大數據對於個人資料保護之挑戰與因應之道

The Impact of Big Data on Personal Information Protection and its Solutions

大數據，乃能蒐集與儲存巨量資料，藉由分析技術而由彙整資料中發現、揭露、預測某些身分、行為、模式、趨勢、關聯性與實用知識，為許多問題提供解決之道。大數據雖有不少益處，但亦形成個人隱私與自主等權益之威脅。大數據運作模式，乃蒐集資料越多越好，保有資料越久越好，除供原始蒐集目的之利用外，尚能供預先設想不到的目的或方式而分析運用之。惟大數據運作模式將與當代各國際組織、國家普遍接受「公平資訊實務原則」（FIPPs）之個資保護原則（如資料最小化、目的明確、利用限制、通知與同意〔告知後同意〕、資料正確性、資料關聯性等原則），有所衝突。因此，歐美諸多組織與學者乃呼籲，奠基FIPPs之個資保護立法架構應大幅更動；然而，歐盟則認為FIPPs個資保護原則仍可適用。面對大數據對於個資保護所形成之挑戰，國際發展趨勢之探討，可供我國釐定因應之道的借鏡。上開立法架構應大幅更動而放寬限制、可更容易利用個資之主張，固有益大數據發展，但實質上則隱藏著激進去管制化目標而仍將危害隱私等；因此，該主張能克服質疑而被普遍採納前，現行個資保護原則仍具適用性惟可因應微幅調整，具體運作上，可借鏡歐盟、英國資訊委員辦公室（ICO）建議之多項遵循措施，以期大數據能善用個資，尚可兼顧隱私等權益之保護。

Big data usually refers to the proliferation in data volumes and types, the increase in the speed for collecting, processing, and using that data, and the improvement of technical solutions to analyze, find, and draw hidden information, surprising correlations, and intelligent inferences from the data. Big data can bring benefits to society, but it also leads to a threat to privacy and autonomy of individuals. Most developed nations and several international organizations have comprehensive privacy laws based on the Fair Information Practice Principles (FIPPs)- a set of commonly accepted protections, which all roughly cover the same ground: transparency, purpose specification, use limitation, data minimization, data accuracy, individual participation, security, and accountability. However, the principle of big data is to collect more and more data in the hopes that the data might be able to be used in unexpected ways. Therefore, several FIPPs, including data minimization, purpose specification, use limitation, informed consent, and accuracy and relevance of data, will be in contradiction with the operation of big data. Currently, the European Union and American scholars try to figure out the solutions to how to fix the problem of the contradiction between the operation of big data and the FIPPs. The solutions are also useful for Taiwan to protect the personal information of individuals while facing the challenges of big data.