ProjectProposal-RolandCarlos

From CS160 User Interfaces Fa06

日本語できる？ - Can you do Japanese?

Contents 1 Problem Context/Forces 2 Problem Description 3 Target User Group 4 Solution Sketch

Problem Context/Forces

One common difficulty that many Japanese language students face is handling the complex system of kanji [1] (Chinese characters) that Japanese uses in its reading and writing system. The scope of this problem can best be quantified in this example. Latin-based languages only use a limited set of characters, small enough that most keyboards use a one key-to-one character correspondence. However, Japanese does not have this same convenience.

While the language does use limited alphabets in its language (hiragana [2] – words in Japanese that have no kanji, and katakana [3] – foreign loanwords in Japanese), it also uses kanji, borrowed from the Chinese language, of which thousands are in common use and tens of thousands exist. Adding to this difficulty is the fact that kanji are read differently depending on how it is used (for example, if a kanji is used by itself in a sentence, or if combined with another kanji into a compound). (Example: 日- this character means “day”　日本 – but using the same character is a compound makes “Japan”)

Therefore, kanji is a problem that all Japanese language students have to face. For students whose native language is Latin-based, the difficulty is obvious. In addition to learning how to deal with a system that uses a large number of characters, many of the characters are completely foreign to them, making the prospect of learning kanji that much harder. However, learning Japanese is even somewhat difficult for Chinese students. While they know may the meaning for each kanji (Japanese kanji, for the most part, retains the meaning of the original Chinese it is based from), they do not know how to voice it into the Japanese language, as the pronunciation between Japanese and Chinese is completely different.

Problem Description

Kanji traditionally has been pointed to as the biggest problem for many Japanese students. The main scope of this project is to help accelerate the learning/translating process for Japanese language students.

Using the Anoto pen and paper system, which allows us to send our pen strokes to the computer, we want to try and use the existing data exchange process to support our kanji learning and translation objectives.

Answering the Stroke Order problem

Some kanji characters are made up of a multiple number of strokes and each character has a specific order in which the strokes are written. The order exists to make writing the characters more natural and easier, but many times students will fall into their own bad habits and write characters the incorrect way.

Think of it this way, when you write the letter “b”, you write the line before you write the semi-circle. The wrong way would be to write the semi-circle before the line. Now multiply this problem by the thousands of characters that Japanese language students have to deal with.

Therefore, the Anoto pen and paper system can be used to help Japanese language students learn the correct way to write kanji. The Anoto system keeps track of what stroke was made and when. The when is the most important part of this system (because how can we know stroke order if we don’t know which was written first?) and luckily it has already been implemented. Therefore the student will write his kanji on the Anoto paper and our proposed project system will check each individual stroke for correctness (kanji is strict about if a single stroke is done left-to-right vs. right-to-left) as well as if all the strokes for a character are written in the correct order. All of this will be accomplished using the time-based information and character recognition software.

Making Flash Cards

One of the most tedious projects for a Japanese language student to undertake is the creation of flash cards. It is messy, takes a long time, due to their paper nature, it is easy for them to be lost or destroyed, ruining the work put in by the student.

Digitizing this should make lives easier for Japanese language students. The student could write the kanji character out on the digital “front” side of the paper. Then the student would “flip” the digital copy, and write/type the meaning of the kanji on the “back”. The flash card would then be saved into an image file (of the user’s choosing). The proposed project system would allow users to create their flash cards, by taking in “front” and “back” copies of a digital card (obviously two different image files), but allowing the user the digitally “flip” them.

Additionally, since they are saved digitally, they can be printed into real cards (either by a template provided by the proposed system or by a printer that prints on both sides) and copies can be easily made again if the physical ones are lost or destroyed.

Dictionary Function

Also, a student can write a kanji character on the paper and let the computer try and figure out what it means for them. Based on the stroke order and structure, the proposed system will attempt to try and find the kanji in its dictionary and give the student the meaning. Since reading characters is an inexact science, the system will also give some other “nearby” (other characters with similar structure or stroke order) characters that the student can also look at if they wish.

Recognizing Characters

It is difficult for a computer to perfectly recognize human handwriting. It is difficult enough with the English alphabet, which only has 26 characters. The thousands of kanji that we have to deal with make the task seem monumental.

However, there has been some work done with taking in natural strokes into readable characters. One of the best examples is with the Microsoft IME (Input Method Editor) [4]. Microsoft created the “IME Pad”, which allows users to write in their strokes, which the system then uses to try and find the kanji the user wants.

The ideal for our proposed system is to be similar to the “IME Pad” system. But the question has to be asked, why don’t we just use the “IME Pad” system for this project then? One main complaint voiced by many Japanese students is that while the IME Pad is useful, it is not new user-friendly and they find difficulty in writing with a mouse.

Therefore, our proposed project will have the advantage of being easy to use and being more natural. What can be more easy to use and more natural than writing?

Target User Group

The obvious target group here is Japanese language students. While it at first seems like a niche market, it is a growing one. Japanese culture has begun to take America by storm, with the introduction of anime and Japanese music into the mainstream. Naturally, more and more people will want to start learning Japanese as well. This project serves to help these people in their studies.

However, this project can also be useful not just for students, but for people looking for a refresher in Japanese. Kanji is a skill that needs constant practice, even for the Japanese people (one can point to the many Japanese game shows that test its Japanese contestants by daring them to identify a difficult kanji). The dictionary functionality will be useful as well, as text-based kanji dictionaries are less intuitive than just writing down a character and getting an answer.

Solution Sketch

[5] [6] [7]

---