競爭智慧(Competitive Intelligence)是以合法方式收集競爭對手及產業環境的資料,藉以協助研判企業處境及未來的趨勢變化,並提供決策的參考因素。由於全球化、產業分工、及傳媒發達等競爭前提,企業如何即時消化資訊已是重要課題之一。針對上述問題,本研究以收集網路即時新聞為標的,將新聞的主要涵義解析為預設的分類後,再分派至競爭對手的團隊成員中。三項主要設計包括:(1)新聞文件的分類,利用文字分析擷取新聞的主要概念,再將其歸類到公司底下預設的新聞類別;(2)供應鏈成員組合,發展知識本體(Ontology)達到彈性且快速重組供應鏈之目的,用以掌握對手團隊成員的動態變化;(3)新聞分派,發展推論規則建立新聞與對手團隊成員的關係,用以因應對手團隊成員變化時,可彈性調整新聞分派的對象。為驗證研究設計的效能,本研究以處理RSS新聞並選定電子產業為實驗對象,將近百家公司實例知識建入知識庫中,由實驗結果顯示:以知識庫為基礎的競爭智慧系統,可以更有彈性調整供應鏈結構,以及反映供應鏈成員網絡的變化,進而再反映到新聞分派上。本研究旨在設計一般化的競爭智慧系統模型,更實務性的應用在新聞文件分派上,可藉由增加專門領域的新聞及事實資料,達到協助企業收集外部環境資料,未來只要領域專家持續增加更多的事實知識至知識庫,將可以使本研究的競爭智慧系統更趨完善。
Competitive Intelligence is a legal way to collect data from competitors and businesses. Whereby, it helps to analysis businesses’ situation and future tendency, and provides reference resources when makes a decision. Due to the competitive presuppositions of globalization, divisions of labor, and developmental communications, it’s one of the most important issues how businesses deal with information immediately. According to the above problems, we conduct an aim at collecting real-time news. After analysis the main implications of news as presumptive classifications, then assign to the group of competitors. Three main designs include: (1) Classification of news, using text analysis to generalize the main idea of news. (2) Association of supply chain members, developing ontology to hold opposite members’ dynamic change. (3) Assignment of news, developing inference rules to establish the relation between news and competitors. In order to test and verify the efficiency of research designs, we handle RSS news and choose electronics as experiment. During the experiment: the base of knowledge-based competitive intelligence system can adjust the supply chain structure more flexible, react to the change of supply chain members and response to the assignment of news further. This study has designed generalization competitive intelligence models that apply to the assignment of news more empirical by increase special domain news and factual data to help businesses collect data of external environment.