Facebook, Inc.

Doordash, Inc.

Aman Dhesi

Alok Gupta

Han Gyul Lee

Robert Bryant Kaspar

Ariel Benjamin Evnine

Louis Benoit Philippe Boval

Zeev Rosenstein

In some examples, a computing device may receive data from a plurality of groups of data sources. The computing device may access a plurality of data synthetization machine learning models configured for generating synthetic data. Respective ones of the data synthetization machine learning models may correspond to respective ones of the groups of data sources. The computing device generates first synthetic data by inputting, to a first data synthetization machine learning model, first data received from a first data source group, and generates second synthetic data by inputting, to a second data synthetization machine learning model, second data received from a second data source group. The computing device determines an allocation of resources based at least in part on comparing the first data and the first synthetic data with the second data and the second synthetic data.

Machine learning with data synthesization

In some examples, a computing device may receive data from a plurality of groups of data sources. The computing device may create a training data set from a first portion of the received data and may create a plurality of validation data sets from a second portion of the received data. For example, each validation data set may correspond to a respective one of the groups of data sources. The computing device may train, using the training data set, a plurality of machine learning models configured for synthesizing data. For instance, respective ones of the machine learning models may correspond to respective ones of the groups of data sources. Further, the computing device may validate the respective machine learning models using the respective validation data set corresponding to the respective group to which the respective machine learning model being validated corresponds.

Exemplary embodiments relate to techniques for determining social networking or messaging user affinity and engagement coefficients (e.g., a measure of the connectedness between two people in a network). The described techniques are particularly well-suited to cases in which only limited information is available, such as when a new user joins a network and only the user's contacts list is available. The available information may be used to determine a group of existing users to which the new user is connected. Some embodiments relate to calculating scores among these existing users in order to infer an affinity for the new user to the existing users. Other embodiments involve calculating bilateral scores that reflect a degree of mutual affinity between two users.

Methods and systems for assigning affinity scores to contacts

Techniques for device configuration of prospective contacts using messaging history information are described. In one embodiment, an apparatus may comprise a client front-end component operative to receive a client inbox request for a user account from a client device, the user account for a communication system; and send an ordered prospective contact list to the client device in response to the client inbox request; a prospective contact list component operative to generate a prospective contact list for a user account, wherein generating the prospective contact list for the user account excludes any existing contacts from an existing contact list for the user account; a predicted interest component operative to determine a predicted communication interest for each prospective contact on the prospective contact list; and a contact ranking component operative to determine a ranking weight for each prospective contact on the prospective contact list. Other embodiments are described and claimed.

Growing community of inventors

San Francisco, CA, United States of America

Aman Dhesi