A short guide to structuring academic presentations
published: 2018-04-23 19:04:56, updated: 2018-04-23 19:06:40
based on a message I sent to a friend
Presenting the paper:
- Present the same material as the paper or accompany the paper? Most conference that I know about have the form where they assume that people viewing the presentation have not read the paper, its abstract, or even its title. I think this is the norm everywhere. So the presentation is structured to provide a high-level description about the paper, its findings, etc. and if they want more details, they refer to the paper.
- Audience - are they all your peers i.e. from the same domain? If not, what things can you assume they know? Always target the present towards the majority. If most people don't know about a term, describe or define it. For me, I have to use the appropriate technical language based on the audience.
- Index - if the presentation is not a short one (5 - 10mins) then it is a good idea to provide an index slide that highlights what you will be talking about. For e.g. if I have to give a 30 min presentation on my PhD, I'll describe myself first (name, position) and then tell them what I am going to talk about: I'll start with the background and motivation of the problem, evaluate available work done in this area, then state the requirements / goals of my contribution, then state the actual contribution, then its evaluation, and then the future work. Of course, this changes with what you're presenting, but the point here is to present it as a coherent 'story' based on how things progressed.
- Describe the problem, describe the motivation - why is this important? why is it relevant? what are its key aspects?
- Work done in the area? Are there any interesting approaches that are DIRECTLY solving the problem? If they are only related, don't bother to mention them. If there are approaches that have shortcomings, mention them. If they have strengths, mention them.
- Creating a problem statement and goals - by this point, you have introduced the problem area and what has been done so far. Based on this, you create a problem statement - how to do X? how to increase Y? and you state goals based on this: X should be a ... Y should have ... These help in the structuring of your work.
- Describe your work - what did you do for each of the goals? how does it help solve the problem? What approaches and assumptions did you take/make? Is there any relevant work that you re-used or improvde upon?
- Evaluation - how do you know that your work was good? how did you evaluate it?
- Publications - if the work or its parts are published or in-press or under review somewhere, mention this here. Provides validity.
- Conclusion - what are the key things you want people to remember in the presentation, summarise them
- Future work - what do you want to do next? How to take this work further?
All of these are highly applicable to a 'technical' presentation. Conferences that revolve more around a single topic should also be structured similarly (IMHO) as the structure of problem-goals-work provides a good narrative and makes it easy for someone to follow the work.