現代的電子文件的量已經巨大地增長，網路科技使用戶獨立地分享信息和知識。語言用來寫文件也有很多種。這種現象引導我們發會放法能精確地檢索文件和以能力解決语言障隘。 本這次研究， 我們發會相似度放法用來從論文和學術論文系統檢索雙語科學文件。我們計算雙語文件相似度(漢語和英語)。 結合一個检索系统以能力解決语言障隘是富挑戰性任務。 在我們的研究的每個科學文件被劃分成4個領域：標題、主題詞、摘要和被援引的參考。要計算每個領域相似度我們使用一個不同的演算法。我們的方法學的結果表示，我們的方法學能準確地檢索雙語文件。 Electronic documents have grown tremendously in quantity nowadays, the internet technology enable users to share information and knowledge independently. The language which is used to write the documents might also variant. This phenomenon has leads us to develop a methodology which can retrieved documents precisely and with the ability to solve language barrier. In this research we develop a similarity-based methodology to retrieve bilingual scientific documents from Theses and Dissertation System. We compute the similarity of bilingual documents (Chinese and English). Integrated a retrieval system with the ability to solve language barrier is a challenging tasks. Every scientific document in our research is divided into 4 fields which are: Title, Keyword, Abstract, and Cited Reference. To compute a similarity of every field we used a different technique. The result of our methodology shows that our methodology is able to retrieve bilingual documents accurately.
