CMSC 601: Basic Research Skills

Spring 2010
Instructor: Charles Nicholas

Subject to revision through the semester. The sequence of topics and papers should be pretty accurate.
Date Topic(s)

Assignments Made

1/27 Overview,
Kung's talk,

M. E. J. Newman, "The structure of scientific collaboration networks" (pdf)
to be discussed in class

Grading policies
H.T. Kung, "Useful Things to Know About Ph. D. Thesis Research",


Finish Kung.
Introduction to LaTeX.

2/3 Reading and summarizing of papers.
Reviewing papers.

Paper summary guidelines
Example: Summary of Newman paper
Paraphrasing guidelines

ICML-03 reviewing form

2/8 SNOW DAY    




The reviewing process. Discussion of Parberry Classification/Analysis.In-depth paper analysis and critique.
  • Dijkstra's Three Golden Rules for Scientific Research

    UPDATED 2/15/10
    Assignment due Monday 2/22/10: prepare summaries of any two of {Smith or Parberry or Palsberg or Gopen} papers (1-2 pages each)

    Zobel, chapter 10 (pdf), to be discussed in the future


    The Collection of Computer Science Bibliographies
    Justin Zobel's writing page
    Bob Grant, "Right your Writing"(html)
    Gopen and Swan, "The Science of Scientific Writing" (pdf)

    Annotated bibliography guidelines(ps, pdf) LaTeX source, and BibTeX file
    Oren Patashnik, "BibTeXing"
    Plain (sorted) and unsorted annotated bibliography BibTeX style files 
    Levin and Rendell, How to Write a Good Systems Paper Local copies(ps,pdf)


    Writing I: Organization, bibliographies.  BibTeX.
    In-depth survey analysis and critique.

  • discuss Zobel Ch. 10
  • Peters Ch. 1,2
  • Purdue Online Writing Lab's APA citation guidelines
  • Zobel Ch. 1 and pp. 20-25
  • Chinneck, How to Organize Your Thesis
    Google Scholar
    Sample topics
    Topic guidelines
    Literature survey tips
    Editing symbols


    Writing II: Style, common errors.

  • Zobel Ch. 2-3
  • Peters Ch. 18

    Assignment: Write a paragraph, or two at the most, describing the area in which you want to do your research project. Be as specific as you can. List at least three papers that you'll look at. Wikipedia doesn't count. Of those papers, indicate which may be seminal, and why.

    I'm requiring that you build your bibliography using BibTex, and the paragraph in LaTeX. So you'll get a chance to practice that!
  • Slides

    example: searching the Patent Database



    How to give a good presentation.

    Generic dissertation outline.

  • Peters Ch. 20
  • Zobel Ch. 11
  • Paper presentation feedback form
    3/1 Research I: Finding a topic and advisor.
  • Peters Ch. 5, 6, 16
  • Slides

    3/3 Research II: Research topics cont.
  • Assignment due March 10: Expand your bibliography file to include at least ten papers to which you might want to make reference in your own research paper. Send it as an email attachment.
  • Outline of a (draft) thesis proposal

    • introduction - topic paragraph
    • survey of literature - what are the seminal papers?
    • summary, summary, summary
    • more recent work - papers which cite seminal papers
    • Foobar et al. used a small data set, but we want to try on 100TB web corpus
    • Dewey, Cheatham and Howe advocated use of FASB 33, but we want to use FASB 444
    • how would we do this, if we had the resources?
    • what evidence is available to substantiate your claim(s)?

    More on research topics


  •  Zobel Ch. 4
  • Mark Hill, Oral Presentation Advice
    Giving a Talk
    Winston, Some Lecturing Heuristics

    Examples of good talks can be found at TED
    Tim Berners-Lee has a talk on web stuff

    Writing III: Tools.

    More on public speaking

  • Zobel Ch. 5, 6, 7
  • Assignment: Read desJardins' paper, to be discussed on Monday 3/22
  • Literature survey review form

    We looked at "Life After Death by Powerpoint" by Don McMillan, available at Youtube
    desJardins, How to Succeed in Graduate School (ps ,pdf )

    3/15-3/19 SPRING BREAK   During the break, I came across this blog post on James McLurkin, and the importance of stretch goals. How do you get to the bleeding edge where the real advances are made?
    3/22 Criteria for listening to a CS talk
    Assignment: have a detailed outline of your paper ready by Wednesday 3/31.

    Applying to Ph.D. Programs in Computer Science by Mor Harchol-Balter

    Outside reader agreement

    3/24 Research III: Empirical methodology.

    Update: What do I mean by detailed outline? As described above on March 3, the document is to have the following structure. For much of this, you'll have a good deal written, but other parts may be no more than bullet points or sentence fragements. But it should look something like this, expanded to maybe 3-5 pages. The final paper will be more like 15 pages. Real proposals are sometimes twice that or more.

    • Problem Statement - you've already written most of this
    • Survey of Related work - from the 10+ papers you've read
    • What could still be done? What would be new? What's your secret weapon?
    • Why is your idea plausible? What evidence do you have, or could hope to get? You can describe evidence that you don't yet have or even know you can get
    • What would need to be done to get from here to there?

    Introduction to statistics for CS - normal distribution, t-test, chi-square, at a high level! Resources include stattrek
    But there are many other tutorials and examples on the web.

    t test spreadsheet

    You may want to become familiar with the R Project, which is a highly regarded, free, extensible stats package.

    Using R to fit a curve to a scatterplot

    multinomial spreadsheet



    Career Development - Time Management

    Randy Pausch's talk (pdf) The video is available at many sites, including the 2007 version.

    Slides on experiment design
    Rob Holte's slides on experimental methodologies
    Research statement guidelines
    3/31 Career Development - Patterson Patterson, How to Have a Bad Career in Research/Academia" Slides on proposals
    4/5 More on Giving Talks Peyton-Jones et al., How to Give a Good Research Talk Advice to a Beginning Graduate Student by Manuel Blum
    CRA-W Career Mentoring Workshops Booklet (includes Berman's "Building a Research Career")
    4/7 Research IV: Proposals, grant writing

    Research portfolio guidelines

    4/12 More on presentations
    Set presentation schedule


    Retirement Advice

    Retirement spreadsheet (xls)

    4/14 Time Management I recommend Peter Drucker's 1999 paper, Managing Oneself (pdf) also available through the research port

    Advice to a Young Scientist, Dijksta (pdf)
    Professional societies: ACM, IEEE, Usenix, CRA, AAAS, AAAI,
    NSBE, Systers

    Science's Next Wave



    Getting Grants

    white papers, brainstorming, talking to program managers, review panels, earmarks

    the role of funding in academic rank

    More on personal finance from Get Rich Slowly

    Mind mapping, and software to help do it

    Web sites for NSF, DARPA, NIH, MIPS, ONR, AFOSR, IARPA

    4/21 Discuss Drucker  
    • Poster
      visual appeal
      legible from where you're seated
      appropriate use of fonts
      appropriate use of colors
      right mix of text and graphics
      grammar and spelling
    • Technical
      at the right level of detail
      right level of formality
      if not, why not?
      central idea - was it presented?
      appropriate presentation of related work?
      future work?
    • Presentation skills
      eye contact
      talking too quickly?
      appropriate use of humor
      talking loud enough
    • Suggestions for speaker:
    4/26 Student poster presentations
    Ballantyne: UAVs (pdf)
    Dabke: AI for Java (pdf)
    Guseman: transfer in machine learning (pdf)
    Hamilton: team formation (pdf)
    Kang: belief update (pdf)

    Some examples of posters:
    McNamee, Nicholas and Mayfield (pdf)
    Wilson and Nicholas (pdf)
    Finin et al (pdf)
    Martineau et al (pdf)

    4/28 Student poster presentations
    Dharurkar: context aware mobile computing (pdf)
    Kasinathan: effectiveness of mobile ads (pdf)
    Lewis: natural language generation
    Sleeman: semantic web data fusion (pdf)
    Gupte: cloud computing architecture (pdf)
    5/3 Student poster presentations
    Qi Ran: node deployment for sensor nets (pdf)
    Ren Zhong: routing in sensor nets (pdf)
    Thompson: optimizing container loading (pdf)
    White: detecting stepping stones using NN (pdf)
    Raabe: musical social networks (pdf)

    Student project presentations
    Graphics .

    Baumel: real-time irridesence shader (pdf)
    Dighade: particles for scientific viz (pdf)
    Mann: spatial input filtering (pdf)
    Mokashi: image analysis for cells (pdf)
    Rockwell: NPR stipple rendering (pdf)
    Kewalramani: sentiment analysis on Twitter (pdf)
    5/10 Student project presentations
    Gui: named-entity disambiguation (pdf)
    Iyengar: context disambiguation in social media (pdf)
    Koduvayur: similarity for semantic web (pdf)
    Rafi: multiobjective optimization using BBO (pdf)
    Rosebrock: RF in thematic space (pdf)
    Trinkle: metadata clustering (pdf)
    5/12 Student project presentations
    Kavita: word prediction (pdf)
    Wilkinson: expanding lexicon using cognates (pdf)
    Lewis: improved debugging system (pdf)
    Leschke: cyberforensics tools (pdf)
    Mahale: security and access control using semantics (pdf)
    Jagtap: data security using negative databases (pdf)

    Kavita has a demo

    Last day of class: May 13

    5/17 Scheduled final exam 6pm Papers due!  

    Stuff I removed from the schedule but don't want to loose:
    Cohen, Schapire, and Singer, Learning to Order Things
    Kamishima and Akaho, Learning from Order Examples

    Kajiya, How to Get Your SIGGARPH Paper Rejected
    Discussion questions
    Doyle and Thomas, Background to Qualitative Decision Theory

    Transcript of Hamming, "You and Your Research" (html,pdf)

    This semester I didn't have students practice presenting papers. It would have been good, but
    I didn't see it on the schedule in time. Also, the class is a little too big to do this and squeeze in
    everything else. But knowing how to present papers is very important in research groups, and
    if I had it to do over I might do it differently.