Preparing slides for a presentation on the new digital-humanities CSC 151

Topics/tags: Talks, CSC 151, Digital humanities

This coming Thursday, I’m giving a short presentation at the 2018 Digital Bridges for Humanistic Inquiry symposium about the new version of CSC 151 that uses the digital humanities as a motivating topic. It’s a bit strange to be presenting about that course now since I’ve yet to develop the full syllabus, let alone teach it [1]. But I was asked to speak, and so I shall.

The Digital Bridges folks are sensible and asked for slides in advance of the presentation. In part, that helps the panel chairs organize things. In part, it helps ensure that we don’t prepare for the talk at the last minute [2]. So I’m spending time trying to get the slides in order. Musing about the task at hand is likely to help me do a better job, or at least will help get me started [3].

What are the guidelines? Let’s see …

We are then asking you to discuss the project you’ve worked on with support from the Digital Bridges grant for 5-7 minutes (tops). If you would like to use images (PPT, Keynote, etc.) we ask that you send us no more than 5 slides.

In your presentation about your project, we encourage you to do two things: (1) Introduce yourself very briefly. (2) Explain what the project is, why it is exciting, and how you are using it—concisely and energetically. (à la 3-minute thesis).

Five slides. That’s not a lot. I usually go through at least ten slides in a five-to-seven-minute talk. I assume one of them should be a cover slide. I’ll do without the traditional future work, questions, and citations slides. In this case, everything I’m talking about is future work. They’ve already planned a Q&A session. And I can just put citations on the slides.

So, what do I want to talk about? I’ve recently decided that the course should have three topical foci: document representation (with SXML), document synthesis, and document visualization and analysis [4]. But I’m in a session entitled Teaching Textual Analysis. Does that mean I should focus only on the last of the three? Nah. So I should give a short overview of what it means to teach those things. I’ll probably want to mention something like topic modeling and what it means to do an intermediate overview of the LDA algorithm [5]. So that’s two of the five slides: One on approaches to digital humanities in the CS classroom and one on understand and build the algorithm.

I need a cover slide to introduce everything. That leaves me with two remaining. A context slide is likely to be useful. That slide can provide past history. For example, CS at Grinnell regularly provides students who serve as programmers for digital humanities projects; many of them have indicated that working on those projects broadened their horizons. But it’s also worth mentioning that we regularly do themed versions of CSC 151 and the reasons we do so [6]. And I should make it explicit that the primary learning outcomes of the course relate to computer science concepts, not digital humanities concepts. Or maybe that should be a separate slide. I can see some value from having a scale balanced toward CS.

That gives with five slides.

• Slide 1: Cover
• Slide 2: Context
• Slide 3: Content
• Slide 4: Comp Sci
• Slide 5: Combining

I could do slides two through five as bullet points. There is, after all, some value to bullet points. For example, having text on the screen helps people who get distracted re-orient themselves [7]. Text-based slide decks are also more likely to stand on their own. Plus, it’s much faster to write bulleted presentations.

But I know that image-based presentations are more effective. So once I’ve thought about bullet points in the notes below, I’ll work on finding appropriate images. I may even mention likely images along the way.

Context

• Focus: Introductory CS course. Hmmm … what image represents introductory CS? Maybe some program code. [8]
• Supporting interactions between CS and the Humanities.
  
• Evidence: Themed courses increase diversity. I’d like to combine two images with an arrow between them.
• Workshop-style course. I can probably find a pair-programming image.
• We use Scheme. I’ll probably use the DrRacket logo.
• This is future work. Maybe a calendar?

Content

• Document format: XML. I’ll generate some sample XML. Should I also draw a tree? Nah, not for this audience. What about the great Paul Graham quote on XML [9]? I’ll paraphrase and not put it on the slide.
• Document creation. I can draw a simple web of arrows and maybe add a bit of Scheme code to show the combination.
• Document analysis. I’m lazy. I’ll just find a picture of a magnifying glass.
• Computer science. I’ll keep the picture from the previous slide.

Computer Science

I can’t resist. I’m redrawing the previous slide with the CS picture overlapping everything else. But what should the bullet equivalents be?

• Algorithms
• Data Structures
• Efficiency
• Elegance
• Ethics
• Reuse
• Robustness [11]
• Recursion

Combining

Okay; that’s a horrible title. I was just having too much fun with Co words. Really, the last slide is an example. And I’m tired. For this one, I’ll stick with bullet points.

• Context: Why is topic modeling a useful tool?
• Algorithm: How does LDA work?
• Data: How do we store/represent the data LDA uses?
• Implementation: Build a simplified version.
• Experimentation: What are the effects of changing [this part]?
• But not: Let’s understand the statistics

Epilogue

Building the image-based context slide took an hour. Revising it took another hour. And, well, it’s a bit of a mess in the static form. Now I remember why I’m always tempted to fall back on text. Plus, now that it’s a day later and I’m working on the exact text, I want to make more changes!

All told, I spent a bit over four hours on the slide deck. I could have spent more, but I had to turn it in. The worst parts are (a) knowing that I really need to redesign the whole thing and (b) realizing soon after I sent them in that I wanted to make more changes.

If you want to see the awful thing, it’s available on Google Docs. It serves as a good example of why it’s a bad idea to try to fit too much stuff onto too few slides [12].

Coming in a future musing: My attempt to write out the text of my presentation. I’m not accustomed to writing presentations; I generally ad lib. But it seems like I might be better writing this one. Plus, I’ve discovered that Grammarly will give me an estimate of how long it will take to read aloud. We’ll see if I have time [14].

[1] I’m spending the fall developing the course and I’ll teach it in the spring.

[2] I will note that I wrote some of my best talks, including my legendary finish in exactly five minutes lightning talk at SIGCSE, the evening before the talk.

[3] I actually did the preparation and the majority of the musing yesterday. Today I’m editing the musing.

[4] Is that four things? From my perspective, visualization is a part of analysis. So I’ll leave it as three.

[5] My students are not expected to know statistics or to be particularly fond of mathematics, so it is unlikely that we will delve deeply into the intricacies of why the algorithm works. But they can implement it.
Should they? I think so. Implementation builds insight. I should implement it first.

[6] Should I include the Maria Klawe quote about diversity and themed intro courses? Possibly.

[7] I’m one of the audience members who regularly needs reorienting. When I hear an interesting idea, I start turning it over in my head and sometimes lose track of the main presentation.

[8] In the end, I selected a Computer Science is more than coding picture from https://www.edsurge.com/news/2015-12-02-computer-science-goes-beyond-coding.

[9] From Paul Graham’s What Made Lisp Different: http://www.paulgraham.com/diff.html

9. The whole language always available. There is no real distinction between read-time, compile-time, and runtime. You can compile or run code while reading, read or run code while compiling, and read or compile code at runtime.

Running code at read-time lets users reprogram Lisp’s syntax; running code at compile-time is the basis of macros; compiling at runtime is the basis of Lisp’s use as an extension language in programs like Emacs; and reading at runtime enables programs to communicate using s-expressions, an idea recently reinvented as XML. [10]

[10] Emphasis mine.

[11] I had not originally included robustness in the list. But it’s an important point I wanted to include. Plus, now I get three R words.

[12] I thought about including the Context slide as an image. But, well, I would not want it as my representational image. And it does make some sense if you watch the animation, rather than trying to take it as a whole.

[14] We’ll also see if Grammarly is accurate.

Version 1.0 of 2016-08-07.