隨著台灣產業投入於技術的研發資源增加，近幾年台灣於美國專利商標局（USPTO）獲得之專利數也呈正比例上升；然而投入研發資源之「量」的增加以及獲得的專利數上升，並不代表研發成果「質」的等比例提升。這一點可從台灣企業及政府研發單位所獲得的專利，其國際專利引證率遠低於平均值，可見一般。 究其原因，一部分的問題在於台灣的企業（及政府研發單位）於最開始各相關資訊的搜集上不夠完全以及沒有一套良好的開發評估程序供其依循，這使得研發方向及標的的確立上有所偏差，而導致研發結果與商業需求有落差。 基於上述背景及動機之下，本研究的重點在於確立一套資訊收集及分析評估流程，以幫助台灣的企業（及政府研發單位）有效降低研發成本、提升研發方向的正確性，更進一步做為其決定研發策略的參考依據。 本研究之目的，即在於以產業分析、技術分析、專利地圖分析三個具邏輯順序性的外部分析為主軸，並結合組織本身核心技術、網絡關係等與組織相關之分析，來對組織的未來研發標的作一評估及策略建議。 本研究最後得到了以下的發現： 1. 市場□技術□專利之研發評估程序具有分析邏輯性。 2. 「市場需求研究」所找到的產品「需求功效」，可推導出達成功效可能的技術，並可藉此確立出專利分析時需要的「技術項目」。 3. 「市場需求」研究所找到的產品「需求功效」，可推導出達成「功效」所需要的「技術項目」，再以此去進行專利地圖的「技術/功效」二維關聯矩陣分析時，技術軸面項目與功效軸面項目兩者因具有因果性，故其相交關聯性極易判斷，也很容易明確定義出哪一個技術/功效相交關聯洞是有意義的、是可實際進行研發的。 4. 以產業市場□技術□專利三程序所完成的的「技術/功效」二維關聯矩陣圖，雖可容易的定義出哪一個技術/功效相交關聯洞是有意義的、是可實際進行研發的，但「哪一個可實際進行研發的技術/功效相交關聯洞具有較高關鍵性、重要性及有較高開發價值？」仍需藉由對技術及產業價值鍊熟悉的專家意見才可做出判斷。 5. 此方式組成的技術/功效二維關聯矩陣不但可看到目前已知的技術/功效相關聯部份（如本研究中的主要關鍵技術與主要關鍵功效相交處）專利佈局狀況，也可看出在未來的技術/功效相聯部份（如本研究中的主要輔助關鍵技術與輔助關鍵功效相交處）專利可能發展方向。 6. 產業市場分析、技術分析、專利地圖分析等三個組織外部分析須再配合組織原有的核心技術與網絡關係的評估，才可正確制定出組織未來研發的路徑及作為。 / Along with the enterprise and government investment increasing to the technical research ＆ development（R＆D） resources, Taiwan obtains the US patent number also to present the proportional rise in recent years; however, the investment of research ＆ development resources and the granted patent number are not proportional to the quality of research ＆ development achievement. This can be proved by the lower patent citation rate of Taiwan enterprises and the government research ＆ development institutes than the international mean value. Investigates its reasons, part of questions is that most of the enterprises (and government research ＆ development institutes) do not collect enough market information and do not have a set of good research ＆ development assessment procedure while the beginning of R＆D project in Taiwan. This causes that the research and development direction and target has a deviation, and causes that the research and development result and the commercial demand have the dropping variance. Based on the above background and the motive, this research key point lies in establish a flow of information collections and the analysis appraisal. It also helps Taiwan’s enterprises and government’s R＆D institutes effectively reduce the research and development cost, and promote research and development direction accuracy, further help for making the decision of research and development strategies. The goal of this research, namely lies in take the industrial analysis, the technical forecast, the patent map study three logic order exterior analysis as a main axle, and combine itself core technology and the network relations of organization to suggest organization's future target of research ＆ development. By this research, we discover 1. The R＆D appraisals procedure of market-technology-patent has analysis logic. 2. The marketing research finds the demands of consumers, and the demands of consumers may infer to the functions of products, and the functions of products may infer to the possible technology sources. 3. The possible technology sources may form the technical items of patent analysis matrix, and the functions of products may form the functional items of patent analysis matrix. 4. The two-dimensional technical/functional patent matrix completed by the market-technology-patent three procedures can easily define which technical/functional intersectional connection hole can be affiliated; however, to judge which technical/functional intersectional connection hole is more important still needs the advice of experts familiar to the technology and the industrial value chain. 5. The three organization exterior analyses of market analysis, technical analysis and patent map study have to coordinate the organization original core technology and the network relations appraisal.

競爭是現代社會中無所不在的行為，國家或企業透過產業競爭分析、企業競爭分析，乃至市場分析及技術預測（Technological Forecast），才能知己知彼並且擬定正確的決策。對科技產業而言，若企業無法隨時掌握技術發展的趨勢，儘早投入技術研發或調整企業的經營策略。不久之後，市場便會被其他競爭對手所佔據。所幸，沒有一項技術發明是直接由發明者的腦袋直接跳到廣泛應用的境地。其間總是經過好幾個連續階段，每一個階段都使得「實用性」及「有用性」更成熟。因此若能掌握科技發展的脈絡，早期投入研發，便能維持企業的競爭優勢。 專利資訊可以用來評估與預測技術發展、規劃研發或技術發展項目、避免誤觸專利權而浪費研發資源、掌握企業發展動向及市場需求。許多企業和政府機關已經發現專利分析的重要性，並且投入大量的人力、物力來進行專利分析的工作。然而，專利的申請日和公開日之間還是存在至少18個月的時滯，若企業過渡倚賴專利資訊的分析，容易使後續的研發資源投入競爭激烈的技術紅海當中。因此若要充分掌握前瞻技術發展的脈動，基礎研究趨勢分析相對於專利趨勢分析，其重要性有過之而無不及。 在分析方法方面，現存的書目記錄以科學與技術類佔大多數，因此，以書目計量學為工具，自然成為研究「科學」技術整體發展的主流。除了傳統的計量分析之外，利用自動化的方法，挖掘大量文件中的隱含及有用知識，也是最近熱門的研究議題。對探勘技術而言，關聯分析、分群、預測等探勘技術，也漸漸成為技術預測不可或缺的工具之一。 過去曾有眾多的研究利用書目計量來分析學術論文或專利資訊，而最近幾年則陸續出現利用文字探勘來分析學術論文或專利資訊，但這樣的分析結果是片段而不完全的。本研究提出整合性的概念，同時結合計量分析（Bibliometrics）與文字探勘（Text Mining）兩種方法，分別對學術論文（Science Citation Index Expanded）與專利資訊（Derwent World Patents Index）這兩種文獻資料作分析，透過互相比較來瞭解技術發展的趨勢。除此之外，也希望透過個案分析，對本研究所提出之方法論本身，探討之間的關聯性。 在選擇個案方面，奈米科技是目前最熱門的科技產業發展方向。其中最具代表性的產業即「半導體產業」和「微機電系統產業」。蝕刻（Etching）製程與設備技術的良劣，直接影響晶圓產品良率的高低，是影響奈米科技未來發展的重要技術之一。故蝕刻技術之發展趨勢值得深入研究。 而本研究之研究結論如下： 1.技術趨勢分析在層次上宜由遠（計量分析）而近（文獻探勘），理論（論文資料）與應用（專利資料）應並重，分析結果才能互補長短。 2.科學發展與市場需求為專利技術生命週期的領先指標。 3.科學發展增加技術商業化的應用，但市場需求則強化了創新擴散的效果。 4.蝕刻技術的基礎研究目前處於成熟期，而技術發展目前處於成長期。在電子產品輕薄短小、高效能的需求下，預期蝕刻技術將持續被商業化應用。 5.蝕刻技術的領先地位，美商已逐漸被亞洲企業所取代，尤其近來南韓的半導體廠商最為積極。台灣的台積電和聯電過去已累積雄厚的技術發展基礎，惟台灣在基礎研究與產學合作方面仍待加強。

全球行動電話用戶數於2002年達到12億，首度超過固定電話用戶數之11億；行動電話用戶數並於2008年達41億，為固定電話用戶數(13億)之3倍以上。行動電話相對於固定電話之主要優勢在於系統之建置成本低及佈建速度快；行動電話之快速普及已成為創新擴散研究之重要題材。 行動電話擴散之研究為選取一成長模型(例如Gompertz、Logistic或Bass模型)並類比該模型以求出擴散之參數(例如成長速率)，以進一步(1)了解相關驅動因子(例如技術創新、市場開放等)對擴散參數之影響，及(2)延伸擴散模型曲線以預測未來之成長。 惟成長模型之選取尚無原理原則可供遵循而具隨機性(ad hoc basis)。為找出模型選用之可能規律，以降低模型選用之隨機性並提高成長預測之準確度，本研究以十二個代表性國家(巴、中、法、德、印、日、韓、俄、星、台、英、美)至2007年之資料以比較三個最常用之成長模型之績效，即Gompertz、Logistic及Bass模型。模型績效逐年比較標準係採用rmse值，並輔以Friedman test檢測模型績效差異之顯著性，再對照模型之機制意涵，以進一步了解最適模型之選用原則。 此外，台灣行動電話普及率於2002年為108%居全球之冠，而中國自2001年起取代美國成為全球具最多行動電話用戶數之單一國家，台灣及中國屬行動電話擴散之重要個案，惟目前尚缺此二個案之實證研究。為補足此一缺口，本研究亦對台灣及中國行動電話擴散之驅動因子進行實證研究，以進一步了解擴散之關鍵驅動力。 研究發現由於目前統計軟體之進步，Gompertz、Logistic及Bass三模型均可獲致極佳之匹配度而難分軒輊，惟模型預測力(延伸曲線)則具差異性。12個模型選用樣本國家中之8個國家(巴、中、法、德、日、韓、英、美)係以Gompertz模型具較佳之預測力；依Gompertz模型機制意涵，代表行動電話擴散早期係受網路外部性(口耳相傳)影響，惟至擴散後期(例如過了擴散極大值之一半)則已不相關。此外，若因市場開放等重大變因造成行動電話之快速擴散，則Logistic模型具有較佳之績效，如台灣及俄羅斯屬之。依Logistic模型機制意涵，代表擴散係受網路外部性所影響。Bass模型應用於行動電擴散時，因該模型所算出之創新係數偏低，績效與Logistic模型相近，而Logistic模型為Bass模型之創新係數為0時之特例。 台灣及中國行動電話擴散之驅動因子研究發現(1)價格下降及(2)預付卡之推行對加速擴散具顯著性，兩者均對低階市場之採用具影響力。鑑於高階市場將先飽和，爰未來加速行動電話擴散之關鍵驅動因子應係與推動低階市場採用具密切相關性。以中國為例，未來市場開放競爭造成價格再度大幅下降，將進一步促低階市場採用，加速中國行動電話之普及。 / The number of mobile telephone subscriptions reached 1.2 billion globally in 2002, exceeding fixed-line telephony subscriptions (1.1 billion) for the first time. The number of mobile telephone subscriptions reached 4.1 billion globally in 2008, over three times the number of fixed-line telephone subscriptions (1.3 billion). The main advantages of mobile telephony over fixed-line are low cost and rapid facility deployment. The rapid diffusion of mobile telephony has become an important topic in innovation diffusion. The conventional approach to studying mobile telephony diffusion is to analogize a single growth model, such as the Gompertz, Logistic or Bass model, and calculate the model parameters, for example growth rate. The significance of certain selected driving forces, such as technology innovation or market competition, to the studied parameters, such as growth rate, is then estimated. The diffusion growth can also be forecast by extrapolating the diffusion curve. Utilizing the growth model analogy is the first step in analyzing mobile telephony diffusion. However, no principles or rules exit for selecting a growth model. To identify rules for model selection to reduce randomness and increase forecast accuracy, this work uses 12 sample countries, namely Brazil, China, France, Germany, India, Japan, Korea, Russia, Singapore, Taiwan, the UK and the USA, employing data prior to 2008 to compare the performance of three most commonly used models, namely the Gompertz, Logistic and Bass models. The root mean square error (rmse) is chosen as the criterion for measuring annual model performance. The work uses the Friedman test to examine the significance of differences in performance between models. The implications of model mechanisms are emphasized to investigate the selection rule for the most appropriate model. The penetration of mobile telephony in Taiwan was 108% in 2002, ranking first in the world. Furthermore, in 2001 the number of mobile telephony in China replaced the United States as number one in the world. Both Taiwan and China are important examples for mobile telephony diffusion. However, no empirical investigation has been performed in these two cases. To fill this gap, this work estimated the driving forces for mobile telephony diffusion in Taiwan and China to learn about the critical drivers of the mobile telephony diffusion. Empirical results indicate that due to improvements in statistical software, providing good fitness for all three models, namely the Gompertz, Logistic and Bass models, distinguishing which has the best fitness is difficult. However, the performance of the three models is distinguishable when forecasting based on extrapolating the diffusion curve. In eight of the 12 examples, namely Brazil, China, France, Germany, Japan, Korea, the UK and the USA, the Gompertz model is the most appropriate model for forecasting. The mechanism of the Gompertz model means that during the initial stage the diffusion is correlated with network externalities (namely word of mouth), however, this correlation reduces during the later stages (such as pass one half of the maximum potential). Moreover, the cases of Taiwan and Russia demonstrated that the Logistic model performs well provided some significant driver of the diffusion exists. The mechanism of the Logistic model means that the diffusion is correlated with network externalities throughout the whole diffusion. Furthermore, using Chinese data, when the Bass model is applied, because of its low innovation coefficient, it performs similarly to the Logistic model, which is a special case of the Bass model in which the innovation coefficient equals zero. Empirical results for the critical driving forces of mobile telephony diffusion in Taiwan and China indicate that (1) reducing prices and (2) the launch of pre-paid services are crucial to mobile telephony diffusion. Both factors are essential to mobile telephony adoption in low-end markets. The high-end market is the first to be saturated by mobile telephony adoption, future drivers of the mobile telephony diffusion should be highly correlated with low-end market demand. Taking China as an example, the opening of the market to further reduce tariffs will attract mobile telephony adoption in the low-end market, facilitating the mobile telephony diffusion.

有鑑於台灣將於2006 年全面轉換現有類比電視系統，進入數位 電視時代，此趨勢亦將帶動數位電視未來的成長與發展，但基於台灣 數位視訊服務與接收設備產品正處於起飛的階段，未來發展的不確定 性仍高，因此其市場商機與未來技術之發展就非常值得進行研究與探 討。故本研究藉由SRI 情境預測法，對台灣未來5 至10 年的有線數 位機上盒(Digital Cable STB)的功能、技術需求與市場狀況做分析預 測，並針對可能發生的情境，提出因應策略，以提供有線數位式機上 盒供應商做未來策略決策時的參考。 本研究係透過包含學界及實務界的專家群會議輔以腦力激盪的 方式循序討論出關鍵決定因素與驅動力量，並以三個不確定軸面形成 情境主軸，發展擴充成為情境內涵，再就各選定之情境(風雲年代、 市場導引、技術革命及夕陽餘暉)內容進行SWOT 及策略發展分析， 並發展出有線數位STB 供應商之市場及技術共通發展策略： 在市場發展策略方面，建議台灣數位有線STB 供應商考慮： 1.形成同業默契，避免掀起價格戰，維持合理利潤。 2.為避免分散力量，造成惡性競爭。應尋求同、異業結盟，擴大業務 規模及市場佔有率，並提高對供應商的議價力量。 3.與TV 業者結盟，推廣HDTV 市場。 4.取得主要CA 供應商授權，共同開發高成長潛力市場。 5.與MSO 業者結盟，合力推動數位視訊市場發展。 在技術發展策略方面，則建議應考量： 1.核心技術之提昇，如CA，寬頻通訊及HDTV 技術。 2.多功能STB 之開發，如DVD, Home Theater, Cable Modem, VoIP 及PVR 等。 3.與台灣半導體業者合作開發關鍵零組件(IC)以降低成本並掌握關鍵 技術。 4.Open Cable 之開發，以因應不同CA 系統之需求。 5.XDSL STB 與無線寬頻技術之研究。 / Taiwan government has formally announced to switch to digital TV program broadcasting from present analog TV program broadcasting system starting year of 2006. At present, digital video service in Taiwan is still in infant stage and with high uncertainty in future d evelopment trend. Therefore, it provides the motivation of studying the market potential and technological development of digital cable industry. This research is based on SRI scenario forecasting methodology to predict future functional, technological needs and the market status of digital cable set-top-box (STB) industry in Taiwan in the coming 5 to 10 years period. Then provide strategic directions to local digital cable STB manufacturers as the reference of strategic decision making. This research will go through a panel discussion with experts from cable industry and with professors’ involvement. To find out the key decision factors and driving forces, then select three uncertainty axes to develop the scenario content. Based on the discussion and voting result to pick up four most possible scenarios (Glory Day, Pull Enforce, Technology Revolution and Beautiful Sunset) for SWOT analysis and common strategies development from market and technology viewpoints. On the market strategies, suggest local digital cable STB manufacturers should consider: 1. Reach consensus with other digital cable STB manufacturers for pricing strategy, in order to maintain reasonable business profit and not fall into price-cutting competition. 2. Strategic alliance with other manufacturers to enlarge the business scale and strengthen the bargain power to suppliers. 3. Allied with TV manufacturers to stimulate the market demand of HDTV. 4. Authorized by major CA provider, co-develop the high potential market. 5. Allied with key MSOs to promote and accelerate digital video market growth. On the technological strategies, suggest local digital cable STB manufacturers should consider: 1. Core technologies enhancement, such as CA, broadband communication and HDTV technologies. 2. The development of multifunctional STB, such as DVD, home theater, cable modem, VoIP and PVR function built in STB. 3. Co-develop the key components (asics) with local semiconductor manufacturer to lower product cost and hold the key technology. 4. The development of open cable to meet different CA system’s requirement. 5. The research of XDSL STB and wireless broadband technology.