Abstract:

In this paper, we develop a web system for learning words by using coil method.

The coil (Combination Of Initial Letters) method has been proposed as an English vocabulary learning method. The coil method aims to consolidate the memory of related English words by learning them all at once. For example, suppose we learn the English word "love". The word "love" is related to the word "life" and "emotion." The coil method is a way of learning to memorize these imagined English words together. A coil question consists of a single keyword and the words related to it.

There is a problem with coil method. There are no teaching materials with images and sounds for the words by using coil method. Furthermore, it is difficult to make multiple coil questions with the same difficulty level of the words.

Therefore, in this paper, we develop a web system for English vocabulary learning by coil method to solve this problem. First, we generate coil questions with the same word difficulty level automatically. Then, the generated questions are made available through our web system that has features of images and
soundes to improve the learning efficiency compared to conventional learning methods.

First, we prepare word lists with English words classfied into each difficulty level and category. We assume that words in the same category are related to each other. Create a coil question from the words in the same category. From the word list, we select a keyword to generate a coil question. Then we look for the English words using the initials of the alphabet that make up the selected keywords in the word list. If the word list contains such a word, that word is used in the coil question. If all words are found for the alphabets that make up the keyword, then one coil question is complete. Then the words used in the coil question are deleted from the word list, and another coil question is generated again by selecting keywords from the remaining words in the word list. We repeat this process, and when the coil question can no longer be generated from the words in the word list, the generation of the coil question for that word list is terminated.

The "coverage rate" is the ratio of the number of words used to generate the coil question to the number of words in the word list. We want to use as many words in the word list as possible since learners need to remember as many words in the word list as possible. In other words, we want to increase the coverage rate. In the following, we focus on the coverage rate.

We generate coil questions by using four methods, respectively and evaluate the coverage rate. The evaluation results showed that the generation method using genetic algorithms showed the highest coverage at all difficulty levels among the four. Furthermore, when the number of times an word could be used twice, the coverage rate was greater than 50% for almost all difficulty levels. Based on these results, we decided to use a genetic algorithm for automatic generation of coil questions in our web system.

We developed a web system to facilitate learning by using the coil method and investigated the learning effect. The results showed that the learning effect of the web system we developed was higher than that of the English vocabulary book.

A questionnaire conducted to users of the web system showed that Many users found it easy to learn.

Keywords:

e-learning system, English vocabulary, Web application.