Research

 A  
Lecture  2:  Elements  of  a  
research  project  

Lejla Batina and Peter Schwabe

Slides credit: James Mc Kinna

Outline  
•  Admin and status
•  Your research project
•  Proposal writing
•  Presenting your work

•  Disclaimer: This is not a course on

Academic (English) writing!
Part 1: Admin and guidelines
Organisa:on  
•  Officially, we should meet on Wednesdays
•  35+(!) students in BB (17-18 groups)
•  => progress meetings from 9:00 and in the afternoon (Wed)
•  Blackboard will be used for
– assignments, reading material
– grading
•  So: make sure you are enrolled, and that you are
comfortable with Blackboard (!)
•  When sending e-mail please put Research A in the
subject line (send it to both teachers)
•  Remainder: next Tuesday is your first deadline
–  e-mail with 2 names, title of your topic, supervisor
Next  deliverable  deadline:  Proposal  submission  
•  Friday, Sept. 27, 15:00!
•  Send e-mail to Peter and Lejla
–  Attach your slides as “Names_proposal.pdf”
•  Presentations on Oct. 2:
–  10 min + 5 min Q&A
–  Both students should speak
–  Timing!
Goals  of  the  Course  
•  Ultimate goal: to produce a research paper
•  ... and the associated proposal and slide presentations
•  To understand what makes a good, or bad, paper
•  Around 2 broad themes:
•  Big Data
•  Internet of Things
•  Work in groups of two; individually – only by agreement
with us (very exceptional)
•  Main focus is on methods (literature search, writing skills,
time management, …) rather than original research
contributions
Standards  
•  Language is English
•  File formats: .pdf for papers and slides
•  Which tools you use to produce them is a matter
of choice
•  Later in Research B, you will be expected to use
LaTeX, so consider learning it now
Obliga:ons  
•  Lectures are optional (slides will appear online)
•  Attendance at presentations is mandatory
•  Additionally: periodic (every 2 weeks, typically) progress
reviews with us and supervisors, also mandatory
•  A few assignments and home works will be posted via
Blackboard
Professional  scien:fic  research  
•  Identifying interesting problems, analyzing and
solving them (if only partially)
•  By means of well-formulated research question
•  By using respected, well known methods,
•  So that the observations and results are
reliable, repeatable and reliably repeatable
From  ques:ons  to  answers  
•  Start with a (partial) question
•  Search for and find material
•  Work on the material to make it suitable for your
research
•  Analyze this suitable material
•  Interpret the results
•  Formulate a partial answer; then iterate
Criteria  for  suitable  ques:ons?  
•  Scope or knowledge area
•  What is the research about?
•  What is the point of view?
•  Relevance
–  Why is it worth doing this research?
–  There must be a reason!
–  Is it a theoretical relevance, or scientific, or practical, or
for society?
Other  criteria  
•  Precision
–  The question needs to be formulated precisely
–  What exactly do you want to learn from the
research?
–  What kind of result should it derive?
–  Describe what you want: A literature study? A
scientific paper? A prototype? Some statistical
conclusions? …
•  Methodology
Methodology  
•  Logical reasoning: do the conclusions follow
from your (experimental) hypotheses?
•  Do you know what you are doing?
•  Can you trust what you are doing?
•  Easy to make mistakes:
–  wrong models, wrong hypotheses, wrong
observations
–  systematic vs. incidental errors
–  draw conclusions outside the range of reasonable
prediction
Literature  
•  Official
–  Strong reviewing process (reliability)
–  Books
–  Papers in scientific journals
–  Papers in proceedings of scientific conferences
•  Unofficial, grey literature
–  Unreliable: little or no refereeing process
–  Unchecked — Wikipedia articles?
–  “Wisdom” of crowds — Google rankings?
–  Journalism
–  Bachelor’s and master’s theses
–  Workshop publications
Part 2: Research Project
FAQs
•  What does it mean to do research?
•  How will I ever get it all done?
•  How to take notes?
•  What is a bibliography?
•  What are background sources?
•  How do I focus my research?
Why  research?  
•  Work in many academic and professional
disciplines depends on it
•  To research = to explore problem systematically
•  A researcher: a careful, critical, systematic
thinker who goes beyond memorizing facts
•  Buying a TV requires also research
Narrowing  and  focusing  the  topic  
•  Search engines again
•  Talk to other people
•  Find out about research questions
•  Develop a working hypothesis: what do you
expect to find out
Steps  to  undergo  for  wri7ng  a  paper  ≈1  week  per  step  
1.  Select a topic, articulate starting questions, begin
background research, schedule a time frame
2.  Build a working bibliography
3.  Read and evaluate sources, take notes on relevant
sources, comment on the importance of sources
4.  Start conducting research, identify gaps
5.  Begin preliminary writing
6. Write the main statement that will guide the rest,
sketch an outline of the paper
7. Write a draft of the paper including bibliography
8. Revise and edit, ask for feedback
9. Print and proofread the final copy
A  working  bibliography  
•  A listing of books and articles on a particular subject
•  When submitting a research shows which sources
you consulted
•  Start creating one in the beginning, which will grow
as your research progress
•  Include accurately all the info
Gather  background  informa7on  
•  Compile a list a key words
•  Background reading in general reference books
•  Interviewing experts
•  Search the web
The  research  paper  
•  A form of written communication
•  Should present information and ideas clearly and
effectively
Steps  for  wri7ng  
•  Determine your purpose in writing the paper (to
describe, to explain, to argue, to persuade the
reader to think)
•  Consider the type of audience
•  Develop the main statement expressing the
central idea
•  Gather your ideas and info in a preliminary list
Steps for writing (cont’d)
•  Arrange materials in an order appropriate to the
aims of the paper
•  Make a detailed outline
•  Write a preliminary draft
•  Read it critically and try to improve
–  Revising, rearranging, adding/eliminating words/
phrases/sentences
•  Proofread the final draft
Taking notes

•  After verifying the publication info for a

source, the next step is to read and evaluate
•  You should not assume something truthful
or trustworthy just because it is in print
•  When you find some reliable material you
should takes notes on it
•  Essential to research but probably no 2
researchers use the same method
•  Careful note-taking helps you avoid the
problem of plagiarism
3 methods for note-taking
•  Summary
•  Paraphrase
•  Quotation (use “” here)

•  Each one more detailed than previous

Plagiarism
•  Using someone else's ideas or phrasing and
representing those ideas or phrasing as our own,
either on purpose or through carelessness, is a
serious offense known as plagiarism
•  From the Latin word “plagiarius” – (“kidnapper”)
Some  literature  
•  J. Gibaldi and W. S. Achert: The MLA Handbook
for Writers of Research Papers, 3rd edition.
•  A. Hult and T. N. Huckin: The New Century
Handbook, 2nd edition.
•  Robert A. Day and Barbara Gastel: How to
Write and Publish a Scientific Paper
Part 3: Your Proposal
Research  Proposal  
•  Checklist
–  Abstract
–  The problem
–  Motivation
–  Theoretical scope
–  Strategy
–  Time schedule
–  References
•  Note that these items do not have to match
the sections
Abstract  
•  The most read part of the proposal
•  Maybe even the most important part
•  Sell your work: point out why your research
is important
•  Sell your work: summarize what you will be
doing
•  Your abstract should not be a summary of
the proposal but of the research
•  Should be understandable even for non-
specialist
The  problem  
•  State the problem
•  Research question(s)
•  Hypothesis?
•  Possible sub-questions
Mo7va7on  

•  Why is it important that someone does this

research
•  Why do you want to do this research
–  Maybe not applicable in all cases
Theore7cal  Scope  
•  Theoretical background
•  Related works
•  Definitions
•  Assumptions
•  Scope: What will be covered by the
research and what not
Strategy  
•  How things are (going to be) done
•  Collecting information
•  Analyzing information
•  Which activities will be done . . .
Time  schedule;  planning  
•  Which activities will be done when
•  Which deliverables will be delivered when
•  Deliverables should deal with logical sections
•  Stick to the plan!
•  . . . or explain why you changed it!
References  
•  List of (probably) relevant references
•  Scientific literature
•  Wikipedia (and many other websites) do not
count as scientific!
Pro-­‐forma  proposal
•  who you are; what is your title;
•  what is the (well-defined) domain of study?
•  what is (are) the algorithm(s) you will study?
•  why is it interesting? and to whom?
•  what question do you hope to answer
•  how do you plan to do it?

•  All this should fit into 5-6 pages

Recap  
•  Statement of a problem
•  Literature review
•  Conceptual framework
•  Research questions
•  Methodology
•  Scope of work (what exactly will be done)
•  Planning
DOs  and  DONT’s  
•  Write clearly
•  Make it easy (and pleasure) to read
•  Abstract and Introduction should be written for a
non-expert
•  Give it to a colleague to read
•  … and be open for (constructive) feedback
Part 4: Presenting your work
 Less  is  more  
•  10 minutes is very little time!
•  1min/slide is hard to keep to
•  don’t read from your slides
•  avoid fancy fonts
•  avoid small writing
Less  is  more  
•  A slide full of text: read or listen
•  Not more than x lines (x =7?)
•  Use pictures (whenever possible -> ideally on
every slide)
•  Text on slides is not for you but for the audience
–  If necessary use notes that appear only on your
screen
•  Your talk is an advertisement for your work, not
a replacement!
•  The take-home message of the talk: Read the
paper!
Pro-­‐forma  talk  
•  who you are and where do you come from
•  what is the (well-defined) domain of study?
•  what is (are) the method(s) and application(s)
you will study?
•  why is it interesting? and to whom?
•  what question do you hope to answer
•  how do you plan to do it?
Pro-­‐forma  delivery  
•  the IBM way:
–  say what you are going to say
–  say it
–  say what you have said
•  too hard in 10 minutes?
•  don’t be tempted to talk quickly; just like too much
text on the slide
Preparing  a  talk:  Recap  
•  Motivation: Why?
•  Difference from other related work: What? How?
•  Evidence that your work is valuable
•  Take-home measage
Giving  a  talk:  Recap  
•  Practice, practice, practice
•  Introduce yourself
•  Concentrate
•  Speaking: don’t rush, use a pause
•  Timing
•  Closing
Dont’s  
•  Laser pointers (only if you use them properly)
•  Looking constantly into your screen (or the slides
showed to your adudience)
–  You should know your slides!
•  “I’m running out of time, so I’ll just…”
–  Leaves a sloppy impression
Homework  
•  Listen the following lecture:
hMp://www.inf.ethz.ch/personal/markusp/teaching/guides/guide-­‐
presenta:ons.pdf  
•  Prepare a 3-minute presentation for next
lecture (September 25), topic – free choice,
no text on slides, just pictures), no notes