Graduate Thesis Or Dissertation


Efficient Private Matching for Private Databases 公开 Deposited




Attribute Name LabelAttribute Values Label
  • Private matching (PM) is a key cryptographic primitive in secure computation that allows several parties to jointly compute some functions depending on their private inputs. Indeed, this primitive has many practical applications. For instance, in online advertising, two companies may wish to find their common customers for a joint marketing campaign. In this scenario, privacy is of utmost importance and it is imperative to ensure that neither company can learn more than their own data and the results of the match. In recent years, secure computation in general and PM in particular has attracted considerable attention from the research community, partly due to the rise of Big Data along with the ever-increasing privacy concerns. This thesis describes three secure private set intersection (PSI) protocols, one private set union (PSU) construction, and one pattern matching scheme. PM can be considered as a main building block in all these protocols. To securely compute the intersection of two sets of size $2^{20}$ our proposed protocols require only 3 seconds which is $4\times$ faster than the previous best protocol. In the multi-party setting, we provide the first implementation that takes only 72 seconds to compute PSI for 5 parties with data-sets of $2^{20}$ items each. For private set union (PSU), our protocol improves prior work by a factor up to $7,600 \times$ for large instances. In addition, our wild-card pattern matching (WPM) protocol shows over two orders of magnitude faster than the state-of-the-art scheme.
License Label
Resource Type
Date issued
Degree Level
Degree Name
Degree Field
Degree Grantors
Graduation Year
Contributor Advisor
Contributor Committeemember
Academic Affiliation


Relationships Parent Rows Label

Rows Empty Text

属于 Collection: