| 摘要: | 目前大數據及開放資料盛行，為了保護個人隱私，故而有去識別化技術之誕生，與此同時，我國亦推出相應之法規範及標準。基於資料有用性與隱私保護之權衡下，是以有去識別化之必要，而技術之擇用則對去識別化過程之成敗至關重要，至於成敗判定之標準，則端視法規範之要求。由於本論文乃以去識別化技術之探析為主軸，而資料有用性與適法性又是不可不考量者，因此，本論文先闡釋及探討常見之去識別化技術，其次，介紹兩種常用的正規隱私量測模型，使得去識別化後之效能得以量測。嗣後，則設計一情境來演示去識別化過程，再就結果進行技術性、資料有用性之分析，並依據我國法規範探討其適法性。希望藉由本論文指出之三構面——技術性、資料有用性與適法性，以及在此三構面下所建立的K-匿名模型之框架，在組織有去識別化之需求時，能於技術擇用、模型建立及效能評估方面給予參考與指引。
At present, big data and open data are prevalent. In order to protect personal privacy, de-identification techniques have been born. At the same time, our country has also intro-duced corresponding legal norms and standards. Based on the trade-off between usefulness and privacy protection, de-identification is necessary, and the selection of de-identification techniques are crucial to the success or failure of the de-identification process. As for the success or failure of the de-identification process, the judgment criteria depends on the requirements of legal norms. Since this paper focuses on the analysis of de-identification techniques, and data usefulness and legality are indispensable, this paper first explains and discusses the common de-identification techniques. Secondly, two commonly used formal privacy measurement models are introduced, so that the effectiveness of de-identification can be measured. After that, a scenario was designed to demonstrate the process of de-identification, and then the results were analyzed in terms of technicality and data use-fulness, and the legality of the results was discussed based on our country's legal norms. It is hoped that through the three dimensions pointed out in this paper—technicality, data use-fulness and legality, as well as the framework of the K-anonymity formal privacy meas-urement model established under these three dimensions, when the organization has the need for de-identification, they can provide reference and guidance in technques selection, model building and performance evaluation. This may be the meager contribution of this research to the protection of personal privacy. |
