110年度跨部會化學物質資訊服務平台(化學雲)應用計畫
| 中文摘要 | 本年度在提升化學雲平台效益，擴增及維護平台資料與功能部分，建立各 部會國內易爆物資料上傳功能、開發廠場分布查詢與資訊查詢功能；擴增3個 系統資料以及更新1個系統資料，持續維護10部會52個化學物質管理資訊系統 定期拋轉。此外，整合與提升化學雲平台相關視覺化功能，完成1項客製化快 報開發並提供介接，開立261個化學雲帳號與持續進行平台資安與運作維護。 另完成辦理16場相關機關協商會議及2場工作坊。 運用科技技術，提升化學雲應用價值部分，建立化學物質流向網絡圖及食 安異常交易之上下游視覺化網絡查詢；完成非監督式學習的自編碼器訓練，產 出6類食品廠異常廠商清單。此外，完成食品業化學品安全監控系統架構設計、 建立輿情分析訓練資料集與訓練模型及輿情分析結果展示模板。而在應用區塊 鏈，技術結合化學物質運作廠場，建立示範場域部分，完成區塊鏈軟體開發及 主機安裝；並完成化學物質交易資料傳輸上鏈及區塊鏈交易查詢網站，進行笑 氣供應商之鋼瓶流向資訊、化學品供應商及其下游廠商資訊上鏈示範。 在整合化學物質相關資訊，研擬管理策略部分，研擬具食品添加相似特性 及效用化學物質清單，在比對歐盟、美國食品藥物管理局(U.S. Food and Drug Administration, FDA)與衛福部公告清單後，完成包括著色劑共13種與防腐劑22 種之物種清單。另完成化學物質指引表與整併表問題檢討及補正，優化化學物 質整併，與食安風險相關化學物質整併結果改正率為2%。定期檢視更新國內外 相關法規及歐盟關切化學物質清單，及完成110年度國內食品相關稽查抽驗結 果蒐集及國內食品安全事件資訊彙整。 | ||
|---|---|---|---|
| 中文關鍵字 | 化學物質、跨部會合作、警示和快報 | ||
基本資訊
| 專案計畫編號 | 經費年度 | 110 | 計畫經費 | 13897.575 千元 | |
|---|---|---|---|---|---|
| 專案開始日期 | 2021/01/21 | 專案結束日期 | 2021/11/30 | 專案主持人 | 李曜全 |
| 主辦單位 | 化學局 | 承辦人 | 林桂如 | 執行單位 | 景丰科技股份有限公司 |
成果下載
| 類型 | 檔名 | 檔案大小 | 說明 |
|---|---|---|---|
| 期末報告 | 110化學雲成果報告_本文.pdf | 14MB |
To Apply the Chemical Substances Management Information Project
| 英文摘要 | In terms of the improvement of efficiency and data maintenance of the ChemiCloud platform, we’ve established domestic explosives data upload function for various ministries and have developed factory distribution query functions during this year. We’ve expanded 3 data systems and updated 1 data system and 10 departments. 52 chemical substance management information systems are maintained on a regular basis. In addition, the related visualization functions of the ChemiCloud platform were enhanced. Furthermore, we’ve developed and provided a customized quick report. A total of 261 ChemiCloud accounts were created and cybersecurity was enhanced. 16 consultation meetings with relevant agencies and 2 workshops were also completed. In terms of big data analysis, we’ve created a visualized time series observation panel template based on the company’s name, chemical substance and time range of the chemical substance flow network diagram. We used unsupervised learning to detect abnormal vendors and used self-encoding device to generate an abnormal score value for the transaction of the manufacturer responsible for testing. In addition, we completed the architecture design of the food industry chemical safety monitoring system, and established a public opinion analysis training data set and training model, and a display template of public opinion analysis result. Besides, in the application of blockchain technology in conjunction with chemical substances to operate factories, we’ve completed the development of the blockchain software and host installation. We’ve also completed the blockchain transaction query website. The demonstration of information on the chain of chemical suppliers and their downstream manufacturers and demonstration of information on the flow of cylinders of Nitrous oxide suppliers on the chain were demonstrated. In terms of other analytical works, we’ve developed a list of chemical substances with similar characteristics and usefulness in food additives. The list of species included 13 kinds of colorants and 22 kinds of preservatives. In addition, we’ve completed the review and correction of the problems of the chemical substance guide table and the consolidation table. We’ve also optimized the consolidation of chemical substances. Therefore, the correction rate of the consolidation of chemical substances related to food safety risks was 2%. We reviewed and updated domestic and foreign central customs laws, regulations and the EU's list of chemical substances of concern regularly. We’ve also completed the collection of the results of the domestic food inspection and finished the compilation of domestic food safety incident information during this year. | ||
|---|---|---|---|
| 英文關鍵字 | Chemical Substance, Inter-Ministerial Cooperation, Alert and Quick Report | ||