3 RSEs demo their own software

RSE Demoes & Panel Discussion

Curious what RSEs in different domains get up to?

We've invited 3 RSEs to showcase projects they have completed. See the description of each demo below:

1. Marianne de Heer Kloots - Language processing models & brains
2. Candace Makeda Moore - CleanX
3. Martine de Vos - Harvesting the Internet Archive

After this, we will have a panel discussion with the three RSEs above, and Alejandra Hérnandez from RIVM. To honour International Women's Day, we will touch upon the topic of gender diversity in the profession.

Join us!

Sounds interesting? Join our NL-RSE Meetup on March 8, 2022.

Presentation summaries

Marianne de Heer Kloots - Language processing models & brains

In this presentation, I will shortly introduce a project that I am currently working on in the first year of my PhD.

We compare data that is generated inside artificial neural network models with brain imaging data of human story reading. For this project I use datasets and pre-trained models that are publicly available online. I will talk about some things I am learning while developing my codebase as a junior researcher, at the start of my research project: about structuring, documenting, and version-controlling my code, but also about working with students and collaborating with other researchers in interdisciplinary projects.

Candace Makeda Moore - CleanX

In this presentation we will discuss the cleanX package in terms of its past, present, and future. CleanX began as an outgrowth of medical professionals trying to create machine learning based solutions for certain specific radiology tasks.

The purpose of CleanX is now to help scientists, medical professionals, and programmers cooperate to create better datasets upon which algorithms related to X-rays or other medical images can be based. It is a Python package for radiology data exploration, cleaning, normalization, and augmentation.

The topics covered in the presentation will include data cleaning with cleanX, and how cleanX can be used to tackle problems of algorithmic bias including as they relate to mammography and imaging of women.

Martine de Vos - Harvesting the Internet Archive

In this presentation, we will demonstrate an AWS workflow for collecting web pages from the Internet Archive. The researcher we are collaborating with assesses the sustainability of European startups by analyzing their websites.

In this project we worked in close collaboration closely with SURF. experts They introduced us to the possibilities of AWS and the amazing tool Terraform (Infrastructure As Code). I also learned a lot of non-technical things. For example, why it is good to share your old code, even if you think it is not good enough. And that documentation and versioning are important in any stage of the project because anything can happen.

Contact

If you have any questions, please contact the host Lieke de Boer