Abstract: Systems and methods of determining a geographic location based conversion via a computer network. A data processing system can receive, from a computing device, an indication of interest in an online content item that is associated with a commercial entity having a location. The system can also receive data points that each include time information and location information of a computing device. The system can generate a valid cluster indicating an activity at the location of the commercial entity by evaluating the time information and the location information of two or more data points with a duration threshold and a location threshold. The system can use the valid cluster to identify the indication of interest in the online content item as the location based conversion based on the activity at the location of the commercial entity.

Abstract: Systems and methods of determining a geographic location based conversion via a computer network. A data processing system can receive, from a computing device, an indication of interest in an online content item that is associated with a commercial entity having a location. The system can also receive data points that each include time information and location information of a computing device. The system can generate a valid cluster indicating an activity at the location of the commercial entity by evaluating the time information and the location information of two or more data points with a duration threshold and a location threshold. The system can use the valid cluster to identify the indication of interest in the online content item as the location based conversion based on the activity at the location of the commercial entity.

Abstract: Systems and methods of generating a valid cluster based on a location of a commercial entity via a computer network. A data processing system can receive the location of the commercial entity from an online content selection data structure, and can receive data points that comprise location information and time information associated with an identifier. Using the data points, the data processing system can create a location cluster having a first data point and subsequent data points that are within a threshold distance from the first data point. The data processing system can determine a cluster duration for each location and compare the cluster duration with a duration threshold to identify a set of valid location clusters. The valid location clusters can be filtered based on a proximity to the location of the commercial entity to identify a valid cluster that can be indicative of activity at the commercial entity.