近年來,隨著智慧財產權的普及,技術的專利化可以說是企業保存競爭優勢的重要手段。然而專利的重要性,不單單是在於衡量競爭對手有多大的能力;而是可以藉由專利分析了解產業動態,進而制定自身研發團隊的發展方向,以避免浪費資源。 傳統的專利分析大部分著重於專利筆數的計數,或者是引用率、被引用率等分析。不同於一般的專利指標研究,本研究試著以另一個角度,也就是使用文字探勘的方法,來讀取專利文件中的關鍵字,由關鍵字的類似程度來建立專利文件彼此間的相似程度,進而建立專利地圖,並找出專利缺口。 本研究獲得如下結論:利用專利文件的關鍵字擷取,除了可以有效的將專利做一更詳細的比較以及分類外,更重要的是可以由建立出的專利地圖察知專利缺口的存在,並可了解產業中各企業的專利分佈趨勢,藉此除了可以有效地了解產業動態,也可以不需要逐篇閱覽專利文件以減少人力的浪費。
As the popularity of the intellectual property, the patents provide a powerful protection to a firm’s technologies. Furthermore, patents become an important asset to a firm. A firm can understand the competition status of the industry by patent analysis. The goal of patent analysis is not only evaluating the capabilities of competitors, but also avoiding wasting resources on wrong R&D direction. Traditional patent analysis focuses on patent counts, or the forward (backward) citation. This research adopts another technique, TEXT MINING, for analyzing the content of the patent. Use the similarity of the keywords to explore the similarity of the patents under-evaluated, then developing the patent map and exploring technology gap. The research finds that, by using text mining, the patents can be further classified and compared. The more important is to find out the technology gap, and the distribution of competitor’s technology development.