| Abstract: | 本研究之目的為透過分析導向模式、語意解析、以及RosettaNet與ebXML的個案研究來綜合考量跨國企業的商業、技術、與策略等方面的因素，為企業與企業間整合性電子商務，建立一套確實、有效率、理論、與實務並重的架構和方法。本研究架構希望藉由闡明企業與企業間整合性電子商務的塑模方法，對於學、研、產界而言，將在理論建立和實務運用方面有突破性之探討。 電子商務是跨領域和全球性之商業與科技集合體，大多數企業或多或少都面臨企業流程之異質與商業模式的協同的問題。由於多元的商業網站、多媒體的頁面、網路服務、資訊資源、和各國語言等日益分歧的異質網路資源，要建立一個企業與企業間電子商務整合的解決方案實屬不易，而在國際、國家、與產業層次上都是非常重要而有價值的探討。 本研究將建立一個新創觀點，一項以分析導向模式為主的統整、轉換、和本體論的方法，以探討企業與企業間整合性電子商務之科技與策略因素可以如何解決？又如何影響企業間供應鍊整合與協同商務的議題？ 異質性與互動運算是企業與企業間電子商務整合的主要問題，但是一般電子商務標準著重在語法更甚於語意。而複雜的電子商務環境中，商業模式與商業流程並非一致不變，尤其在面對跨國交易夥伴時，商業訊息、詞彙、資料字典、通訊協定、目錄服務等異質和不一致性情形更形嚴峻，是企業與企業間電子商務不可避免要正視的問題。雖然已有許多方法提出為解決資訊整合的問題，然而它們大多缺乏異質整合時的作業流程觀點與協同運作能力。再則，這些方法未能有效以具有系統性與方法性的模式和技術，多半為概念架構之隨意式解決方法，不應以個別特例處理，而應以問題根源，逐項提出結構性解決方案。 因此，本研究提出一項新創觀點，以需求分析導向為主的理論觀點，應用異質性與互動運算和語意網之技術，就問題根源逐項處理企業與企業間整合的作業流程觀點與協同運作能力的困難。本研究提出一項以統一塑模語言、延伸性標籤語言、本體論為基礎的研究方法，建立一套具有系統性與方法性的模式和技術，分析商業流程與資料模式，建立電子商務需求分析和本體論。 在第一年的研究提案中，我們提出一個新創的糢式，表達企業與企業間語意整合，此模式創新之處在於融入和實現處理異質需求的分析技術。首先檢視資料、流程、和控制的塑模，異質資訊整合的根本問題在於需求分析的塑模，也就是先檢視商業流程與資料模式的差異性和不一致性還有變異性；接著我們延伸統一塑模語言來擷取與編輯不同企業的資料與流程的需求和語意，再應用分析結果將其轉換成商業領域的本體論，例如詞彙、概念、關係、與限制、資料字典、通訊協定、目錄服務等。在第二年的研究提案中，我們建立本研究方法的雛型架構和系統，做為研究實作中，驗證本研究模式和技術可行性的衡量和測試工具。我們將實作兩項主要電子商務標準測試本研究，它們分別為：垂直型產業電子商務標準─RosattaNet，水平型產業電子商務標準─ebXML。實作該兩項電子商務標準並作為個案目的是驗證企業與企業間電子商務整合之需求分析可轉換為語法和語意的整合與更新。 在第三年的研究提案中，我們提出如何將需求分析方法轉換成延伸式標籤語言的雙向應對和運算，以達成本研究欲建立一項在能夠在早期並自根源處來解決整合性與異質性的研究目的。因此，我們在第三年的研究提案中，以一項新方法來解決雙向轉換與應對的運算需求，那是使用學名結構導向的觀點，以學名結構基礎進行，分別對統一塑模語言與延伸標籤編碼語言之間的結構和語意進行解析和轉譯。這將使本方法趨於泛用型，可擴充到較廣領域，我們亦將建立一項雛形系統以驗證此方法的可行性和正確性。
The purpose of the proposed study is to articulate a theory and practice oriented model and method for cross national business-to-business integration electronic commerce by exploring an analysis-driven schema and semantics resolutions among business process integration and business model interoperability with emphasis on RosettaNet and ebXML case studies to incorporate the multi-national corporate culture, technology, and strategy. A framework illustrating the ways modeling and method that affects the cross national business-to-business integration electronic commerce would represent an advance in theoretical development and practical insights to scholars, professionals, and executives. Electronic commerce within and across national boundaries is universal. Most firms if not all have problems in one way or another with business process integration and business model interoperability. On both theoretical and pragmatic levels, due to increasing diversity in web pages, services, data sources, and languages in all countries, developing a framework of cross national business to business integration resolution is important at international, national, and intra-national levels. This study will develop a framework and a method to explore the way in which technology and strategy factors influence integration and interoperability over schema and semantics in business-to-business integration electronic commerce. Issues relating to business-to-business integration among the schema and semantics due to business process and business model will be the particular focus of cross national differences in this study. We choose to focus on this issue because it is of vital interest to Taiwan in particular the high-tech and the banking industry. The dynamics of Internet and intelligence of unified modeling language, extensible markup language, and ontology interplay with inter-organizational culture in this issue, making it an excellent basis for exploring the proposed framework and method. Heterogeneity and interoperability are the key issues in business-to-business integration. Electronic commerce standards are usually syntactic rather than semantic. In the complex electronic commerce environment, the business model and business process are not always the same, especially when facing cross national trading partners. This means reviewing the business messages, vocabularies, and data dictionaries is inevitable when establishing each business-to-business connection. Business-to-business integration hence seeks to reconcile the semantic conflicts between messages, protocols, and catalogues. Various approaches have been proposed to study business-to-business integration issues. However, they lack the process perspective and the interoperability with the data integration. They suffer from system and methodology but adhocracy. Much is needed in the systematic and methodological enhancement. We intend to tackle this inadequacy by addressing heterogeneity and interoperability issues in business-to-business integration on a global basis. This research proposes a unified modeling language, extensible markup language, and ontology-based framework that has a set of methods to reconcile and represent the conflicts and correspondences in the business process and business model. In the first year research proposal, we propose a novel approach to address the research issues in the business-to-business semantic integration. We tackle the issues at an early stage. We examine the analysis and modeling difficulties first. We believe the root of heterogeneous information integration lies on the requirements analysis, that is, an analysis of the heterogeneous business process and data schema. We hence adopt and extend the unified modeling language to capture and compile the heterogeneous data and process requirements. We utilize these analysis results and convert them into domain ontologies such as terms, concepts, relationships, and constraints. This method is novel because a new perspective of heterogeneous requirements analysis is incorporated and realized in the study. In the second year research proposal, we propose a prototype of the research method developed will be constructed. The research implementation is created to illustrate the feasibility and validity of the method. We choose two main electronic commerce standards to test our method. One is a vertical electronic commerce standard, RosettaNet. The other is a horizontal electronic commerce standard, ebXML. The prototype is used to run through these case studies and demonstrate the business-to-business integration. The testing in the case studies further illustrates the proof of the syntactic and semantic integration and resolution. In the third year research proposal, we propose to continue the study converting the analysis method into the extensible markup language so that we achieve our final goal of developing one analysis-integrated approach from the beginning and at the early stage. A systematic and methodological approach is created in this three-year study. We adopt a new method to convert and transform the data models between the modeling language and the encoding language. The new approach is based on generic constructs. The orientation of generic constructs provides us with a more generalized conversion method in tackling the transformation between a modeling language and an encoding language. A prototype of this bi-directional transformation will be developed to illustrate the feasibility and validity of the method. |
