Abstract:
Systems and methods of matching identifiers between multiple datasets are described herein. A system can transmit a first identifier vector to a third party server. The first identifier vector can include a first identifier, first parameters, and second parameters. The system can receive, from the third party server, the first identifier vector encrypted based on a third-party encryption. The system can receive, from the third party server, a second identifier vector encrypted based on the third-party encryption associated with the third party server. The second identifier vector can include a second identifier, third parameters, and fourth parameters. The system can determine a correlation count between the first identifier vector and the second identifier vector. The system can determine that the first identifier corresponds to the second identifier based on the correlation count. The system can generate one identifier key for both the first identifier and the second identifier.
Abstract:
Systems and methods of cryptographically protecting location data transferred between servers via a network to tune a location engine are described herein. A first encryption module determines a first hash value for each location determination, encrypts the first hash value for each location determination using a first encryption protocol to generate a first encrypted data set, and transmits the first encrypted data set to a server. The first encryption module receives, from the server, a second encrypted data set with second hash values generated by a second encryption module of the server using a second encryption protocol. The first encryption module creates a first double encrypted data set from the second encrypted data set. A tuner compares the first double encrypted data set with a second double encrypted data set received from the server to adjust the location engine.
Abstract:
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for detecting and utilizing conversion data are disclosed. In one aspect, the system can receive an indication of a user interaction with content item presented at a user device. Based at least in part on a device identifier of the user device being received with installation data corresponding to the installation of an application, the system can attribute a conversion of the content item to the user interaction.
Abstract:
Systems and methods of cryptographically protecting location data transferred between servers via a network to tune a location engine are described herein. A first encryption module determines a first hash value for each location determination, encrypts the first hash value for each location determination using a first encryption protocol to generate a first encrypted data set, and transmits the first encrypted data set to a server. The first encryption module receives, from the server, a second encrypted data set with second hash values generated by a second encryption module of the server using a second encryption protocol. The first encryption module creates a first double encrypted data set from the second encrypted data set. A tuner compares the first double encrypted data set with a second double encrypted data set received from the server to adjust the location engine.
Abstract:
Aggregated transaction data from a transaction data provider may be encrypted and exchanged with a content item selection system using commutative encryption algorithms. The transaction data provider and content item selection system may utilize a set of common identifiers that are each encrypted using a respective commutative encryption algorithm of the transaction data provider or content item selection system. The other of the transaction data provider or content item selection system encrypts the single-encrypted common identifier using a respective commutative encryption algorithm to generate double encrypted common identifiers. The double encrypted common identifiers may be used to match a set of common identifiers with transaction data. The transaction data may be encrypted and/or may include random offset values.
Abstract:
Systems and methods for cross-device conversion estimates determine an observed number of cross-device conversions based on different sets of devices being logged into the same online accounts. The observed cross-device conversions are extrapolated out based on a percentage of one set of devices having logged into the same accounts as the second set of devices.