Online Language Learning AI Assistant that Grows with People
Project Overview
- Funding Name: NEDO “AI That Evolves With People“
- Theme Category: 1-2 “Explainable AI”
- Project Name: “Online Language Learning AI Assistant that Grows with People”
- Principal Investigator: Yoichi Matsuyama (Waseda University)
- Contractor: Waseda University
- Co-Contractor: Waseda University Academic Solutions Inc.
Responding to the growing need for online education brought about by the COVID-19 pandemic, our research group provides advanced English-speaking education support by fully utilizing the highly interactive conversational AI technology. Using the experience and know-how of Tutorial English, a communication-oriented English-speaking program developed and implemented by the Waseda University, English-language education experts will work together with AI to assess and improve students’ English-language communication abilities. The ultimate goal of this project is to realize an online learning AI assistant system that grows and engages with people over time.
Research Themes
1. Massive Conversational Data Collection and Analysis Ecosystem
We are building large data sets of English conversation data and actual online English conversations based on scenarios designed to take a well-balanced look at students’ English speaking abilities. >>Read more
2. Dialog-based Automated Speaking Skill Assessment
English speaking ability assessment has generally been dependent on the knowledge and experience of individual judges, leading to generalization issues. In this study, we have proposed a collaborative framework where humans and AI can grow together. A highly-experienced human judge will make final judgments, giving priority to cases where AI judgment is determined to have low accuracy, by using an active learning framework that can explain the judgment process by preincorporating the English speaking ability assessment framework into the structure of the neural network. >>Read more
3. Smart Online Language Learning Experience Environments
We are using developed English speaking ability assessment system to develop a framework for providing both students and educators satisfactory feedback by combining knowledge gained from analysis of large-scale real-time English conversation data with supervisory guidance from English language education experts to elucidate and explain linguistically how the English language abilities of students participating in online English conversation classes develop in the short and long term. >>Read more
The current research uses the Common European Framework of Reference for Languages (CEFR) as its theoretical basis. According to the CEFR, humans are defined as social agents and are organized broadly and comprehensively according to how they should be evaluated for quality communication skills and the strategies for each particular task. It is expected that the English-speaking support AI, which is designed based on such theory and large-scale data, will serve as the basis for technology and services that can explain short- and long-term learning effects to both students and educators.
Background
The NEDO Project: “AI That Evolves With People”
This section introduces the NEDO project which is being implemented as a five-year project from 2020 to 2024 to address explanatory and reliability issues in AI reasoning results. >>Read more
Waseda University’s Tutorial English
This section introduces Tutorial English, a communication-oriented English-speaking class at Waseda University. >>Read more