In Japanese, there are a large number of notational variants of words. This is because Japanese words are written in three kinds of characters: kanji (Chinese) characters, hiragara letters, and katakana letters. Japanese students study basic rules of Japanese writing in school for many years. However, it is difficult to learn which variant is suitable for a certain context in official, business, and technical documents because the rules have many exceptions. Previous Japanese writing support systems were not concerned with them sufficiently. This is because their main purposes were misspelling detection. Students often use variants which are not misspelling but unsuitable for the contexts in official, business, and technical documents. To solve this problem, we developed a context sensitive variant dictionary. A writing support system based on the context sensitive variant dictionary detects unsuitable variants for the contexts in students' reports and shows suitable ones to the students. In this study, we first show how to develop a context sensitive variant dictionary by which our system determines which variant is suitable for a context in official, business, and technical documents. Finally, we conducted a control experiment and show the effectiveness of our dictionary.

One of the essential factors in community sites is anonymous submission. This is because anonymity gives users chances to submit messages (questions, problems, answers, opinions, etc.) without regard to shame and reputation. However, some users abuse the anonymity and disrupt communications in a community site. These users and their submissions discourage other users, keep them from retrieving good communication records, and decrease the credibility of the communication site. To solve this problem, we conducted an experimental study to detect submitters suspected of pretending to be someone else to manipulate communications in a community site by using machine learning techniques. In this study, we used messages in the data of Yahoo! chiebukuro for data training and examination.

In this paper, we propose a camera system which translates Japanese texts in a scene. The system is portable and consists of four components: digital camera, character image extraction process, character recognition process, and translation process. The system extracts character strings from a region which a user specifies, and translates them into English.