This PhD will work on Algorithm Debt, which corresponds to sub-optimal implementations of algorithm logic in the scientific software. Scientific software is the one development for scientific enquiry or data-driven decisions, such as statistical software, data science software, machine and deep learning algorithms.
You will be working with both Python and R code (reading both, coding more in Python), and also with practitioners’ surveys and insights to define algorithm debt, uncover its causes, and see how it changes in different subdomains. You will be completing a systematic literature review and mapping, and will also work studying real-world, software to study how algorithm debt manifests in OSS.
Notes: although everyone is welcome to apply, women, non-binary and female-identifying people are especially encouraged to submit.
The candidate must be admissible with a fee waiver, otherwise you will not be able to apply for scholarship. Please carefully read the information provided by our College before you do anything else. In the section Pre-application process > Step 1, you will find a self-assessment of eligibility. If you are eligible, then read the rest and continue on this page.
You must be able to move into Australia for commencement.
If the above two apply, then contact me following the steps below. Do not apply at ANU before we have talked and you have received written confirmation of my intention to support your application.
How to Apply:
Please contact me via email. Briefly summarise your academic background and work experience, describe highlights such as the courses that you found most inspiring or challenging. To signal to me that you have read this carefully, start your subject line with “[PhD Application]”, and include the word planipennate somewhere in the email body.
After that, in the body of your email you will need to show that you have some understanding of what I do, demonstrate that you’ll bring something interesting, and that you’ve identified something in the offered topics that particularly intrigues you. You can find my latest papers on my website: https://melvidoni.rbind.io/#about
You will have to include the following documents about you:
- A current resume, including publications (if existent) and GitHub/GitLab (if existent).
- Two academic letters of recommendation, including contact details (if you are successful, you’ll need 3 people for the application at ANU).
- Evidence of your English level (if you have it). Please read ANU's requirements for English Language Tests.
Additionally, you will have to include the following documents about the project:
- Attach a 1-2 page essay (11pt Arial font) discussing your thoughts about Algorithm Debt as presented/discussed in the following two papers (if you cannot access the PDF, write me an email for a copy): TechDebt in Deep Learning Frameworks and SATD in R Programming. If you cannot access them, you are welcome to email me for a copy.
- Select only one of the following open-source datasets. Wrangle the data (in either Python or R), and create one single plot to rule them all (elaborated, meaningful, clear, aesthetic). Your document must have the plot, discuss the information presented, and why it is relevant (1-2 pages only). The datasets are not related to the papers above, so don’t try to cross them. Datasets: Option 1, Option 2, Option 3, Option 4. Note that you can use either Python or R to analyse this dataset.
- Optional, but encouraged. Pick a fiction book you have read and love. Discuss, in 1 page, what you don’t like about that book. It doesn’t matter if the book is old, not translated to English, or not mainstream; I don’t care about the book’s genre as long as it is fiction. If you cannot pick a book, then comics, visual novels, animated series, TV series, even story-based games will do. Why to do this? It is difficult to be critical of things we love, and art is especially subjective. In research, critical thinking is a fundamental skill. Besides, if I read the book (or watched the series), this will be a great icebreaker.
If I hear from you, my first consideration will be whether you are genuinely interested in my research. I will also note whether you’ve taken the time to carefully read these instructions. Please don’t expect a response if you have not.
These requirements may sound onerous, but remember that a PhD is an enormous commitment, both for you and your advisor. Taking the time to carefully and thoughtfully engage with prospective advisors is an investment that is sure to pay off greatly as you embark on this academic adventure.