| 摘要: | 近年來，企業、社群網站累積了龐大的資料量，企業導入大數據分析(Big Data Analytics，BDA)要能夠成功發揮效益，有賴於結合有效的外部資訊與異質資料(Heterogeneous Data)，才能發揮更大價值。未來，大數據(Big Data，BD)相關之企業與產業上下游組織，必將形成利益共生之大數據產業鏈。大數據要能夠成功支援企業創造優勢，需要具備跨企業的資料處理的能力，因此跨企業的大數據整合與合作是必然的趨勢。本研究將探究大數據產業鏈的應用瓶頸與障礙，探討大數據產業未來可能的商業模式(Business Model，BM)的發展與企業數據商業化的可行性，提出能夠為企業資料創造價值之大數據產業創新商業服務模式的整體解決方案。本研究經初步研究與評估，預計提出下列三種大數據創新的服務模式：(1)虛擬企業服務模式，使企業能夠根據市場機會組成大數據虛擬企業，使企業共享及整合數據、共創資料價值；(2)大數據市集服務模式，提供企業內資料在去隱私化後，透過交易再創價值的機會，也為苦於資料不足者，提供獲取資料的管道；及(3)客製化大數據分析媒合服務模式，對有資料而無分析技術能力之企業，提供客製化大數據分析之媒合服務。企業資料資產化已是趨勢，若能夠建立起具交易公正性及安全性的虛擬式大數據服務平台，就能讓資料的交易更便利。為達上述目標，本計畫的主要研究工作，包含： (1)企業數據應用與技術需求分析；(2)大數據商業化瓶頸、障礙及策略分析；(3) 跨企業資料合作與分享之問題分析與關鍵因素評估；(4)大數據產業之企業資料與知識整合技術研究，(5)跨企業之資料所有權、存取控制、安全、與個人隱私議題探討；(6)大數據交易與合作安全風險評估方法設計；(7)安全多層式雲端化大數據商業服務平台架構設計；(8)跨企業多樣化巨量資料整合技術研發；(9)虛擬式跨企業資料倉儲(Virtual Big Data Warehouse，VDW)與資料超市(Virtual Big Data Mart，VDM)儲存架構設計；(10) VDW與VDM資料模型分析與設計；(11)本體論為基的跨企業多樣化資料整合語意網路模型建構；及(12)服務平台的測試、改善、與實際應用。
Enterprises and social networking sites have accumulated a substantial amount of data in recent years. Successful attainment of the effectiveness and value of big data analytics (BDA) implemented by enterprises depends on its effective integration with external information and heterogeneous data. In the future, BDA enterprises and the related upstream and downstream organizations must form a big data industry chain with shared interests. To support enterprises in creating competitive advantages, BDA must incorporate cross-enterprise data processing; therefore, big data integration and cooperation across enterprises is an inevitable trend. This study will examine the application bottlenecks and obstacles faced by the big data industry chain and investigate the potential development of business models for the big data industry and the feasibility of enterprise data commercialization, thereby proposing a comprehensive solution for the innovative business model of big data industry that can create value from enterprise data. Through preliminary research and evaluation, this study plans to propose the following three big data innovative service models: (1) Virtual enterprise service model: This model enables enterprises to form big data virtual enterprises according to market opportunities; therefore, the enterprises can share and integrate data and cocreate data value; (2) Big data marketplace service model: This model provides enterprises the opportunity of value recreation through transactions after their internal data are deprivatized. It also provides data acquisition channels for enterprises that are troubled by the lack of data; and (3) Customization BDA matching service model: This model provides customization BDA matching service for enterprises that possess data but are in capable of data analysis. Treating enterprise data as assets has become a trend. If a virtual big data service platform that ensures fair and safe transactions can be established, the transactions of data can be more convenient. To achieve this goal, this study will perform the following research work: (1) analysis of enterprise data application and skill requirements, (2) analysis of the bottlenecks, obstacles, and strategies of big data commercialization, (3) analysis of problems concerning data cooperation and sharing among enterprises and evaluation of the key factors involved, (4) research on cross-enterprise data and knowledge integration technology, (5) investigation of cross-enterprise data ownership, access control, security, and privacy issues, (6) method design for the security and risk assessment of big data transaction and collaboration, (7) architecture design of a safe multi-tier cloud-based big data business service platform, (8) research and development of diverse big data integration methods for cross-enterprise application, (9) architecture design of Virtual Enterprise Big Data Warehouse (VDW) and Virtual Big Data Mart (VDM), (10) data model analysis and design of VDW and VDM, (11) construction of an ontology-based semantic network model for integrating diverse data across enterprises, and (12) prototype service platform testing, improving, and application. |
