How to prepare your class presentation for CPSC 545:

Here are a few suggestions on how to prepare your presentation.

  • Two weeks ahead of time: start preparing your presentation, i.e. read your paper and directly related papers, compile a list of your questions, clarify these questions as far as possible by reading related papers, books, online resources and asking your colleagues. Prepare a digital draft version of the talk. This version should address the following key questions:
    • What is the main goal of the paper and what are the main claims made by the authors?
    • How do the authors support each of their claims?
    • Which related research has been previously published by the same or other authors? What is the novely value of the research presented in this paper? For this you may need to do a small literature search on your own.
    • Which methods or algorithms do the authors employ or propose to support their claims?
    • How do these work? You need to understand and be able to present the details of the these methods.
    • Do the algorithms and methods do what the authors claim? What is their time/memory requirement? What are other key features? Be sure you mention the input and output to each method.
    • What is the overall logical flow of the paper? Does this make sense or are there logical (or other) flaws?
    • (for some papers) Which data sets do the authors choose to support their claims?
    • (for some papers) How does the analysis pipeline work?
    • What are the main conclusions that the authors draw?
    • Are all conclusions well supported? Why? Why not? Argue in a watertight way and, if appriate, propose and explain potential solutions.
    • Which ideas do the authors propose for future follow-up research?
    • What are your own ideas for future follow-up research? For this, you may need to read related papers by other authors and consider the wider context of that research area.

  • One week ahead of time: clarify any remaining questions you may have by emailing me or seeing me in my office hour.
    • Decide on all figures and tables to use in your presentation. Make a wise selection of those in the paper and contemplate using or making additional ones. In any case, carefully acknowledge the source of each figure/table. Figures are often very useful for explaining tricky concepts.
    • Check whether all terms in all of your formula are defined before they are being used. Double-check that you can explain what they mean.
    • Check whether all new technical terms, i.e. "intron" are clearly introduced (or have been introduced in an earlier presentation). Again, double-check that you know exactly what they mean.
    • Check each of your slides to see if the text or any other aspect could be further improved. Avoid using full sentences, but rather use bullet point list of concise text.
    • Get the feedback of one or two colleagues from the course on your draft presentation. Can they understand your presentation only based on your slides?
    • Check the overall logical flow of your presentation. Do you define all new terms/words before using them? Does the overall logical of your slides make sense?
    • Check that you can give your presentation in the allocated time. Do one or two trial runs. This will also help you decide on want to actually say. Based on this, you may need to revise some slides.
    • Check that you can explain all details of the presented method and algorithm. Make sure the key features are clearly explained on your slides and be prepared to explain the omitted details on a white-board if being asked in class (by me or a colleague).
    • Add your own thoughts on what could be further improved regarding this analysis and potential future analyses.
    • Discuss the material you are presenting in the wider context and the context of earlier presentations, if appropriate.

  • Two days ahead of time: make sure your laptop is ready for the presentation and you have a back-up copy of your file with you on a USB-stick (just in case)
    • it is your responsibility to check that everything works fine with the data projector in class
    • if you don't have a laptop, you are welcome to email me the pdf-file of your presentation at least 24 hours before your talk (please wait for my email confirming that the talk has been successfully received and that it can be displayed properly, otherwise you will need to get a different laptop for your presentation)

  • On the day of your presentation:
    • set up everyting early enough so we can start on time
    • during your presentation, stand up and try to speak freely and while facing the audience
    • after your presentation, get my input/feedback
    • collect the input/feedback from your colleagues on your presentation
    • email me the pdf-file of your presenation for my record and for grading

    Updated: December 16, 2011, Irmtraud Meyer