中文詞彙屬於一種開放集合,目前並無一個可將所有中文詞彙收入的詞典或方法。分詞系統最常因為參考詞彙不足而產生錯誤的詞彙切分,解決此問題最有效的方式就是補充各領域的詞典,以加強詞彙切分的正確性。本研究將輸入劇本文件,應用關鍵詞彙的詞彙切分技術方法,先將分析出來之關鍵詞彙轉換為分鏡腳本形式,然後再搜尋出與分析文件相關場景與場景組件圖檔,並以此建構出分鏡場景,以利編劇與腳本設計家可以利用此分鏡腳本與分鏡場景做討論對照並做劇本修正。系統將分為三大部分:第一部分為資料管理系統,將建立場景、時間、人物、場景物件等分析劇本所需要之詞彙庫資料庫;分鏡場景建構系統所需要的場景與場景物件圖片資料庫。第二部分為詞彙切分系統,在詞彙切分處理上利用詞庫中的「詞彙使用次數」、「詞彙長度」與「詞彙出現次數」為演算法基本函數,以提升正確率,並以此結果建立分鏡腳本資料。第三部分為分鏡場景建構系統,將相關場景圖檔建構出分鏡場景。
Chinese vocabulary is a kind of open-collective; currently there is not a dictionary or a method that can acquire all Chinese terms. A segmentation system often results in faulty vocabulary segmentation due to insufficient word bank. The most effective way to solve this problem is to supply dictionaries of different fields so that the vocabulary segmentation can be properly done. This study uses the keyword method of vocabulary segmentation on the imported script text. The analyzed keyword is first converted to the shooting script -writing framework; the word is then used in searching and analyzing for the image files that are relative to the document, thus the shooting script is constructed in the process. The screenwriter and the scriptwriter can make use of the shooting script to discuss and compare it with the scenario and make revisions to the script. The system consists of three major parts: First is the data management system. Second is the vocabulary segmentation system. Third is shooting scene construction system.