基於安全性的排名優化

隨著雲端儲存服務的普及，越來越多使用者選擇將資料儲存於雲端服務平台上（如：Dropbox、Google Drive、Apple iCloud Drive等），以方便後續管理與使用。然而，對於一些較隱私的資料，使用者儘管仍欲使用服務供應商所提供的儲存與管理服務，但因擔心可能遭受他人窺視、攻擊以及資料外洩的風險，而往往會傾向不將該類資料放於雲端平台上。因此，為提供使用者一個更好且安心的使用環境，許多雲端服務所需具備的隱私要求（privacy requirements）與相關加密與搜尋技術相繼被提出。其中一部分為有關搜尋模式（search pattern）的隱藏，即當使用者下達一組關鍵字查詢（query）後，對於最終回傳的排名結果（ranked list）應做適度隱藏。以避免雖已對查詢所下達的關鍵字做隱藏的動作，但因回傳的排名結果間的相似，而致第三方推測出兩筆查詢間有所關聯。並減少因洩漏使用者的行為模式或較常存取哪些特定文件等資訊，以使第三方有機會對此作特定攻擊。　　然而對於搜尋模式的隱藏，在現有辦法中皆存在著一些問題，如未考慮排名結果間次序性的重要度，搜尋模式隱藏時未對排名結果的效能做維持，或者耗費過多時間做隱藏等。為此，我們基於文件與查詢字眼間所計算出的相關分數（relevance score），提出同儕的概念。在維持排名結果品質的情況下，讓同一組的關鍵字查詢盡可能產生不同回傳結果，以隱藏使用者的搜尋行為。

關鍵字

安全性；隱藏搜尋模式；排名效能；排名關聯度；基於同儕

並列摘要

Since cloud service are wide-spread used nowadays, more and more users store their data on cloud storage (e.g., Dropbox, Google Drive, Apple iCloud Drive, etc.) for management. However, because of the risks of information leakage, user might tend not to store sensitive personal data in cloud. Thus some privacy requirements and corresponding technologies are being proposed. One of the requirements is search pattern privacy (i.e., hiding the query search’s result). To avoid a third party infer query correlation from result correlation (i.e., similar query results might come from similar queries). And preventing focused attack because of private interest leakage. Unfortunately, existing search pattern hiding techniques have some shortcomings. For example, the method does not consider rank order, do not maintain ranking performance, or is inefficient. Thus, based on the relevance score between document and query keyword, we propose the concept of peer. Not only to generate as different result as possible by the same query, but also maintain ranking performance. That means achieve search pattern privacy.

並列關鍵字

security ； search pattern privacy ； ranking performance ； rank correlation ； peer-based

參考文獻

[2] D. Chaum, "Untraceable electronic mail, return addresses, and digital pseudonyms," in Communications of the ACM, vol. 4, no. 2, pp. 84-90, Feb. 1981.

[5] O. Goldreich, R. Ostrovsky, "Software protection and simulation on oblivious RAMs," in Journal of the ACM, vol. 43, no. 3, pp. 431-473, May 1996.

[9] E. Stefanov, E. Shi, "Oblivistore: High performance oblivious cloud storage," in Security and Privacy, 2013 IEEE Symposium on, pp. 253-267, May 2013.

[12] M. Shokouhi, "Segmentation of search engine results for effective data-fusion," in Proceedings of the 29th European conference on IR research, pp. 185-197, 2007.

[14] S. Wu and F. Crestani. "Methods for ranking information retrieval systems without relevance judgments," in Proceedings of the 2003 ACM symposium on Applied computing, pp. 811-816, Mar. 2003.

國際替代計量

基於安全性的排名優化

主題瀏覽