BACK
Contact details
Submitted by: Lauren Lichtman, Partnership Lead
Emails:
Website: https://learningequality.org/
Reflections on refugee and host community educators on the process: https://blog.learningequality.org/hackathon19-debrief-7f1911d9b109
Social: #design2align
Introduction to the project
Country
Kenya
Duration
March 2017 - December 2020
Description
UNHCR, Learning Equality, Google.org, Vodafone Foundation, and UNESCO convened a hackathon. This hackathon aimed to prototype a tool or set of tools to automate aspects of the curriculum alignment process. The intention is that this will become a public good that can support curriculum alignment in emergency and crisis contexts, show the potential to be integrated with policy work, and strengthen cohesion around existing projects in the space to benefit both refugee and Open Educational Resource (OER) communities.To take an initial step in prototyping this work, we started by focusing on the curriculum itself (and not the content within curriculum) as an entry point.
One of the outcomes behind this tool is that it will be easier to automate aspects of the process of curricular alignment in host countries, this will ease the pressure if the curriculum being followed by refugees is the host country curriculum.
Project aims
Google.org, Learning Equality, Vodafone Foundation, UNHCR and UNESCO have joined forces to explore the question, “How can we create a tool that helps to automate the mapping of digital learning resource to national curricula in an effective and efficient manner?” The aim is to produce public goods supporting a semi-automated alignment tool (or tools) that can be used by curriculum specialists to improve upon this process, but can also benefit other education stakeholders including ministries of education, curricular bodies, and Open Educational Resource communities.
Resources used
What is notable is that we have taken a multi-stakeholder and consultative approach that has engaged more than 400+ individuals, including beneficiaries, in the process. We have convened a series of consultations—including design-thinking style sessions in Kakuma Refugee Camp, and workshops at the 2017 WISE Summit and 2017 and 2018 UNESCO Mobile Learning Week— which led to a design sprint in March 2019 in Paris, France, following UNESCO Mobile Learning Week. There, we developed designs for open-source tools to make the process of curriculum alignment easier and faster, usable by a variety of platforms and projects, with a focus on primary and secondary education. Most recently, we organized a hackathon in October 2019 in San Francisco, California, USA to take the first step towards prototyping, and after a productive three days we were able to:
- Obtain an initial understanding of how to automate the digitization of curricula that are often in non-machine-readable PDF or hard copy formats.
- Create a rubric to inform how a machine can make curricular alignment predictions.
- Develop a user interface to collect human judgment data from comparing two parts of curricula to train and improve future predictions made by the machine learning model.
- Develop a user interface for incorporating the machine learning recommendations into an overall user journey.
These 2019 sessions were unique in that they brought together those with experience and expertise in curriculum, machine learning, and education in emergency and crisis contexts. Refugee and host community educators, as well as UNHCR Education staff from Kenya, who would most benefit from the production of these public goods contributed remotely in Paris and in person in San Francisco to work alongside others in actively developing these public goods. Engaging them in this community of practice, receiving their input and early buy-in will be especially helpful throughout this process.
Partners
- UNHCR
- Vodafone Foundation
- UNESCO
- Learning Equality
- Support from Google.org
Challenges and how they were overcome
We are working to overcome two challenges:
(1) identifying curriculum to digitize.
(2) engaging individuals with relevant skill sets to generate data to inform a machine learning model.
We’re still in the early stages of the development of the necessary public goods, although our prototype tools from the hackathon demonstrate that we have developed some tangible prerequisite outputs that have value.
Results of the Good Practice
- Understanding how technology that will benefit them is developed in practice
- Empowering to be a part of the development process
Next steps
We are looking for copies of curricula to digitize, and curricular experts to make the necessary judgements in comparing two parts of curricula to improve the machine learning model. All the outputs of this work will be shared back openly with the community to inform our respective projects.
