產業研究是企業掌握產業內提供類似產品或服務公司之情報，乃至於上下游供應鏈與需求鏈的分析研究，所謂「知己知彼，百戰百勝」，產業研究做得好，企業對內能夠更明確地擬定方針，優化營運策略以及提升決策品質，對外亦能了解產業內競爭者的現況與動向，進而採取因應措施甚至打擊威脅。然而，企業之間相較於過往非合作即競爭的關係，現在的企業更偏向「競爭中求合作，合作中有競爭」的競合關係，以此來最大化利潤，最小化成本。
對於新興產業，進行產業研究是勞神傷財的，研究人員定義完問題後，需要找到想要關注的廠商做質性訪談，或耗資購買產業報告，前者點對點的訪談未必能勾勒出產業的整體輪廓，後者也不一定對定義的問題有所幫助。即使是期刊、媒體報導等次級資料，透過人工搜集、篩選、整理、分析須耗費大量時間，也容易發生人為失誤，有鑑於此，本研究以台灣離岸風電為案例，藉由網路爬蟲自動化資料搜集過程，並以自然語言處理以及文字探勘等技術進行資料清理，再透過社會網路圖視覺化呈現產業結構以及廠商間的競合關係，最後提供互動介面網站，使產業研究人員能夠在定義新的問題後，迅速對產業或企業進行探索與評估。

Industrial research has been a vital information for companies to analyze their competitors, supply chains and consumer demand. The more industrial research companies do, the better probability they will grow against the tide. For instance, internally optimizing operating strategies, formulating policies with efficiency and enhancing the quality of decision-making. Externally, knowing the current status and identifying market trends like the back of the hand. However, it is not possible for every company to do the industrial research due to time urgency and the lack of capital. As the industries continue to advance, coopetition, cooperation and competition, has sprung up in the fields. It is coopetition that help economize manpower and material resources. The original approach of doing industrial research might go to defining the problems, then doing qualitative interviews with related manufacturers or spending a sum of money on existing reports. However, the former peer-to-peer interview may not be able to outline the industry, nor the latter which covers a large area be helpful for the solution. Not to mention the secondary data such as journals and media reports. Therefore, coopetition has been rolled out. In the study, taking the offshore wind power industry as an example to demonstrate “coopetition”. First, data can be automatically operated by means of web crawlers. Second, exploring the data by natural language processing, text mining and word embedding. Meanwhile, industrial structure and coopetition relationship can be visualized through the social network diagram. Last, setting up an interactive website to enable the researchers to evaluate industries and companies within the shortest possible time.

論文審定書 i
誌謝 ii
摘要 iii
Abstract iv
目錄 v
圖目錄 vii
表目錄 viii
第一章 緒論 1
第一節 研究背景 1
第二節 研究動機 1
第三節 研究目的 2
第四節 論文架構 2
第二章 文獻探討 3
第一節 台灣離岸風電發展概況 3
第二節 競合理論 4
第三節 文字探勘技術 5
第四節 社會網路分析 7
第五節 詞嵌入模型 8
第三章 研究方法 11
第一節 研究對象 11
第二節 研究架構 11
第三節 研究步驟 12
第四章 結果與討論 18
第一節 文集基本統計 18
第二節 資料預處理 23
第三節 命名實體識別 24
第四節 共現矩陣與相關矩陣 25
第五節 互動式社會網路圖 27
第六節 探索產業關係 29
第七節 結果驗證 32
第八節 個案分析與討論–以宏華營造為例 39
第九節 研究限制 43
第五章 結論與建議 45
第一節 結論 45
第二節 建議 45
第三節 研究貢獻 46
參考文獻 47

Barnes, J. A. (2016). Class and Committees in a Norwegian Island Parish: Human Relations. https://doi.org/10.1177/001872675400700102
Bengio, Y., Ducharme, R., & Vincent, P. (2001). A Neural Probabilistic Language Model. In T. K. Leen, T. G. Dietterich, & V. Tresp (Eds.), Advances in Neural Information Processing Systems 13 (pp. 932–938). MIT Press. http://papers.nips.cc/paper/1839-a-neural-probabilistic-language-model.pdf
Brandenburger, A. M., & Nalebuff, B. J. (1997). Co-Opetition (1 edition). Currency Doubleday.
Company Relation Extraction from Web News Articles for Analyzing Industry Structure—IEEE Conference Publication. (n.d.). Retrieved September 26, 2020, from https://ieeexplore.ieee.org/document/7889512
Global Wind Speed Information Database | 4C Offshore. (n.d.). Retrieved July 23, 2020, from https://www.4coffshore.com/windfarms/windspeeds.aspx
Kao, A., & Poteet, S. R. (Eds.). (2007). Natural Language Processing and Text Mining. Springer-Verlag. https://doi.org/10.1007/978-1-84628-754-1
Mérő, L. (1998). John von Neumann’s Game Theory. In L. Mérő (Ed.), Moral Calculations: Game Theory, Logic, and Human Frailty (pp. 83–102). Springer. https://doi.org/10.1007/978-1-4612-1654-4_6
Pennington, J., Socher, R., & Manning, C. (2014). Glove: Global Vectors for Word Representation. Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), 1532–1543. https://doi.org/10.3115/v1/D14-1162
RE100. (n.d.). Retrieved August 17, 2020, from https://www.there100.org/
Shiny. (n.d.). Retrieved July 24, 2020, from https://shiny.rstudio.com/
Text2vec. (n.d.). Retrieved August 16, 2020, from http://text2vec.org/
Travers, J., & Milgram, S. (1969). An Experimental Study of the Small World Problem. Sociometry, 32(4), 425. https://doi.org/10.2307/2786545
trinker/entity: Named Entity Recognition version 0.1.0 from GitHub. (n.d.). Retrieved July 23, 2020, from https://rdrr.io/github/trinker/entity/
UDPipe | ÚFAL. (n.d.). Retrieved July 23, 2020, from http://ufal.mff.cuni.cz/udpipe
Wang, Y., Liu, S., Afzal, N., Rastegar-Mojarad, M., Wang, L., Shen, F., Kingsbury, P., & Liu, H. (2018). A Comparison of Word Embeddings for the Biomedical Natural Language Processing. ArXiv:1802.00400 [Cs]. http://arxiv.org/abs/1802.00400
Wellman, B., & Berkowitz, S. D. (Eds.). (1988). Social structures: A network approach (pp. xii, 513). Cambridge University Press.
World’s largest semiconductor foundry joins RE100 and calls on peers to step up on renewable energy. (2020, July 27). The Climate Group. https://www.theclimategroup.org/news/world-s-largest-semiconductor-foundry-joins-re100-and-calls-peers-step-renewable-energy
林頌堅（2010）。以詞語共現網絡分析探勘資訊傳播學領域的研究主題與關係。取自http://catweb.ncl.edu.tw/literature/page/29684
王建平&吳曉云（2019）。競合視角下網絡關係強度、競合戰略與企業績效。取自http://www.cnki.com.cn/Article/CJFDTotal-KYGL201901012.htm
郭家瑋（2017）。離岸風電產業開發商之競爭策略分析。國立台灣大學土木工程學研究所碩士論文, 台北市。取自https://hdl.handle.net/11296/p435rk