EPISODE · Aug 31, 2018 · 34 MIN
Using Machine Learning for Humanitarian Aims (emf2018)
from Chaos Computer Club - archive feed · host Simon Bedford
At the beginning of 2017 I joined a volunteer Data Science community called Data for Democracy and became involved in a project focused on putting together a submission for a worldwide competition run by UNITE ideas (part of the United Nations) on behalf of the Internal Displacement Monitoring Center (IDMC) in Geneva. The main aim of this challenge was to build a machine learning-driven tool that can automatically identify news and other written articles about ‘Internally Displaced Persons’ (people who have been forced from their homes by conflict or disaster). The platform also had to be able to extract key facts about the underlying events mentioned in the articles. I would like share the story of how our team, that had never met in person, put together a platform using almost 100% open source tools, and ultimately won the competition. We were subsequently invited to Geneva to help implement our platform, and we presented it to the broader non-profit community at an event at the International Red Cross museum. During my talk the areas I would like to address are: 1. My experience of learning & getting involved in data science/programming/machine learning as a non-professional programmer 2. What it was like to work on something constructive with a team from around the world that had never met each other 3. The motivation behind needing this platform and how we put together a winning solution using open source tools about this event: https://c3voc.de
What this episode covers
At the beginning of 2017 I joined a volunteer Data Science community called Data for Democracy and became involved in a project focused on putting together a submission for a worldwide competition run by UNITE ideas (part of the United Nations) on behalf of the Internal Displacement Monitoring Center (IDMC) in Geneva. The main aim of this challenge was to build a machine learning-driven tool that can automatically identify news and other written articles about ‘Internally Displaced Persons’ (people who have been forced from their homes by conflict or disaster). The platform also had to be able to extract key facts about the underlying events mentioned in the articles. I would like share the story of how our team, that had never met in person, put together a platform using almost 100% open source tools, and ultimately won the competition. We were subsequently invited to Geneva to help implement our platform, and we presented it to the broader non-profit community at an event at the International Red Cross museum. During my talk the areas I would like to address are: 1. My experience of learning & getting involved in data science/programming/machine learning as a non-professional programmer 2. What it was like to work on something constructive with a team from around the world that had never met each other 3. The motivation behind needing this platform and how we put together a winning solution using open source tools about this event: https://c3voc.de
NOW PLAYING
Using Machine Learning for Humanitarian Aims (emf2018)
No transcript for this episode yet
Similar Episodes
Jun 9, 2026 ·50m
Jun 1, 2026 ·31m
May 25, 2026 ·39m
May 17, 2026 ·37m
May 8, 2026 ·39m
Apr 28, 2026 ·25m