Researches

Research Fields

Machine Learning, Artificial Intelligence, Information Science, Statistics, Educational Technology

Research Topics

Bayesian Network, Active Mining, SVM, Decision Tree, Collaborative Filtering, Recommendation Systems, e-Learning, e-Testing, Expert Systems, Web Mining, Data Mining, Text Mining, Data Analysis System

Baysian Network

Our extensive study coveres both theoretical research to system implementation. The figure shows the interface of our original software “Bayesian Discovery”.

Bayesian Discovery allows users to fuse his/her prior knowledge, and infer the resulting Bayesian networks.

We have been building our own learning methods, principles and algorithms from fundamental theories.

Web Recommendation System （Joint development with Recruit Co.,Ltd.）

Collaborating with Recruit Co.,Ltd., we developed a system which presents an optimized page for each individual by using collaborative filtering. The system use web browsing and purchase history data of customers in the web-site ISIZE. ISIZE has over 30 million web pages and over 3 million customers.

The system provides appropriate information for each customer.

Internal Engine：The world's largest Bayesian network — Internal Engine：The world’s largest Bayesian network

Learning Management System “Samurai”

I started to study “e-Learning” before the word appeared.
The Learning Management System does not teach, The system induces a natural learning.

The online analysis system for the messages on the electronic bulletin board.

This system extracts important keywords from the electronic bulletin board automatically.

This system generates the relationship diagram between keywords and learners automatically.

This system generates messages using the intelligent agent：Generate messages by compareing a user's and high achiever's learning history. — This system generates messages using the intelligent agent：Generate messages by compareing a user’s and high achiever’s learning history.

National Project：：e-Testing System

National Center for University Entrance Examinations is the largest national exam in Japan, and there are also over 500 national exams. These exams will be all e-Testing. Ueno Lab. has planed, designed and implemented these e-Testing systems.

e-Testing = Statistics＋Psychology＋Computer Science

Item bank of Information Technology Engineer Examination

Automated construction of multiple equivalent test forms

Item database system of National Center for University Entrance Examinations