In the research of CAI systems for individual instruction, various devices to select items for sub-sequent presentation according to individuals' answer have been developed. However, most of them can only be used with learning materials which have systematic or hierarchical structures. This study aims at proposing and testing a new spaced learning method which can be used for materials without such structures. The method is called "the Low-First Method", and is derived from psychological findings relating to the cause of spacing effects based on the change of human memory activity. The Low-First Method is based on the following two principles : (a) sorting all the items at the end of each learning session by their weighted cumulative recall rates (P_ns) in ascending order for the subsequent session ; (b) omitting items whose P_ns have reached a certain level. In the experiment, 24 participants learned HTML using a CAI with a mere repetition method and that with the Low-First Method. The results showed that the Low-First Method was not only more effective but also more efficient than the mere repetition method.

View full abstract

In recent years, NIE (Newspapers In Education) has attracted a great deal of attention, but Japanese elementary school students and learners of Japanese as a second language cannot read all the kanji in newspaper articles. In this study, we investigated which kanji would be most effective for learning to read the newspaper. The results show that : (1) Total kanji usage accounts for 41.46% of all characters, while hiragana and katakana usage account for 34.06% and 6.34% ; (2) The Basic 500 kanji, and the kanji assigned for teaching in elementary schools account for 70.41% and 89.84% respectively of the total kanji usage in the newspapers ; and (3) There are many high frequently kanji which are not taught in elementary schools. Based on these findings, we develop a kanji list for NIE and show what kind of kanji should be introduced to help students read newspaper articles.

View full abstract

This paper discusses the present education of journalization and the functions of a system which should be used in it. Moreover, it describes an ITS (Intelligent Tutoring System) which the author has been developing, that is, an intelligent CAI system named ISAK (Intelligent System with Accounting Knowledge). This system is an ITS for elementary bookkeeping. It has knowledge base for accounting and teaches a variety of accounting knowledge to students. ISAK includes several functions : a function to set study points, a function to make questions, a function to create answers, etc. It tries to enhance students' levels by using these functions. ISAK is easy to operate because it is a system for Windows. Although Intelligent Tutoring Systems have long been researched, they have not been realized yet. This paper focuses on the above functions of ISAK, accounting knowledge and student model and discusses functions an ITS should have.

View full abstract

The authors conduct an effectiveness research of e-mail discussion review support system with summary extraction method. The support system we have developed can automatically extract summary sentences from the colloquial e-mail messages using reference relationship of e-mails they have discussed. Japanese natural language processing technology is applied in the proposed method. In order to evaluate the effectiveness of the system, we conducted some experiment using questionnaire method and protocol analysis. We compared the two system ; the system with and without summary sentences in the table of e-mail content. As a result, following fact-findings were obtained. The system with summary sentences could promote reading strategy such as utilization of table of contents and comprehension of e-mail message structures. On the other hand, the system without summary sentence makes the reader give attention to the detail information such as names. From above described result, we concluded that the system with summary sentence is effective to understandings of relationship of e-mail messages.

View full abstract