AI-powered Education

How can we improve childhood learning with personalized education

How can we improve childhood learning with personalized education

ABOUT

The AI-powered Education Project is part of an AI4Good project series, between the AI for Good Foundation and the Applied Data Science with Venture Applications Course at SCET, UC Berkeley.

Education has always been the heart of all nation’s growth. The government has the vested interest in the improving both quality and access to proper education for their people. With the improvement of technology and the increasing amount of highly accessible and inexpensive online learning materials, there exists a demand for the search of appropriate learning material amongst the sea of information. With our data-driven model for learning, we help to personalise learning to each individual, providing them with materials not only suitable for their cognitive level, but also in a manner that suits their learning styles best.

PROBLEM

Problem of classes with children of different ages, with different backgrounds and therefore very different academic level.

Lack of personalization in the education makes children give up, stop going to school.

Problem of the qualities in education people are not currently satisfied with in some areas in the world.

MAIN IDEA

Create a tool to provide a more personalized education to young children. Making it easily available, easy to use and fun.

Gamification of learning, trying different teaching approaches. Using learning models to identify which approach is the more appropriate for the child.

Collect data on children’s learning with the app to improve teaching.

OUTPUT AND INTERFACE

A personalised model of each child’s learning abilities in different areas and preference in learning style.

An online learning channel (web or app) that generates questions and educational tools based on the child’s model.

Overall anonymised dataset of educational levels, needs and preferences of children in specific schools, region or country as reference for educators and policy-makers to make better investments decisions and efforts.

Possible data set requirements: Each dataset would be generated on an individual basis, though possibly compared to another source of data to understand the individual’s needs relative to the population.

Curricula of various subjects of interest

ABOUT

AI-powered Education Project is part of an AI4Good project series, between the AI for Good Foundation and the Applied Data Science with Venture Applications Course at SCET, UC Berkeley.

Education has always been the heart of all nation’s growth. The government has the vested interest in the improving both quality and access to proper education for their people. With the improvement of technology and the increasing amount of highly accessible and inexpensive online learning materials, there exists a demand for the search of appropriate learning material amongst the sea of information. With our data-driven model for learning, we help to personalise learning to each individual, providing them with materials not only suitable for their cognitive level, but also in a manner that suits their learning styles best.

Look at the presentation video!

Look at the presentation video!

PRINCIPAL QUESTIONS

How do we provide children (especially in developing countries) with personalized education content in a fun way to improve their learning process?

How do we provide educators/policy-makers with data visualization based on anonymized education data from children, in order to help promote the quality of educational investment for the specific children/area?

PRINCIPAL QUESTIONS

How do we provide children (especially in developing countries) with personalized education content in a fun way to improve their learning process?

How do we provide educators/policy-makers with data visualization based on anonymized education data from children, in order to help promote the quality of educational investment for the specific children/area?

DATA CATALOGUE

Since our aim is to create a highly personalized learning model for the individual user, the main data source will be from the user themselves. We will first inquire about the user’s general personal data such as education level and age which starts them out at the baseline. From there, we will then fine tune the dataset based on the user’s question and answer inputs using an underlying machine learning model.

AI2 Science Questions

A dataset of middle school science quiz questions.

OECD

The Programme for International Student Assessment (PISA)

DATA CATALOGUE

Since our aim is to create a highly personalized learning model for the individual user, the main data source will be from the user themselves. We will first inquire about the user’s general personal data such as education level and age which starts them out at the baseline. From there, we will then fine tune the dataset based on the user’s question and answer inputs using an underlying machine learning model.

AI2 Science Questions

A dataset of middle school science quiz questions.

OECD

The Programme for International Student Assessment (PISA)

RESEARCH TEAM

DATA – X TEAM SCI-KID LEARN

RESEARCH TEAM

DATA – X TEAM SCI-KID LEARN

NIEMA ELBOURI

Researcher

MICHAEL LAWRENCE

Researcher

SHINE SHAN

Researcher

NIEMA ELBOURI

Researcher

MICHAEL LAWRENCE

Researcher

SHINE SHAN

Researcher

ANDY SPEZZATTI

Researcher

ADA TAN YIN DAWN

Researcher

ALJAZ KOSMERLJ

SCIENCE lEAD

ANDY SPEZZATTI

Researcher

ADA TAN YIN DAWN

Researcher

ALJAZ KOSMERLJ

SCIENCE lEAD