| 摘要: | 華語學習隨著中國熱延燒全球，而保有最完整中文漢字的台灣，在這場全球語文學習重新洗牌的變動中，扮演關鍵的角色。另外台灣島內近來因國際化及政策開放，在台外籍人士及外籍配偶人數急遽增加，在台人士學習華語的需求殷切。有別於本國人學習母語的方式，以外國人為對象的華語學習，有許多先天上的限制，尤其在漢字的書寫上，更因中文字筆畫複雜、字數又多的特性，讓慣於拼字拼音文字的外籍人士在學習上產生極大的障礙。 部件教學法是將中文字相同的部件統一歸納教學，讓學習者在學習中文時可以利用部件方塊拼字的觀念，將中文部件拼湊、結合來組合成有意義的中文字。本論文希望使用者藉由線上手寫輸入所提供的實筆資訊、虛筆資訊、筆壓變化與筆劃前後的時間性，並依筆劃的相依性、筆畫的內聚性及連續筆劃間方向的改變等特徵，建立一個特徵資料庫，並利用SVM分類器來訓練部件斷點，藉此科學化的分類手法抽取出相關的中文字部件，以提供更有系統的筆順教學來幫助外籍配偶及外籍人士可以有系統與效率的學會書寫中文字。 Due to “Chinese Mania,” study on Chinese language is now globally getting more and more popular. Taiwan, the place which preserves the most complete and traditional Chinese characters, plays a key role in this global language study reshuffle. Moreover, the numbers of foreigners and foreign spouses are rapidly increasing, because of the internationalization and the policy of Taiwan. It is a eager need for them to study Chinese language is eager. However, different from the way how Chinese native speakers learn Chinese, Chinese language study for foreigners has certain limitations, especially in the aspect of writing. Due to the complication and huge numbers of Chinese characters, Chinese language study is particularly difficult for those foreigners who get used to spelling system. “Partial (character) teaching method” is to induce and generalize the same parts of Chinese characters and allow learners to use the combinations of partial characters to form meaningful Chinese characters. The purpose of this thesis is to set up a database for users with real stroke information, virtual stroke information, point press variation, time between strokes, the dependency of strokes, cluster of strokes, the changes of directions of constant strokes, etc. Meanwhile, this database will use SVN classifier for partial break point to sort out related partial characters with scientific classification method. With the build-up of this database, we can provide a much more systematic stroke teaching method to help foreigners and foreign spouses to learn how to write Chinese characters more efficiently. |
