The number of people storing their data on the cloud has increased significantly in the past decade due to the advancement in technology. Thus, with an increase in the amount of data on their servers, cloud service providers have begun outsourcing their data to remote service providers in order to save costs. However, data security becomes an important issue in such cases. It is, therefore, necessary that cloud service providers encrypt their data locally before storing them on remote servers. Searchable symmetric encryption is a technique which facilitates searching and retrieval of data that has been stored in an encrypted form on a remote server, without having to decrypt it. In this paper, we tackle the problem of similarity search over large scale textual and image data using searchable encryption with feature-based and efficient indexing mechanism. This helps against adaptive chosen keyword attacks. The proposed method shows a promising result in terms of recall and speed.

在过去的十年中，由于技术的进步，将数据存储在云中的人数已大大增加。因此，随着服务器上数据量的增加，云服务提供商已开始将其数据外包给远程服务提供商，以节省成本。但是，在这种情况下，数据安全性成为重要的问题。因此，在将数据存储在远程服务器上之前，云服务提供商必须在本地对其数据进行加密。可搜索对称加密是一种便于搜索和检索以加密形式存储在远程服务器上的数据而无需对其进行解密的技术。在本文中，我们使用具有基于特征的高效索引机制的可搜索加密解决了大规模文本和图像数据的相似性搜索问题。这有助于抵制自适应选择的关键字攻击。提出的方法在召回率和速度方面都显示出令人鼓舞的结果。