資料科學於我國食安風險偵測之應用概況   全文下載

An Overview of Data Science Application on Food Safety Risk Identification in Taiwan

隨著巨量資料時代的來臨，世界各國家政府皆面臨數位轉型，運用資料分析將資料轉化為有用的資訊，進而輔助決策。本文彙整我國食品藥物管理署導入資料科學輔助各面向食品安全管理之應用策略，說明食品自邊境輸入到後市場流通之現況與挑戰，透過導入機器學習方法，精準鎖定高風險產品及業者，並強化食品追溯追蹤風險控制；另針對國人高關注食安議題，介紹如何擬定風險偵測方式，輔助實務稽查應用，防堵相關食安事件。隨著各式新興科技日益蓬勃發展，未來可透過拓展資料來源、引入相關技術、進行人才培育等方式，持續強化食品巨量資料之加值應用，提升政府精準決策之能力。

As the emergence of the Big data, the governments all over the world are facing the challenge of digital transformation. With the aid of Big data technology, primary data can be transformed into useful information to assist decision making. In this report, we summarized the preliminary achievement by applying data science technology on food safety management by Taiwan Food and Drug Administration (TFDA). Current situations and challenges of food safety issues were demonstrated from border inspection to post-market surveillance by implementing the machine learning methodology, precisely searching for potential high risk food products and food product dealers to enhance food products traceability, and risk control. In order to strengthen the ability to prevent the food safety events, we focused on food safety issues with high public attentions and introduced methods to establish risk detection systems. As variety of new technologies come into existence, the Big data applications on food safety management can thus be improved via expanding data sources, incorporating relevant technical skills and talent cultivation, etc., to enhance the government policy-making capabilities.