Abstract: A method, implemented by a processor, for combining multiple data sources in a product purchase study includes acquiring, by a processor, first product purchase data for a product from a first data source, the first product purchase data uniquely identifying the product; sending, by the processor, the first product purchase data to a remote server; receiving, by the processor, a signal from the remote server based on the first product purchase data, the signal comprising a request for additional product purchase data; acquiring by the processor in response to the request, second product purchase data from a second source independent of the first source to the remote server; and sending the second product purchase data to the remote server.

Abstract: A computer-implemented method for collecting and managing TV viewership data from multiple TV metering data providers is disclosed. The method includes: receiving an event log file at a distributed computer system that includes multiple computers; dynamically selecting one or more computers according to a predefined sharding function; at each of the selected computers: allocating a set of compressed event records, which corresponds to a subset of the event log file, at predetermined locations within the memory of the computer; and in accordance with a predefined schedule, replicating the compressed event records from a respective one of the selected computers to one or more other computers of the distributed system such that there are at least two replicas of any event record on at least two computers of the distributed computer system.

Abstract: A method for providing efficient and accurate estimates of TV viewership ratings through a distributed computer system that includes multiple computers is disclosed. The method includes: receiving a query from a client at the distributed computer system; dynamically selecting one or more computers according to a predefined sharding function; at each of the selected computers, determining a count of qualified event records that satisfy the query; aggregating the respective counts of qualified event records determined by the selected computers; statistically projecting the aggregated count of qualified event records into an estimated total count of qualified event records on the distributed computer system; and returning the estimated total count of qualified event records to the requesting client.

Abstract: A cluster of computer system nodes connected by a storage area network include two classes of nodes. The first class of nodes can act as clients or servers, while the other nodes can only be clients. The client-only nodes require much less functionality and can be more easily supported by different operating systems. To minimize the amount of data transmitted during normal operation, the server responsible for maintaining a cluster configuration database repeatedly multicasts the IP address, its incarnation number and the most recent database generation number. Each node stores this information and when a change is detected, each node can request an update of the data needed by that node. A client-only node uses the IP address of the server to connect to the server, to download the information from the cluster database required by the client-only node and to upload local disk connectivity information.