Systems and methods for pacing information campaigns based on predicted and observed location events

A system, comprising:

• one or more databases storing therein datasets associated with mobile devices, wherein each respective dataset of at least some of the datasets includes a respective time stamp and at least one respective event involving an associated mobile device at a time indicated by the respective time stamp;
• one or more processors;
• and
• one or more memories into which one or more computer programs can be loaded for execution by the one or more processors to function as one or more modules, units and servers including:
• a model training unit configured to machine train a location prediction model for an information campaign, the information campaign having a set of campaign parameters including parameters specifying one or more places of interest (POIs), a time span for the information campaign, and a budgeted number of conversions to be achieved by the information campaign during the time span, wherein a respective conversion is associated with a respective mobile device and corresponds to the respective mobile device having at least one location event at any of the one or more POls during a specific time frame, wherein the model training unit includes a feature engineering module configured to construct a feature space corresponding to a plurality of mobile devices using datasets associated with respective ones of the plurality of mobile devices and having time stamps in a training time period, and to determine a plurality of labels corresponding to the plurality of mobile devices, and a machine learning module configured to train a general prediction model using the training feature space and the plurality of labels, wherein a respective label corresponding to a respective mobile device is determined based on whether the respective mobile device has had at least one location event at any location of a group of one or more predefined locations during a training time frame;
• a front-end server configured to receive and process information requests, and to generate processed information requests that are stored as datasets in the one or more databases, wherein, during a respective time unit of one or more successive time units across the time span of the information campaign, for each particular information request of a plurality of information requests received during the respective time unit, the front end server is configured to determine if the particular information request qualifies for the information campaign based on data associated with the particular information request and the set of campaign parameters of the information campaign;
• and
• a prediction unit configured to predict probabilities of mobile devices to have location events at any of the one or more POls during certain time frames of an information campaign, wherein, the prediction unit is configured to, in response to the particular information request qualifying for the information campaign, apply the general prediction model to a particular feature set constructed using datasets associated with the particular mobile device to obtain a prediction result and to predict a conversion probability for a particular mobile device associated with the particular information request using the prediction result, wherein the conversion probability corresponds to a predicted probability of the particular mobile device having at least one location event at any of the one or more POls during a particular time frame;
• wherein the front-end server is further configured to:
• determine a respective target number of conversions to be achieved by the information campaign during the respective time unit;
• and
• determine a response to the particular information request based at least in part on the conversion probability and on the respective target number of conversions.