update README.md

Author: WeiqiNs

README.md

@@ -1,3 +1,11 @@
+# Reduce Leakages in Query Processing on Encrypted Multi-dimensional Data with Practicality
+
+Xinle Cao, Weiqi Feng, Quanqing Xu, Chuanhui Yang, and Jian Liu.
+
+(**Abstract**) Encrypted databases (EDBs) have been a prominent focus in the database community with the rise of cloud computing. However, many challenging open problems in EDBs make most existing solutions undeployable in real-world applications. One significant issue is how to avoid dangerous unexpected leakages when querying encrypted multi-dimensional data. For example, given one query SELECT $user\_id$ FROM $T$ WHERE $\mathsf{Garde}=\mathsf{A}$ AND $\mathsf{Depart}= \mathsf{CS}$, the minimal leakage should be the records that satisfy both the two conditions. However, most existing EDBs reveal separate frequency leakages: 1) records that satisfy $\mathsf{Garde}=\mathsf{A}$, and 2) records that satisfy $\mathsf{Depart} = \mathsf{CS}$. This incurs additional leakage of frequency information which has been widely recognized as very sensitive and dangerous information in EDBs.
+
+To this end, this work aims to reduce leakages as much as possible when querying multi-dimensional data. Specifically, we propose several new novel constructions to guarantee only minimal leakages revealed in a series of common queries over multi-dimensional data such as filtering single/multiple values within multiple columns, equi-join, and general equality test. We evaluate our schemes by comparing them with a naive (insecure) baseline and two SOTA works, the results show that our schemes indeed outperform prior works in both efficiency and practicality in deployment.
+
 ## Project Structure
 
 - The [experiment](experiment) folder consists of files to run experiments that we report in the paper.