大數據分析與個人資料保護之衝突：從收視行為調查談起

Abstract

隨著數位化時代來臨，「大數據」被視為解決傳統收視行為調查弊病的解方；然而，在利用此一新興工具時，大量被鉅細靡遺留存下來的個人收視行為紀錄，也引發了對於智慧隱私的威脅、過濾罩效應，甚至是因自動化演算程式而產生歧視等種種問題。\n在比較法上，關於個人收視行為紀錄的保護，不論是採取部門式立法的美國，或是制定有全面性個人資料保護規範的歐洲，均有所著墨；但在迎接大數據分析的挑戰時，相關規範皆面臨個人資料保護與個人資料利用之間的衝突與拉鋸。\n我國也不例外，隨著個人資料保護意識的提升，其與個人資料利用之間的關係也愈趨緊張。雖然在現行個人資料保護法下，以大數據分析開展收視行為的調查及應用，同時兼顧個人隱私的保障，並非毫無機會；但若要在個人資料保護與個人資料利用的緊張關係之間尋求更佳的平衡點，現行法仍有未臻周全之處，有待修法解決。\n本文除了就現行個人資料保護法下，以大數據分析進行收視行為調查，提供相關執行面的建議外；也嘗試針對現行法有待改善的部分，提出未來修法的建議方向。期待在大數據應用不斷發展的同時，一併打造出足以令人信賴的個人資料利用環境。

With the coming of the digital age, “big data” is believed to be the panacea for the problems of traditional audience measurement methods. Nevertheless, a great deal of detailed personal viewing behaviors is collected and stored while using this emerging tool. As a result, it raises concerns over intellectual privacy, filter bubble, and even discrimination which results from automatic parsing algorithm. From the comparative law perspective, both the United States, which adopts a sectoral data protection framework, and the European Union, which has a comprehensive personal data protection law, have already addressed the need to protect information regarding personal viewing behaviors. When facing the challenges of big data analytics; however, both legal regimes are still confronted with conflicts between the protection and use of personal data. There is no exception for Taiwan. With the rising awareness of personal data protection, the tension between the use and protection of personal data is also increased. Although under the current Personal Information Protection Act (PIPA), it’s possible to employ big data analytics on audience measurement, while at the same time, conforming to the minimum protection of individual privacy, if we want to strike a better balance between data protection and data use, the PIPA needs to be revised. In addition to providing practical suggestions for conducting audience measurement using big data analytics in accordance with the PIPA, this thesis also tries to provide a proposal for future revision of the PIPA. This thesis hopes that as big data applications continue to develop; we can also foster a more trustworthy environment for the use of personal data.