Overview of Adaptive learning

Adaptive Learningkenji ODA

Massive Open Online Courses=MOOCs

Overview of Market

• Market size• Potential Market• World population ： 7.1billion• English speakers: 1.2billion• Internet user: 2.8billion• Internet user who use English:

1.3billion• Existing Market• Coursera 2012: 6.5 million• Edx 2012: 1.6million• Udemy 2010: 2 million• Approximately 7 ～ 10 million

existing users for MOOCs and increasing

Ave. 34yfulltime

Bachelor or higher

Edtech

Request for Startup by YC

MOOCs

Problems of MOOCs

• Only 7-9 % students finish their courses.• High educated and rich people tend to take courses.(Primarily purpose is

to provide free courses to people all around the world especially poor people)• Monetization is not succeeded.

[Causes]Motivation of studying is not to get certification, but just for curiosity or enhance themselves.→Give them Nanodegree （ Udacity ） for monetize.

Adaptive Learning

• One of the hot trends in edtech in the U.S.• Providing “Personalized courses”

Knewton knowledge graph

From Knewton’s white paper

Item Response Theory(IRT)

student ability (θ), item difficulty (β), and discrimination (α)

From Knewton’s white paper

Control of adaptive presentations of instructive contents of e-Learning systems(2005)

• This system delivers the appropriate contents to the student every trial. We used the contents of mathematics for senior high school students. And we tried different methods based on students' inputs and examined the results. • Using k-NN(k-Nearset Neighbor) and SVM(Support Vector Machine) to

discriminate the contents.

http://www.code.ouj.ac.jp/media/pdf2-1-3/No.3-08ronbun02.pdf