Applied AI & AI in business /

Using AI to tailor the education of primary school students 

May 30 2021/4 min read

Differentiation in the classroom is an important skill for teachers to give all their students the best chance of reaching their full learning potential, regardless of the student's abilities, strengths and weaknesses.

Differentiation means tailoring the learning process in order to meet students’ individual needs. This tailoring is based on a continuous assessment of the student through testing and observation.

Our goal with Edulitics is to combine the potential of e-learning with what we've learnt about the need to create a solution that makes it easier to identify problem areas and give each and every student the support they need.

02/ The educational situation in primary schools today

At present, the average primary class in OECD countries has more than 21 students per teacher. That’s a lot of students for one teacher to observe and keep track of. Moreover, while students in secondary and upper education are able to self-evaluate, primary education students have not yet developed such a skill and are, therefore, unlikely to self-assess and realize that they are struggling and need to seek the teachers’ help.

Though tests do provide a sense of what the student knows, they give no insight into how the student feels. Positive emotions, such as happiness and surprise are linked to a better learning experience as they boost students’ interest and, thus, engagement. Despite this, emotional wellbeing is seldomly used as an assessment tool in education.

The Covid pandemic has brought a lot of these problems to light. Distance learning made it impossible for teachers to be able to look over their students’ shoulders to find out how long it was taking them to complete an exercise, to observe their students in class in order to be able to pick up on which of them were struggling. Distance learning made it clear that we need to start using new technologies that go hand-in-hand with existing e-learning resources in order to be able to best adapt them to each individual student.

The Covid pandemic has brought a lot of the problems with education to light

03/ The business idea and AI model

Edulitics is a browser add-on that acts as a teacher’s second pair of eyes. The add-on records the student’s screen, as well as picks up on students’ facial expressions every couple of seconds. The underlying AI technology then takes that facial expression, analyses it and labels it as representing one of five emotions: sadness, anger, neutrality, happiness and surprise. Edulitics acknowledges that learning is a complex process that involves not just cognition, but also emotion. We are essentially integrating affective neuroscience and artificial intelligence in order to create an advanced analytics tool for education.

The Covid pandemic has caused a surge in the use of technology and e-learning platforms in schools, hence there’s an opportunity to provide a solution to combat the negative consequences of not having in-person teaching (e.g. teacher observation becomes difficult/impossible, which means some students’ struggles pass unnoticed).

Our application can be used as an observation tool both during an in-class e-learning session or outside of class if students are given e-learning activities as homework. The potential of AI technology for education is manifold. If our project gains the traction we strive for, the add-on could be converted to an education browser, which could potentially allow the application to detect emotion during live teaching.

The application can be used as an observation tool during an in-class e-learning session or outside of class for homework

04/ The dataset

For training the AI model, we used labeled images of human faces. The dataset contains approximately 35.000 examples of 48x48 pixel grayscale images of faces divided into ‘train’ and ‘test’ datasets. The images are categorized based on the emotion shown in the facial expressions (i.e. happiness, neutral, sadness, anger, surprise, disgust, fear). For the purpose of simplifying our dataset, we decided to remove 'disgust' and 'fear' from the dataset provided below.

05/ Target audience

Our primary market is public schools in Spain as these schools receive government funding for the implementation of technology in the classroom, which would allow them to purchase an institution-wide subscription to Edulitics. If this proves successful, we would eventually expand to the rest of Europe.

Once we have a thriving community of teachers and users, our goal is to explore an additional market: e-learning platform creators and providers. The data we collect might not only help teachers adapt their instruction to cater to their student's individual needs but might also help e-learning providers adapt their platforms to better suit their users’ wants and needs. This could be another potential business funnel.

06/ Future mission

However, we want to be bigger than just an education analytics platform. Our mission is to aid primary school teachers in tailoring education to each individual student’s strengths and weaknesses, interests and motivations, personal preference, and optimal pace of learning in order for them to reach their full potential. 

To do this, analytics are not enough. We know that it’s how the data we provide is interpreted and used that matters the most. That’s why we’re dedicated to providing in-depth information on how to interpret our data, as well as up-to-date suggestions on how to modify e-learning activities to best suit each student. We want to build a community, have a panel of experts on board whose knowledge we can leverage to create infographics and get teachers sharing insights with each other.

  • Elvira Martos-Barrios

    Finalist in the AI for Business Challenge

    Elvira studied Business & Marketing in the UK before moving to Copenhagen to do a Masters in Organisational Innovation and Entrepreneurship at CBS. Her own teaching and learning experiences, together with her time at different edutech startups, opened her eyes to the unprecedented opportunities e-learning solutions offer, particularly when it comes to adapting education to best fit each individual student's strengths. She hopes her venture Edulitics is a step towards avoiding students' struggles going unnoticed and unaddressed.

  • Tomáš Haverla

    Finalist in the AI for Business Challenge

    Tomáš is a full-stack developer with experience in graphic design, animation and project management. He studied web development in Aalborg, where he founded his own freelance company, Pharostudio, before moving to Copenhagen to work as a full-stack developer for satellite communication systems manufacturer Copham SATCOM. He is now working on developing his Microsoft Azure Cloud services skills, as well as on using his years of experience in front- and backend development to support edutech initiatives that he's passionate about.

02/ More on Business & Applied AI