Spotlight on asylum claims
This is part of the Women Driven Development hackathon, #TechItForward with Pride.
Proposed by Megan Stodel, this project aims to look at outcomes of asylum claims made the grounds of sexual orientation and/or gender identity (SOGI).
Our first challenge is to collate from the government website the decision documents of SOGI-related cases. Our current approach relies on a web scraper using lots of regular expressions. This involves parsing the contents of decision documents to classify them as successful or unsuccessful (or abandoned/withdrawn, i.e. no decision), and then identifying the subset of those cases that are SOGI-related.
Looking forward, we would like to consider the possibility of a machine learning approach: replacing the regular exprssions with a model trained to differentiate between successful and unsuccessful cases, and identify those that are SOGI related.
Link to presentation
Megan has also written a blog post about the project.
Megan Stodel: @MeganStodel
Ben Knight: @benknight135
David Thai: @dafuloth
Rish Kumar: @rishkum
The project focuses on better understanding the outcomes of asylum claims made by people on the grounds of sexual orientation and/or gender identity (SOGI), using two key sources:
- Decisions posted by the Upper Tribunal (Immigration and Asylum Chamber)
- Official statistics: Asylum claims on the basis of sexual orientation
The hackathon was on Friday, 26th July 2019. We continue to work on it.
The Trello board for this project can be accessed here.