Adaptive Learning Path and Detection App for Dyslexic Children

Inspiration

The motivation behind this project came from a strong desire to make education more inclusive for children with dyslexia. Dyslexia is a common learning difficulty that can affect children in different ways, from struggling with reading to writing and processing information. The lack of personalized learning tools that can adapt to the specific needs of dyslexic children inspired me to create an app that not only detects the type of dyslexia but also tailors learning paths to each child’s ability.

What I Learned

Through the course of building this app, I learned several key concepts, both technical and educational:

• Adaptive Learning: How personalized learning paths can enhance skill development by adjusting content based on performance.
• Machine Learning in Education: I learned how to apply machine learning models to classify dyslexia types and predict learning progress.
• Front-End and Back-End Development: I expanded my knowledge of front-end frameworks like React Native and back-end cloud interactions for real-time data storage and analysis.
• Importance of Data in Education: Gathering relevant data about student performance and learning preferences was key in improving the user experience and adapting to individual needs.

How I Built the Project

Technologies Used

• React Native: For building the mobile app, providing a user-friendly interface that allows children to interact with the learning modules.
• Machine Learning Models: A combination of supervised learning models was used to classify dyslexic types (surface, phonological, rapid naming, double deficit) based on test scores and survey data.
• Data Analytics: I generated both real and synthetic datasets to simulate various learning scenarios and dyslexic characteristics. Features such as reading speed, memory, visual and auditory discrimination were essential for personalizing the learning experience.

Main Features

• Dyslexia Detection: The app begins with an assessment phase, which evaluates the child’s current reading and writing abilities and classifies the type of dyslexia, if any.
• Adaptive Learning Path: Based on the assessment results, the app generates a personalized learning path. Topics are broken down into steps, and each step includes 10-15 questions of varying difficulty.
• Parent/Guardian Dashboard: A separate dashboard for guardians allows them to monitor their child’s progress and receive helpful insights via a chatbot. The chatbot is fine-tuned to answer specific questions about dyslexia and the child's performance.
• Mastery-Based Question Distribution: The learning path adapts in real time by redistributing questions based on the child’s mastery level, helping them focus on weaker areas.

Challenges Faced

• Data Collection and Balance: Ensuring the dataset was balanced for all dyslexic types while generating both synthetic and collected data was challenging. I had to experiment with various methods to make sure the dataset was both realistic and usable.
• Question Generation: Dynamically generating questions that are both educational and personalized for dyslexic children required careful attention. Making the questions progressively harder while ensuring they remained understandable was a fine balance.
• Back-End Development: With little experience in back-end systems, integrating cloud services and ensuring smooth communication between the front-end and back-end took significant effort.
• Real-Time Adjustments: Implementing a real-time, adaptive learning system where the difficulty of questions adapts on the fly based on user performance was complex. Tuning the models to predict progress accurately based on varying learning paces required additional iterations.

Conclusion

This project taught me the importance of personalization in education and how technology can be a powerful tool for inclusion. By combining machine learning with educational principles, I was able to create a tool that supports dyslexic children in their learning journey. The challenges I faced helped me grow as both a developer and an educator, and I hope this app can make a difference in the lives of children who need it most.

Built With

• google-cloud
• react-native
• tensorflow