臺灣食品大數據風險分析──以社會網絡分析探討油品之交易流向管理為例

Big Data in Food Safety-Social Network Analysis for Oil Supply Chain

大數據的整合分析為政府推動國家發展極為重要的方向，也是食品藥物管理署（下稱食藥署）正面臨的關鍵挑戰。過去國內曾爆發食用油遭混摻、造假事件，此重大食安問題引起社會大眾對於油品安全性產生高度關注，本文將透過食品巨量資料庫，結合跨部會資料，以油品交易流向管理為例，運用社會網絡分析（Social Network Analysis, SNA）方法，並以網絡密度與網絡中心性作為指標，探討整合分析食品相關大數據提供更精準的決策建議。研究結果顯示，結合社會網絡分析於關係複雜的油品交易網絡，可成功偵測食品業者於非食用油（例如飼料油或廢食用油等）交易網絡所擔任之角色，釐清食品業者購買非食用油的產銷鏈網絡之重要性，重要性愈高表示該食品業者於產品流通上具有舉足輕重的地位，藉以整合業者與業者之間所形成的「關係」資訊，與「業者」本身業別資訊，進而作為疑慮產品非法流入食品鏈風險偵測之綜合研判參考。因此本研究建議非食用油是否流入食品鏈之監控機制，可融入油品交易流向社會網絡分析，由大數據統計視角進行多元監測，期使油品管理事半功倍，為國內流通之油品把關。 The integration of various databases for big data analysis is an important policy for national development in Taiwan, which is also a critical challenge faced by Taiwan Food and Drug Administration. The results of big data analysis provide a foundation to formulate critical management policies, especially with the increasing numbers of food safety events. Furthermore, big data analysis can be employed to detect and monitor risk factors in order to prevent food safety events and continuously to track the incidents; hence, civilians could be protected. There was a series of incidents which affect food safety in Taiwan; among them, adulteration of cooking oil with recycled waste oil and animal feed oil was the worst and beyond imagination. This article addressed the Big Data Integration Model of cooking oil in Taiwan and Social Network Analysis (SNA) was used to demonstrate the statistical methods in detecting criminal frauds in the oil supply chains. SNA in a complex network of oil supply chains that could successfully monitor the relationships between individual oil industry and the transaction activities of animal feed oil in the upstream and downstream of the supply chain, and thus provided managerial insights to prevent further cooking oil frauds.