thermal  imaging,  and  ultrasonic  reflection in order to detect  deformation or failure.  The use of active image processing concepts in Structural  Health Monitoring of static and moving civil structures is largely unexplored and represents an interesting area of research.

The  aim of this dissertation is to evaluate the use of visual imaging from  unmanned air vehicles to autonomously monitor the Structural Health of Wind Turbine blades in order to create a real-time model which can be used to assess its structural health.

Preliminary experiments  and  simulations have provided promising results, and the use of capable drones in combination with specialist algorithmics can be used to identify small-scale deformities. Current  research  targets  challenges such as3d reconstruction of scenes and objects from  unmanned aerial vehicles imagery, proactive non-destructive monitoring of moving mechanical structures using visual imaging etc.

 


Towards Autonomic Uplift of Information

Student:             Sarah Alzahrani
Supervisor:       Prof Declan O'Sullivan
Course:             MSc in Computer Science (Networks and Distributed Systems)

In the past few years, there has been a considerable amount of growth in the semantic web adoption. A large amount of the information available on the web has been modelled into various data models, and with each model comes a different form of standardization, format and rules to govern the data. Thus, people are increasingly uplifting data into Resource Description Framework (RDF) which is a standard model for data interchange on the Web. However, because data schemas and ontologies that are used to describe data in RDF may change frequently over the time, the need for an approach to represent the mapping statements in an analyzable form has been arise.

Most of current ongoing research in this area attempting to demonstrate generic solutions for mapping or translating data that exist in different format from the standard data model. However, not much attention has been given to the representation of the mapping statements in the previous solutions. In this research, focus is placed on achieving the mapping process in an autonomic fashion. In other words, representing a query-based mapping (XQuery queries) using RDF (Resource Description Framework), which appears (after significant investigation) to be a novel idea in the area. Therefore, this representation will enable the manipulation and analysis of mapping statements.


 

 

Volume Ray Casting in Virtual Reality                                                                                                                        
Student:              Huanxiang Wang
Supervisor:         Prof John Dingliana
Course:                MSc in Computer Science (Interactive Entertainment Technology)

 

 

Demonstrations and Installations (Large and Small Confrence Rooms, O'Reilly Institute)

Course: MSc in Interactive Digital Media

Into Kilmainham: A VR Story – a Virtual Reality Game

            
Students:          Philip Hickey, Neil O'Carroll, ALina Ekye, Conor Snyed, James O'Sullivan
Supervisor :      Prof Mads Haahr

A VR Storyis a 3D virtual reality game for Google Cardboard. The player finds themselves drawn into Kilmainham Gaol one night, and must find a way to escape. Echoes of the past still linger in the Gaol, and the player encounters the ghosts of various inmates and staff from the prison’s history. The player must listen to their stories and insights in order to solve puzzles and advance through the Gaol towards freedom.

 

A Mad Tea Party - an Installation

Students:         Gemma Cagney, Niamh May Goodwin, Cian O’Gralaigh, Gearóid Mac Giolla Mhuire and Richard Varney
Supervisor :     Susan Gill

What, if anything, do you think Einstein might have learnt from Alice in Wonderland? What if you were told that Carroll’s story of Alice is 50 years older than Einstein’s theory of General Relativity? Both works are celebrating important anniversaries this year, and you are invited to party to celebrate, a tea party that is and a mad one at that!
Journey down the rabbit hole with us, to a world where your actions affect the world around you, and nothing is as it seems

Avian - an Installation

Students:   Julia O’Flynn O’Brien, Sophie Murphy, Alessa Gross, George Morohan, Jack O’Dea
Supervisor:  David Kaneswarren    

Avian, defined by the English Oxford Dictionary: a bird, relating to birds. It is certain that computer is not a bird, but it is also certain has the capability to relate to birds, if programmed to do so.
The ‘birds eye view’ is something we perceive second degree, on a map or a screen, static and motionless. AVIAN is an immersive interactive, visual and physical experience, upon which physical birdlike inputs provide visual birdlike outputs. Finally the day, the day whereby if you flap, you can see flat; skyscrapers and grass fields look like mere geometric constituents to the jigsaw of the universe. Indeed, the panoramic visualisations in front of you make it impossible to see otherwise.

Hello World? - an Installation

Students:  Robin Challis, Therese Walsh, Mark Dunleavy, Evan Burbidge, Joe Deegan
Supervisor:  Kris McGlinn

Feel, speak, connect, think.  
Communication is at the very core of humanity. Technology has become a tool to transmit our feelings and ideas but how will we communicate in the future? Will we need language? Perhaps we will communicate directly: mind-to-mind. And what of the finer details of human interaction: inner feelings and emotions? Will the self itself change as we merge with the network?
Imagine the future of communication in a digital era. Will it work? Is it human? We beg the question. You try it out.

 

Postgraduate programmes on offer at the School of Computer Science and Statistics

Full-time over 1 year

MSc in Computer Science

  • Interactive Entertainment Technology
  • Mobile and Ubiquitous Computing
  • Networks and Distributed Systems

MSc in Interactive Digital Media

Part-time over 2 years

MSc in Health Informatics (evening/weekend)

MSc in Management of Information Systems (evenings)

MSc in Technology and Learning (evening/weekend)

Postgraduate Certificate in Statistics (evenings)

Further information at: www.scss.tcd.ie/postgraduate/

Structured PhD programme

The School has a large and vibrant group of over 200 researchers, of which more than 150 are PhD students. TheSchool collaborates with leading industry players in the disciplines of Computer Science and Statistics.

PhD students have a wide and varied choice of research areas in one of our main disciplines: computer systems, information systems, intelligent systems, software systems and statistics.

Further information at: www.scss.tcd.ie/postgraduate/structuredphd/

Undergraduate study

Day courses

BA (Mod)/MCS in Computer Science

BA (Mod) in Computer Science and Business

BA (Mod) in Computer Science and Language

BA (Mod) in Management Science and Information Systems Studies

BA, BAI, MAI in Engineering/Electronic and Computer Engineering

Evening courses

Diploma in Information Systems; a two-year course intended for IT professionals and those who wish to work in Information Technology

Degree in Information Systems; a two-year programme following on from the Diploma in Information Systems leading to a BSc (Hons)

Further information at: www.scss.tcd.ie/undergraduate/

Trinity College Dublin is the top-ranked university in Ireland and is ranked in the top 100 universities of the world within the disciplines of Computer Science and Information Systems.

 

Attendance at the event included: Murex, Workday, Fullbright Commission, PageFair, Bluebear, Meez, Ding, Surewash, Ericsson, Lighthouse BCS, Citi, Fieldaware, go2web Ltd, Intercom, Accenture, Escher Group, Amazon, Infosys Software, Irish Software Innovation Network, Salesforce, DIGIT Game Studios, HSE, Daft.ie, Synopsys, Synchronoss Technologies, First Derivatives, Intercom, Havok, Swrve, SIG, Artomatix, Intercom, Cisco, Corvic, Fungus Ltd, Brandcreate, SAP.