How data science is changing education
Educational Technologies (EdTech) continues to transform traditional education. From the rise of online teaching to Massive Online Open Course (MOOCs), where many companies brought teaching services online to enable learning to take place anytime, anywhere and via any device, the face of education is changing.
In alignment with this development, educational content expanded from paper resources to smart content such as digital textbooks and video lectures. During this time, administrative tools continued to remain popular for schools, with a shift towards the adoption of better cyber-security practices.
Now, online teaching and assessment can now bring more value to users with the introduction of learning adapted and customised to the students’ needs. How can this work? It depends on users' responses. An adaptive online teaching platform generates a customised curriculum.
Similarly, an adaptive online assessment platform generates personalised questions for students that are just right in difficulty level and remediation content that covers the students' learning blindspots. Students' learning data can also be aggregated for easy viewing and interpreted by a human tutor, who can then follow-up with further interventions.
Artificial intelligence (AI) is a new era for educational solutions as the incorporation of AI further improves the value propositions for its users. For one, AI can enhance adaptive learning with neural architectures that overcome challenges in recommendation algorithms such as complex relation modelling, modelling temporal dynamics and cold start. AI-powered adaptive models are beginning to show great promise in predicting user responses accurately.
Practically, in a traditional adaptive assessment, questions get easier after the student attempts wrongly. However, the next easier question may come too late as the student already feels discouraged by the wrong attempt. Hence, the ability to predict the correctness of a student's response, and with high accuracy, can allow the recommendation engine to mimic the pedagogy technique of scaffolding effectively.
Besides adaptiveness, AI also enhances administrative and security tools with image recognition in areas of attendance marking and security supervision. Smart classrooms with cameras in-built also track students' attentiveness during the lesson.
AI can also generate smart content on the fly. For example, AI can dynamically generate personalised textbooks and questions that are suitable for each student's revision. There is also substantial research and development in the vision to create AI-teachers.
The main challenge in automating education is that while the tasks in the education supply chain are laborious, they are not simple. However, with the evolvement of AI, machines can now perform complex tasks, making practical automation in education a real possibility. Even though the research and development into deep tech for education are almost ripe for day-to-day integration, we believe that AI still cannot fully substitute a human teacher. Teachers are character role models and teach a wide range of soft-skills to students through human interactions.
Hazel Wei is the Co-Founder of Joni.AI, an AI-powered adaptive assessment platform that personalises education and helps students to make sense of their learning data. Joni.AI received S$100,000 investment from Accelerating Asia and joined the portfolio in January 2020.
If you’re interested in investing alongside us or connecting with Joni.ai, please reach out to us via email.