Learning kanji characters for Japanese second language learners is a major component in their development of reading pro®ciency. The focus of this study is to evaluate ®ve web-based kanji programs to determine what features may promote learning for autonomous learners. Discovering the features of an Internet program that will promote vocabulary learning will provide programmers with a basis for creating and improving existing programs. In order to discover what is necessary for Japanese second language kanji learning, we ®rst examined the existing literature on autonomous kanji learning and the evaluation of the effectiveness of CALL for kanji learning. From this theoretical foundation, we created our criteria to evaluate existing Internet programs. The main objective is to provide suggestions for improving WWKanji, a web-based program. The results of our evaluation indicate that for future versions of WWKanji, features such as grouping by radicals, mnemonics and ideographs, drills and quizzes, additional compounds, and ®nally kanji learning in context should be added in order to provide the autonomous learner with more resources for learning. Furthermore, each feature that will be added must be rigorously tested to determine its effectiveness for autonomous learning of kanji characters.

The purpose of this study is to examine methods of measuring syntactic complexity by analyzing an original corpus of written Japanese data from native speakers and learners of Japanese. We compared two measures, mean dependency distance (MDD) and mean hierarchical distance (MHD), which have been examined using in English in previous studies. Our research question is to compare the two methods and evaluate them in order to develop an index for measuring Japanese learner's syntactic complexity.

No abstract