MSc Projects Outline
MSc in Computer Science 2015-6
Oliver Ray
04/12/2015
Contents
•
•
•
•
•
•
•
•
•
Project Timeline
Research Skills
Project Interviews
Project System
Project Types (I, II, III)
Factors when choosing a Project
Other sources of ideas
Student Case Study
Staff Case Study
Project Timeline
Project
proposals
…
dec
Project
interviews
jan
Project specification
(Research Skills)
feb
mar
project execution
(Individual Project)
apr
may
jun
jul
aug
Project
writeup
sep
SV
SAP
SRW
TB 2
EV
TB 2
JAP
CV
TB 1
07 14 21 28 04 11 18 25 01 08 15 22 29 07 14 21 28 04 11 18 25 02 09 16 23 30 06 13 20 27 04 11 18 25 01 08 15 22 29 05 12
workplan
thesis
research review
literature list
Poster
demo
synopsis
Project allocation
https://www.cs.bris.ac.uk/project/MSc/
progression
A note on Progression
• Marks for taught units are considered by an exam board in
June and the results will be announced at the end of June
• If you fail any units you won’t normally be allowed to start
your project until you have passed any resits in September
– this is official university policy!
• In this case, you will give your poster/demo in late December
and submit your thesis in mid January
– This will delay your graduation!
https://www.bris.ac.uk/engineering/currentstudents/gradschool/summerresults.html
Research Skills
https://www.cs.bris.ac.uk/Teaching/Resources/COMSM2202/
Project Interviews
• You should start thinking about your project in TB1 as you
must find a supervisor by wk14 at the latest
• This is a competitive interview process so you will need to go
prepared and be able to sell yourself
• All students need to define their own project using the skills
they will learn on Research Skills unit
• But staff will help you to get started by putting some initial
draft proposals into our Project System
• Students really need to be proactive in all aspects of their
projects (self-motivation is the key to success)
• Start approaching staff as soon as possible; don’t get left
behind
https://www.cs.bris.ac.uk/project/MSc/interviews.html
Project System (MSc)
https://wwwa.fen.bris.ac.uk/COMSM2202/projects/index.jsp
Project Types
• Type I: Implementation – build a novel piece of software /
hardware using concepts from Computer Science. The added
value will mainly come from the design and evaluation.
• Type II: Investigation - work on an open research problem to
find new results or new methods. The added value will come
from the analysis of requirements, design and evaluation.
• Type III: Theory - developing models / theories to explain an
idea in Computer Science. The added value will come from
the mathematical analysis above that already in the literature.
• Any combinations of the above!
https://www.cs.bris.ac.uk/project/MSc/MScProjectGuidelines.pdf
Factors when choosing a Project
•
•
•
•
•
CV/Portfolio - What do you want to do in the future?
Passion – Which topics do you enjoy the most?
Skills – What are your strengths and weaknesses?
Personality – Which members of staff to you get on with?
Ownership – You must take responsibility for your project!
– Regularly check the projects list and do some background reading
– Contact the supervisors: email them; go to their office, talk through the
project; demonstrate your understanding; be willing to contribute your
thoughts on the directions the project could take
– Remember most projects can be adapted to suit your interests and
skills; most projects will evolve over time anyway
– Remember other projects can be suggested; you are encouraged to
propose your own ideas, but you must find a supervisor to champion
them
Other sources of ideas: PhD
https://www.cs.bris.ac.uk/admissions/phd/projects/phd-projects.jsp
Other sources of ideas: UG
https://www.cs.bris.ac.uk/project/ideas/staff.html
Other sources of ideas: Web
http://www.cs.bris.ac.uk/~oray/
Other sources of ideas: Papers
https://scholar.google.co.uk/citations?user=3vwgzlUAAAAJ&hl=en
Other sources of ideas: Groups
http://www.bris.ac.uk/engineering/departments/computerscience/research/
The Bottom Line
• The task of specifying a good MSc project is a continuous
process of refinement that requires a significant amount of
effort and background knowledge
• It is simply not practical for staff to formalise all of the ideas
or interests they have, so there is no substitute for doing your
own research and going to talk to staff in person
• The best projects will most likely be done by students who
are already thinking about their projects and talking to staff
(i.e. taking responsibility without being told what to do)
• SO GET OUT THERE!
Student Case Study: Nov 2015
Hi Oliver,
I'm looking forward to your lecture. I'm just emailing to ask for some initial
advice about choosing a realistic project: from either an area in which I have
more experience (Law-related technology), or one for which I have less
experience but far more enthusiasm (dance-related technology).
After you mentioned that I would be taking Research Skills this year, I
decided to take a look at the unit pages and read some of the handbook. I
was really excited to learn of your interests in social dancing and in legal
reasoning.
I hope this is not too early, pre-emptive or presumptuous, but I'm hoping you
could advise me about choosing a realistic project from one of those areas.
I've had and researched ideas for both (below), but it would be really helpful
to know which direction would be most advisable to explore more. My
current ideas are as below. If you do have the time, and it's an appropriate
stage, then many thanks in advance for your advice.
Student Case Study: Idea #1
Law: More efficient machine classification of legal documents - trying to
find quicker ways (both computationally, and pragmatically in a business
sense) of accurately identifying what long legal documents are about.
(Should be comfortably within my knowledge and ability.)
Student Case Study: Idea #2
Dancing: Learning exercises to teach social dancers how to hear and
respond to the music's beat more accurately. This would involve
researching the biological/knowledge-gap some people have problems
hearing and moving to the beat, and designing some software exercises to
help train them to do this better. (Should be comfortably within my
knowledge and ability.)
Student Case Study: Idea #3
Dancing: "Silent" tap shoes: Using wireless, pressure sensing and MIDI
technology to let tap dancers dance with minimal noise, sending synthesised
sounds from their feet to a headset, rather than out as a nuisance to the
world. May also involve using physical sound-proofing. (I would need to learn
a lot to implement this in hardware, but if the circuitry is kept relatively
simple, execution should still be possible for me).
Student Case Study: Idea #4
Dancing: Typing using dance moves: designing a workable 'alphabet'
from a well-defined set of dance moves, and using motion detection, gesture
recognition, and keyboard input technologies to implement it. (This would
require the most new learning, and would be the most risky in terms of my
ability to execute it. But it's the one I would most love to do! Wondering
whether even doing one part of it - the alphabet design, using some
computer science technique to find the most commonly combined and least
strenuous dance motions to include - would be an appropriate project?)
Staff Case Study: Nov 2015
1) Modelling of collective decision making in social insects
https://sites.google.com/site/computationalantlab/home
2) Energy aware Software Compilation
http://seis.bris.ac.uk/~cb0094
3) Sports Analytics
4) Relational Data Mining / Nonmonotonic Reasoning
5) …
THE END