本文首先透過所建立之雲端運算專利檢索公有雲，以LED水族燈研究之載具，由創新性改良之搜尋法檢索出LED水族燈之繁體和簡體中文及英文專利文件，並進行LED水族燈繁體和簡體中文及英文專利的IPC分析。此外本文改良繁體和簡體中文專利之零組件元件字的搜尋及斷詞方法。由繁體和簡體中文專利的文件之內容，歸納出兩種創新零組件元件字搜尋法。本文由繁體和簡體中文及英文專利語意分析及斷詞斷字系統，由新增加第二層技術的技術詞句斷詞點半自動半人工歸納出相關繁體和簡體中文及英文專利之技術字群與功能字群。並應用所搜集的LED水族燈相關繁體和簡體中文及英文專利，收集同義字，建立繁體和簡體中文及英文同義字對照表。接著建立LED水族燈相關繁體和簡體中文及英文專利之第一層及第二層技術/功能矩陣，並進行以技術觀點為主的LED水族燈繁體和簡體中文及英文專利分析。本文並應用判斷一件專利所歸屬的技術或功能類別之方法。此方法為先利用上述建立的繁體和簡體中文及英文同義字表，配合元件字、技術字群及功能字群之關鍵字之出現次數，建立關鍵字相關的模糊值，進而應用次數常態化概念，建立相關公式，算出該繁體和簡體中文及英文專利可能歸屬之技術或功能類別的概率值。
接著本文以完成整理之LED水族燈的繁體和簡體中文及英文專利進行IPC分析、發明人分析、公司別分析、所屬國別分析以及整體趨勢分析。並由繁體和簡體中文及英文專利之元件字、技術詞和功能詞的相關斷詞及斷字分析，建立技術功能類別之字群，進而產生第一層及第二層技術功能矩陣。本文進而由第一層及第二層技術/功能矩陣建立第一層及第二層各分項技術與公司別、核心競爭力公司分析圖、競爭者研發活動力趨勢圖及技術層活動力趨勢圖。最後本文實際投入兩篇相關專利驗證其歸屬之第一層及第二層技術類別，證明本研究之方法為可行。上述所搜尋的相關LED水族燈的繁體和簡體中文及英文專利，以及前述之IPC分析和技術分析之結果，可提供給企業及工程師作為參考，以節省研發時間，作為新專利產生之參考，此外亦可使企業了解競爭者之核心競爭力。

First of all, the paper retrieves public cloud through the established cloud computing of patents, takes the carrier of LED aquarium light research, uses the innovatively improved search method to retrieve the traditional and simplified Chinese and English patent documents of LED aquarium light, and carries out inter-process communication (IPC) analysis of the traditional and simplified Chinese and English patents of LED aquarium light. Besides, the paper improves the search and term segmentation methods of the words of parts and components of the traditional and simplified Chinese patents. From the contents of the traditional and simplified Chinese patent documents, the paper induces two innovative search methods of the words of parts and components. From the semantic analysis as well as term and word segmentation system of the traditional and simplified Chinese and English patents, and from the term segmentation points of the technical terms of the newly increased Level 2 techniques, the paper semi-automatically and semi-manually induces the technical word group and functional word group of the related traditional and simplified Chinese and English patents. The paper also applies the searched relevant traditional and simplified Chinese and English patents of LED aquarium light, and collected related synonyms to establish a List of Traditional and Simplified Chinese and English Synonyms. After that, the paper establishes Level 1 and Level 2 technical/functional matrices of the related traditional and simplified Chinese and English patents of LED aquarium light, and carries out analysis of the traditional and simplified Chinese and English patents of LED aquarium light that mainly take technical points of view. The paper also employs the method of judging of which technical or functional category a patent belongs to. This method firstly uses the List of Traditional and Simplified Chinese and English Synonyms established above, together with the appearance frequency of the keywords of component words, technical word group and functional word group to establish related fuzzy values of the keywords. Furthermore, applying the concept of frequency normalization, the paper establishes the related equations to calculate the probability value for a traditional or simplified Chinese or English patent to belong to which technical or functional category.
After that, the paper conducts IPC analysis of the well-arranged traditional and simplified Chinese and English patents of LED aquarium light, analysis of inventors, analysis of company types, analysis of country types and analysis of the overall trend. From the related term and word segmentation analysis of the components, technical terms and functional terms of the traditional and simplified Chinese and English patents, the paper establishes the word group of technical and functional categories, and further produces Level 1 and Level technical/functional matrices. Furthermore, from Level 1 and Level 2 technical/functional matrices, the paper establishes Level 1 and Level 2 item techniques and company types, analysis diagram of core competitive companies, diagram of competitors’ research and development (R&D) energy trend, and diagram of technical-level energy. Finally, the paper practically puts in two documents of related patents and judges them to be of Level 1 and Level 2 technical categories, thus proving that the method developed by the study is feasible. The related traditional and simplified Chinese and English patents of LED aquarium light being searched above, as well as the results of abovementioned IPC analysis and technical analysis can provide corporates and engineers a reference for saving their R&D time, and a reference for invention of patents. Besides, with these analytical results, corporates are able to understand the core competitiveness of their competitors.

1.杜家宏，「整合繁體和簡體中文及英文斷詞斷字系統之發光二極體檯燈專利分析研究」，碩士論文，國立台灣科技大學機械工程學系，台北(2011)
2.Lawrence, S., and Giles, C. L., (1999), Searching the World Wilde Web General and Scientific Information Access, IEEE Communication, 37(1):116-122.
3.Trappey, C. V., Trappey, A. J. C., and Wu, C. Y., (2010), Clustering patents using non-exhaustive overlaps”, Journal of Systems Science and Systems Engineering, 19(2):162-181.
4.Trappey, C. V., Trappey, A. J. C., and Wu, C. Y., (2009), Automatic patent document summarization for collaborative knowledge systems and services, Journal of Systems Science and Systems Engineering, 18(1):71-94.
5.Singhal, A., and Salton, G.., (1993), Automatic Text Browsing Using Vector Space Model, Technical Report, Department of Computer Science, Cornell University, U.S.A..
6.Gruber, T., “Ontolingua: A translation approach to portable ontology specifications,” Knowledge Acquisition, pp.199-200,(1993)
7.Oleary, D. E,“Enterprise knowledge management,” Computer IEEE, Vol.31, pp.54-61 (1998).
8.拓墣產業研究所，2007深入剖析LED產業市場發展，拓墣科技，台北市(2007)
9.陳哲宏、陳逸南、謝銘洋、徐宏昇，專利法解讀，月旦出版公 司，第14-15頁，民83年。
10.「國際專利分類表第七版使用指南」，經濟部智慧財產局，民90年，7月。
11.「國際專利分類之研究」，經濟部中央標準局。
12.Overview of the Classification System, http://www.uspto.gov/main/patents.htm
13.The EPO guide to patent information on the internet, http://epart.epo.org/dwl/espacenet_manual.pdf
14.Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 7296 LNCS , pp. 370-376
15.MIPRO 2010 - 33rd International Convention on Information and Communication Technology, Electronics and Microelectronics, Proceedings , art. no. 5533317 , pp. 344-349