應用網絡與主題分析法探析智慧製造之研究發展;Applying Network and Topic Analysis on Smart Manufacturing Research

NCUIR > School of Management at National Central University > Graduate Institute of Business Administration > Electronic Thesis & Dissertation > Item 987654321/85837

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/85837

Title:	應用網絡與主題分析法探析智慧製造之研究發展;Applying Network and Topic Analysis on Smart Manufacturing Research
Authors:	張黎琴;Chang, Li-Chin
Contributors:	企業管理學系
Keywords:	智慧製造;智慧工廠;網絡分析;文獻耦合;主題建模;Smart Manufacturing;Smart Factory;Network Analysis;Bibliographic Coupling;Topic Analysis
Date:	2021-06-08
Issue Date:	2021-12-07 11:32:12 (UTC+8)
Publisher:	國立中央大學
Abstract:	第四次工業革命興起，智慧製造的評估與導入成為全球工業熱絡的討論議題，許多學者專家為了協助製造業推展技術創新與改善已進行了許多相關的研究。為了探析智慧製造研究領域歷年來的學術研究發展情形，本研究從 Web of Science 平台中檢索出近三十年智慧製造相關的期刊文獻，提出以網絡鏈結關係強度與時間鏈結關係為基礎的網絡分析方法進行調查，並提出主題分析法對大量的文獻摘要進行文字探勘。本研究結果發現有關智慧製造相關的研究文獻產量自 2016 年起每年呈現跳躍性的增長，我們探析出五個面向的智慧製造學術網絡集群及五個主要的研究主題。網絡分析的結果，揭示出五大面向的學術網絡集群包括（1）四個熱門的作者協作網絡集群；（2）四個主要的文獻來源出版品網絡集群，研究發展期間最久的網絡類型依序為工程與人工智慧計算領域及產業資訊整合系統與生產製造領域；（3）四個主要的關鍵字共現網絡集群，分別為以系統字詞、物聯網字詞、模型字詞，以及以設計字詞為首的集群；（4）六個文獻耦合網絡集群及其代表性文獻，其中近年來的主要研究網絡分別為探討將社會服務層面納入智慧製造系統、智慧製造垂直整合機制、在工業 4.0 下的製造業及工廠技術發展，以及智慧製造大數據分析技術；（5）五個關鍵研究議題的歷史研究網絡關係及各時期具影響力的文獻，包括整體討論度最高的數據驅動概念研究網絡、近十年討論度最高的深度學習演算技術在產業應用研究網絡等。此外，主題分析的結果則探析歷年來智慧製造文獻的五大研究主題，分別為演算法及學習網絡模型改善製造效率、物聯網網絡安全與通訊技術、工業4.0 工廠科技與管理、監測設備與程序監控，以及智慧化控制系統架構設計。本研究探析出智慧製造相關領域中熱門及隱藏的議題方向及討論內容，可提供企業決策者評估與分析智慧製造相關的生產與投資決策，多面向的探析出智慧製造學術網絡中知識鏈結的過程，有助於技術創新者與學者專家掌握相關研究領域最新的理論與技術的研究發展，進行更進一步的研究與創新。 ;The rise of the fourth industrial revolution prompts smart manufacturing to be an important issue for companies. Many scholars and experts have also researched fields related to smart manufacturing to help manufacturers innovate and improve their production. To further understand the research development of smart manufacturing, distance-based and time-based network analysis approaches were proposed to examine the smart manufacturing related research articles from the Web of Science database in last thirty years. Topic modeling method was also adopted to conduct text mining on the abstract of journal papers to identify the key patterns of smart manufacturing literatures. Accordingly, five types of network clusters and five main research topics related to smart manufacturing were identified. The network analysis results show five different types of major academic clusters, including (1) the four co-authorship network clusters; (2) the four citation-by-sources network clusters, of which the longest time duration was the engineering and artificial intelligence cluster, followed by the industrial information integration system and manufacturing cluster; (3) the four co-occurrence keywords network clusters; (4) the six bibliographic coupling network clusters and the representative literature in each cluster, in which the major research networks in recent years were incorporation of social services into smart manufacturing systems, vertical integration mechanisms in smart manufacturing, technological development of the manufacturing industry and factories in the Industry 4.0, and big data analytics technology in smart manufacturing; and (5) the five historical research network constituted of key research topics and the influential papers in each period. In addition, the results of topic analysis indicate that the five most common research topics among smart manufacturing literature over the years were as follows: researching algorithms and learning network models for improving manufacturing efficiency, Internet of Things network security and communications technology, factory technology and management in the Industry 4.0, monitoring equipment and procedural monitoring, and architectural design of intelligent control systems. The results from this research can help corporate decision-makers evaluate and analyze smart manufacturing production and investment decisions. Besides, technological innovators, scholars, and experts can gain comprehensive insights into the latest theories and research development in smart manufacturing, thereby facilitate further research and innovations.
Appears in Collections:	[Graduate Institute of Business Administration] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	51	View/Open

社群 sharing

Loading...