從專利引文網絡分析探討國家關鍵技術和技術生命週期：以台灣為例

Identifying Essential Technologies And Mapping Technology Life Cycle By Patent Citation Network Analysis: A Case Of Taiwan

卓大順(Ta-shun Cho)

指導教授 ： 施信佑

國立暨南國際大學/管理學院/國際企業學系/博士(2012年)

https://doi.org/10.6837/NCNU.2012.00116

摘要

一個國家致力於追求技術競爭優勢，能掌握其關鍵技術和技術生命週期必定是制定該國家技術戰略和政策之至關重點。許多世界知名的研究機構，如國際管理發展學院（IMD）和世界經濟論壇（WEF），視專利作為評價一個國家研究與發展地位的關鍵指標。基於網絡分析方法的最新進展，更大幅提升考量以專利引文網絡的視覺圖對專利技術流動的了解。本研究以台灣為例，提出一種實用的定量研究方法，即專利引文網絡分析，來探索找出台灣關鍵技術中之核心技術與新興技術，並比較分析台灣核心技術體系之網絡結構特性，以及監測台灣已發展成熟階段之重點技術領域的技術生命週期(TLC)來進行技術預測。
資料收集則是依據美國專利和商標局（USPTO）在1997至2008年間，專利授予台灣為樣本，並以國際專利分類(IPC)基礎用來代表專利技術領域。據資料分析顯示，台灣產出的專利數量為42,650，其專利引用數則是282,714，包括引用與被引用台灣專利。與此同時，在電力學和物理學的二大核心專利技術領域各佔台灣大多數的專利數量分別為22,420（52.6％）和10,447（24.5％），而其個別技術體系之專利引用數則分別為148,299和47,205。本研究方法以網絡分析指標包括密度、集中化、中心度、結構洞、派系和區塊模型，分別考慮在台灣專利引文網絡分析裡以顯示出精確的網絡模式。從網絡的角度來看，首先找出台灣主要核心技術和新興技術，並進一步比較台灣電力學和物理學的二大核心技術體系之網絡結構，最後透過監測台灣發展最成熟技術體系裡之電力學領域和其前十大技術子類(subclass)之技術生命週期，藉此優化台灣政府、產業與企業之技術投入與有效整合資源配置，並強化台灣核心技術體系之網絡結構以利於全球科技競爭優勢，且經由監測技術生命週期的發展階段與專利數量，可建議一個技術體系之技術預測與資源分配與佈局。
本研究乃屬探索性研究，採用專利引文網絡分析方法，其目的為找出台灣五個核心技術和五個新興技術之關鍵技術領域，並比較分析台灣在電力學和物理學二大核心技術體系之網絡結構特性，以及觀察台灣重點技術領域之電力學(electricity)與前十大技術子類(subclass)之技術生命週期(TLC)，以利於概觀台灣整體國家之技術政策與創新發展。本研究成果與貢獻可明白指出，在台灣政府方面，可明白提供追求最適當的技術和創新管理制度，並做出最好的專利戰略和技術政策。在科研院所，產業界和學術界方面，也能建議他們在技術研發的選擇上，哪些是應投入資源，給予正確研究方向，以促進產生更明智的，更高質量的專利，利於增進台灣關鍵技術的發展與創新，並強化台灣技術體系的結構完整性，以及提供一個技術體系之技術預測發展模式。

關鍵字

核心技術 ； 新興技術 ； 技術體系 ； 技術生命週期 ； 技術預測 ； 專利引文網絡分析 ； 台灣

並列摘要

Identifying essential technologies and mapping technology life cycle is critical for formulating national technology strategies and policies for pursuing technological competitive advantage. Many world-known research institutes, such as International Institute for Management Development (IMD) and World Economic Forum (WEF), view patents as a key indicator in evaluating a national research and development (R&D) position. Recent advances in network analysis methodologies greatly enhance the understanding of patent technology flows by considering a network view of patent citations. A case of Taiwan, this study presents a useful quantitative method, namely patent citation network analysis, for identifying core technologies and emerging technologies, and comparing the network structures of core technological system, and monitoring the technology life cycle (TLC) in the maturity stage of the Taiwan essential technologies fields for the national focused technological forecasting.
Data collection included United States Patent and Trademark Office (USPTO) patents issued to Taiwan during 1997-2008 and citations to Taiwan patents held. The International Patent Classification (IPC) categories were used to represent patent technology fields. Then data analyses showed that the number of Taiwan patents was 42,650, and the number of patent citations was 282,714, which links to Taiwan patents, including the cited patent and the citing patent. Meanwhile, electricity and physics technology field’s patents occupy mostly the number of Taiwan patents respectively was 22,420 (52.6%) and 10,447 (24.5%), and the number of patent citations within individual network respectively was 148,299 and 47,205, including the cited patent and the citing patent. The network indicators of density, centralization, degree centrality, structural holes, cliques and blockmodeling considered in the Taiwan patent citation network analysis revealed precise network patterns; moreover, a network perspective was applied to firstly identify core technologies and emerging technologies, and further compare the two core network structures of electricity and physics technological system, and monitor in the maturity stage of the Taiwan electricity technological system and analysis top 10 subclasses of the technology life cycle in electricity of Taiwan in order to optimize technology investment and the effective resource allocation and to strengthen the network structure of Taiwan's core technological system for competitive advantage in global technology and to monitor the development of technology life cycle from the number of patents to suggest technological forecasting and resource allocation distribution.
This study is an exploratory research, the purpose is to gain an overview of technology policy and innovation development in Taiwan by analyzing patent citation networks and by identifying five core technologies and five emerging technologies and by comparing the network structures of electricity and physics in the two core technological systems and by mapping the electricity technological field and top 10 subclasses of the technology life cycle in Taiwan. The findings and contributions indicate the most appropriate management of technology and innovation and the best patent strategy and technology policy that the Taiwan government should pursue. Research institutes, industries and academia are also given research directions for choosing the technologies in which they should invest resources and can foster to generate more informed, higher quality patents in order to strengthen the Taiwan essential technologies development and innovation and to strengthen the network structural completeness of Taiwan’s core technological system and to provide a model of the technological forecasting in the technological system.

並列關鍵字

Core technology ； Emerging technology ； Technological system ； Technology life cycle ； Technological forecasting ； Patent citation network analysis ； Taiwan

參考文獻

1.Achilladelis, B. (1993). The dynamics of technological innovation: The sector of antibacterial medicines. Research Policy, 22, 279–308.

2.Achilladelis, B., Schwarzkopf, A., and Cines, M. (1990). The dynamics of technological innovation: The case of the chemical industry. Research Policy, 19, 1–34.

3.Alba, R.D. (1973). A graph-theoretic definition of a sociometric clique. Journal of Mathematical Sociology, 3, 113–126.

4.Alba, R.D. (1982). Taking stock of network analysis: A decade’s results. Research in the Sociology of Organizations, 1, 39–74.

5.Akamatsu K. (1962). A historical pattern of economic growth in developing countries. Journal of Developing Economies, 1(1):3–25, March–August.