Aurai, like War Child Holland, is based in Amsterdam, and also like War Child, we do not see ourselves simply as Amsterdammers; we are world citizens responsible for each other regardless of geographies. Aurai began as a mix of human resources, sociology, data science and engineering and machine learning applied to future-proof the longevity between talent and organizations. Our company has grown as we assist a diverse number of missions by engineering data analysis through machine learning. One of these partnerships was with War Child’s Can’t Wait to Learn Team as they work to improve the education gap in conflict affected areas.
Technology Solutions for Conflict-Affected Children
Young students come to class eager to play and grow; no child should be forced to wait to learn. One in six children worldwide live with the effects of armed conflict. Education, a fundamental right, is frequently overlooked when access to food, safety, shelter or water becomes scarce. Nevertheless, in emergency, post-conflict and refugee contexts, it is important to reach as many students as possible. Engaging children in their natural tendency to play is all the better as it increases hope and resiliency. The current boom in education technology and e-learning has offered practical and cost-effective resources to combine learning and play, so that no student is waiting to learn. War Child brought together a diverse partnership to create Can’t Wait to Learn to provide an educational technology solution to the teachers and students who need it the most.
Predictions With Data Analytics
Can’t Wait to Learn began as E-Learning Sudan in 2012 in order to provide learning opportunities in a country where over three million children do not have access to school. Since then, the program has also been deployed to Lebanon, Jordan and Uganda. In November 2018, Aurai partnered with War Child to research and test whether log data from children’s game play could be processed with data analytics to predict which children are at risk of drop out.
The cooperation between War Child and Aurai began within our eight-week full-time intensive candidate training with a custom, and condensed version of a Design Sprint. A standard element of the Can’t Wait to Learn design process at the start of a new partnership is to give insights into the students’ and teachers’ experience with the program with, among others, qualitative data. This step crucially helped us to add a human dimension based on actual field experience to our approach based on hard science and data.
Ensuring no child has to wait to learn through practical and cost-efficient e-learning programs
The Art of Machine Learning
The art of machine learning involves finding the right algorithms for the right data sets. We applied ours to find trends and correlations as the learning management system follows the individual student’s progress through the educational games on his or her personal tablet. The Can’t Wait to Learn briefing helped us match the gameplay data to the educational objectives. In order to make predictions with the log data, we had to separate and identify signposts marking the routes of gameplay and their correlation to students acquiring reading, writing and math skills. Data points such as learning curves, attendance, and engagement in activities over time proved the best indicators for tracking and monitoring learning outcomes. We linked the indicators to reflect the performance and habits of the end user as he or she progresses through the game’s levels. Using existing data from two countries, we were able to tailor a custom predictive program, which took into account the actual obstacles and opportunities in the field. Knowing the past outcomes of the data sets provided meaningful statistics and dropout predictions from the gameplay.
Doing Well While Doing Good With Technology
With the anonymous data sets arranged correctly, the data scientists could make suggestions for the tablet interface to help increase engagement in the gameplay. We congregated and analyzed the number of games played, the level of difficulty achieved, inactive periods and the rate of stopping or quitting a game, attendance, and the rate of skill accomplishment. The results of the analysis were geared towards helping teachers identify at-risk or struggling students, or conversely, students who may need more stimulation or material. These automatic assessments are crucial when considering many students in post-conflict contexts may be too stressed, shy, or distracted to communicate their difficulties.
Our data scientists had the opportunity to gain valuable skills and competencies by incorporating human, social and cultural factors, and bringing more empathy into our Design Sprint. We are very proud of our contribution to the Can’t Wait to Learn program because it exemplifies how professional technical skills and the latest technological developments can and should be used to improve our shared future. Moreover, the implicit challenges in adapting machine learning to Can’t Wait to Learn increased the expertise and abilities of the staff and pushed us to advance our technologies, proving that alliances between businesses and non-profit organizations are by no means unprofitable: a company can do well while doing good in the world.