Notes on Course Projects

Course Survey Responses

• Though I sprang the question on you, responses were helpful.
• Surprised to see strong preference for course project; especially as so many don’t have a solid idea yet
• Range of stages on project idea development:
  • Some look promising
  • Others need refining
  • Still more need ideas
• Need to define the scope and expectations of the course project more thoroughly

What is a “good” course project?

• A bit more of a challenge this year, b/c of schedule: less time to complete one.
• In the past, it’s usually been an interactive tool, run from ArcGIS Pro, perhaps with some Python
• Allowing much more flexibility this year: aim for a simpler, less-ambitious product
  • ArcGIS Pro tool based on a simple workflow: what we do in section 1 applied to a new situation
  • Python-based workflows
    • Iterate a workflow across many inputs
    • Scrape data from on-line sources
    • Access ESRI’s Python based tools: Deep Learning, Image analysis
  • Visualization of workflow product:
    • AGOL StoryMaps, Dashboard, Experiences
• ==Any project that demonstrates innovation and learned expertise==

What is a good course project topic?

• Topic should focus on the technology, not the subject matter.
• Far too many topics proposed go well beyond GIS itself.
• Example: A course on MS Word to learn how to write your MP…
  • Course will teach you formatting, embedding images, footnotes, etc
  • Course will not teach you anything on forestry, fisheries economics, climate change
• ==Up to you to determine the question GIS will answer==
• ==This course reveals techniques to get there more effectively and efficiently==

An example:

Antarctic Marine Protected Areas

• Given a set of user provided shapefiles on conservation need and conservation opportunity, each labeled with reasons/justifications of each, I want to create a tool that highlights areas where both specific needs and opportunities exist.
• To do this, I need to develop some sort of query-able spatial database - maybe a hexagon mesh - where each polygon is tagged with the number of studies in which it designated a “need “or and “opportunity”. I think I might be able to do this using the ArcGIS Pro model builder or perhaps some Python scripting.
• Once this dataset is created, I will create an interface, that allows the user to explore areas where needs and opportunities meet or exceed a threshold value. I think I might be able to do this in ArcGIS Pro or ArcGIS online.

Note how the central question is really quite simple: count the number of occurrences found within a specific spatial unit. The focus is on how to get the technology to do the work and generate the products.

Note also that how to apply the technology is still uncertain. This is an acknowledged challenge with developing a course project: we don’t quite know what we can until we learn a topic. That is why we develop topics together. And why we shouldn’t be too ambitious.