Skip to content

Projects

Jump to bottom Edit New page
YY Ahn edited this page Jan 15, 2024 · 46 revisions

The final project is the culmination of what you learned in the course!

For your project, you will need to produce a public-facing (although making it public is not strictly required) video content, such as a YouTube video or a series of shorts (whatever you call it -- YouTube shorts, instagram reels, or TikTok videos), that explain some cool network science. Although it is totally ok to record a "normal" presentation as your final presentation, I would like to strongly encourage you to think more creatively!

For instance, you can make a science explainer video about some of the cool network science topics that you learned just like other amazing explainer videos on YouTube (e.g., see Veritasium, Kurzgesagt, minutephysics, etc.). Maybe it's about interesting math (check out 3B1B!). Maybe it's about digging into a really interesting network data to explain how the data came about and what could be learned from the data. Or maybe you want to document the journey to learn and implement something in network science (e.g., see Sebastian Lague, Jabrils, carykh, suckerpinch and so on).

Whatever you make, try your best to add something new and unique, and your own perspectives and ways of understanding, rather than simply repeating something from books, papers, other videos, etc. Show how you really made the topic your own.

The video should be accompanied by a final paper that documents the details about the science presented in the video (a sort of written version of your video). If your project is largely a research project, you can think of the paper as a potentially publishable paper draft that presents your results. If your project is largely an implementation project, you can check out some papers in the Journal of Open Source Software. The format will depend on what you're making, but whatever you do, it will be important to communicate why anyone should care about what you are presenting. Also you can find the "standard" research paper format below. Please contact the instructor if you are unsure about how should you make your video/paper.

Team

You can either create a small team (2-4 people) or work alone. It is ok to change the team and topic after your proposal if things are not turning out well.

Authorship

If your project is eventually developed into a publication, I don't expect to automatically be a co-author of your project paper based on my input through the class. If you would like to develop your project into a publication, I will be happy to discuss it but I'll assume that I'll not be a coauthor by default.

If your project is something that you have been doing before the class or you are "using" your project across multiple courses, you should disclose that and clarify which part was done within the scope of this course.

Some project ideas

Types of projects

There can be many different types of projects! Discuss your crazy ideas with us! Here are some common types (but don't get constrained by them):

  • Hypothesis-driven projects: well, it's the usual "normal" science. You use network science to solve a substantive problems in social, biological, or any other complex systems.
  • Data-driven projects: You find or create an interesting network dataset and analyze it with available tools and algorithms, to describe the properties of the network. Note that there are journals that publish datasets, such as Scientific Data.
  • Method-driven projects: Develop and/or implement a new network metric or method, and document the journey!
  • Replication projects: replicate an existing paper. Replication is one of the best ways to learn research processes and also a good service to the research community, especially if you share your implementation.
  • Implementation projects: Many methods can't be used easily just because there is no open-source implementation of them. Why don't you implement an existing method and publish it so that others can easily use it? Note that an open-source package can be published in places like the Journal of Open Source Software.
  • Explainer video: identify a really interesting part of the course. Do a deep dive and make a cool explainer video.

For research-oriented projects, the expectation is that you will create, at the end of the semester, a mini research paper, that can be a seed for a future publication or working prototype of a research product (e.g. software packages or methods).

For those that are more geared towards explainer, it will be all about how cool the content is and how well it gets delivered.

Results from the past

Regarding research Multiple papers came out from this course! And it will certainly keep happening. :) Here is a selected list of papers that came out of the course:

Deliverables (check Canvas for the due dates)

Proposal

You will submit the written proposal and a team contract. The proposal will be a two to four-page document with:

  • Title
  • Team members
  • Abstract: a brief summary of what your project is about.
  • Introduction: why your project is important or interesting? Why should we care? What's your unique angle?
  • The proposal: So what are you trying to do? What is your approach? Why is it feasible? What is your plan? Can you lay out concrete steps that you want to take?
  • References and acknowledgments

Check Canvas for the link to the template for the team contract.

Rubric

  • (10pts) The proposal contains all required elements (Title, team members, abstract, introduction, proposal, references, and acknowledgment) as well as the team contract.
  • (30pts) The proposal tackles an important/interesting problem and provides a well-written, convincing motivation and introduction that is robustly supported by the concrete data and literature. It also clearly explains what has been done and what is missing.
  • (30pts) The proposal clearly states potential approaches and provides good rationales for the chosen approaches. It also lays out concrete plans and milestones towards the objective. It identifies concrete datasets and methods in each step.
  • (10pts) The proposal properly cites all the materials that it is built on and acknowledges any materials, helps, and ideas from colleagues.

Examples

  • A method implementation project: the proposal identifies a method that showed great potential in solving an important open question. It explains why the question is important and why the identified method has a good potential to address the question. It then reviews similar method implementations and what is missing out there. The proposal proposes a plan for the implementation (choice of language, major milestones, major tests and evaluations to implement, workflows, etc.) as well as the rationales for these choices.

Project pitch

Create a short video (1-2 minutes, think about TikTok posts) that explains why your project will be awesome.

Final presentation

Record a video (or a series of shorts/reels) and upload on YouTube. You can use other social media platforms if you'd like to. You can either publish it or make it unlisted, but I'd encourage you to make a public video for more self-imposed pressure for quality. 😉 The format is up to you. Pack it with serious science and make it interesting and fun!

See Presentation recording for some recording how-tos.

Rubric

  • (30pts) The presentation video makes a very strong case of why it is worthwhile, interesting, and important.
  • (40pts) The video explains really cool science!
  • (30pts) The quality of production; it has a powerful narrative structure, attention to detail, and good production quality.

Final paper

You should submit the final paper as well as self/team evaluations.

A good structure for a research-oriented paper would be:

  • Abstract
  • Introduction
    • Motivation & Review of literature
    • Objectives, Hypotheses, research questions, etc.
  • Methods
  • Results
  • Discussion & Conclusion
  • References
  • (Optional) Appendix

The structure may diverge if your project is not a traditional research-oriented project. For instance, if you are creating an open-source package, you can think of papers in the Journal of Open Source Software and focus more on creating a good documentation for the project. If you are unsure, consult with YY!

Rubric

The rubric items may be slightly different depending on the nature of the project. For instance, if it is about a method implementation, the rubric will be similar to https://joss.readthedocs.io/en/latest/review_criteria.html, paying more attention to the engineering and open source project side of it (e.g., how well it is tested, how well it is documented, etc.).

If it is a scientific research, the following rubric will be applied.

  • (10pts) The paper contains all required elements (Title, team members, abstract, introduction, proposal, references, and acknowledgment) as well as the self/team evaluations.
  • (40pts) Introduction: The paper tackles an important/interesting problem and provides a well-written, convincing motivation and introduction that is robustly supported by the concrete data and literature. It also clearly explains what has been done and what is missing.
  • (30pts) Methods: The paper clearly describes the methods employed by the project with sufficient details (enough to replicate what has been done).
  • (40pts) Results: The paper executes the employed methods correctly and clearly presents the results (having negative or unanticipated results will not negatively affect the grade because it's outside of your control).
  • (30pts) Discussion: The paper extensively identifies and discusses its limitations. The paper also discusses future directions.
  • (10pts) The proposal properly cites all the materials that it is built on and acknowledges any materials, helps, and ideas from colleagues.
Add a custom footer
Add a custom sidebar
Clone this wiki locally