Carolina Naim: March 4

Private Multi-Group Aggregation

Carolina NaimRutgers University

We consider the private multi-group aggregation (PMGA) problem. This setting involves n users, and each user belongs to one of k distinct groups and holds an integer value. A central server wants to find the aggregate (sum) of the values in each group (with high accuracy) under communication and local differential privacy constraints. The privacy constraint guarantees that the user’s group remains private. This is motivated by applications where a user’s group can reveal sensitive information, such as religious and political beliefs, health conditions, or race.  
In this talk, we will introduce the Query and Aggregate (Q&A) scheme for PMGA. The novelty of Q&A is that it is an interactive aggregation scheme. In Q&A,  each user is assigned a random query matrix, to which he sends the server an answer based on his group and value. We will compare Q&A to the Randomized Group (RG) scheme, which is non-interactive and adapts existing randomized response schemes to the PMGA setting.