REAL TIME CLICK (RTC) SYSTEM AND METHODS

• US 20090094073A1
• Filed: 10/03/2007
• Published: 04/09/2009
• Est. Priority Date: 10/03/2007
• Status: Abandoned Application

• Current Assignee: Oath Inc. (Verizon Communications Inc.)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Siddiqui, Muhammad Ali, Sorkin, Brian Wayne, Ahmed, Rana Anjum, Wong, Robert K., Hu, Eugene Freeman, Cheung, Dominic
• Application Number:
  US11/866,548
• Publication Number:
  US 20090094073A1
• Time in Patent Office: Days
• US Class Current: 705/14.47
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0248 Avoiding fraud
• CPC Class Codes: G06Q 40/12 Accounting

Abstract

A real-time click system to process advertisement (ad) clicks includes a real-time listener operative to listen for, and store in a memory, a plurality of click event packets emitted by an ad server when corresponding ads are clicked by web users. A sequencer stores the click event packets in a database. A collector is coupled with the real-time listener and the sequencer and is operative to retrieve the plurality of click event packets, upon request, from the memory at a predetermined time interval. The collector also determines a partition number associated with each of the plurality of click event packets and sends to the sequencer the click event packets having a partition number corresponding to the sequencer.

Claims

1. A real-time click system to process advertisement (ad) clicks, comprising:

a real-time listener operative to listen for, and store in a memory, a plurality of click event packets emitted by an ad server when corresponding ads are clicked by web users;

a sequencer to store the click event packets in a database; and

a collector coupled with the real-time listener and the sequencer and operative to retrieve the plurality of click event packets, upon request, from the memory at a predetermined time interval, wherein the collector determines a partition number associated with each of the plurality of click event packets and sends to the sequencer the click event packets having a partition number corresponding thereto.

2. The system of claim 1, wherein the listener collects the plurality of click event packets by listening for light-weight event system (LWES) events emitted by the ad server, and wherein the memory comprises a circular buffer memory.

3. The system of claim 1, further comprising a plurality of the listeners, a plurality of the collectors, and a plurality of the sequencers to form a plurality of data pipelines in which the plurality of collectors can retrieve click event packets from any of the plurality of listeners, and only click event packets of a pre-determined partition number are sent to corresponding sequencers.

4. The system of claim 3, wherein only one collector retrieves packets from any given listener at a time and wherein each collector maps the pulled click event packets into a plurality of data packets of discrete time periods (N) for each advertiser.

5. The system of claim 1, further comprising:

a priority queue memory of the sequencer operative to enable updating priority of the click event packets corresponding to each of a plurality of advertiser accounts based on a pluggable implementation of at least one business rule.

6. The system of claim 5, wherein the at least one business rule comprises a rule that tracks at least one of a number of clicks and a cost for the number of clicks.

7. The system of claim 5, wherein the click event packets are processed within the priority queue memory and then stored in the database of each sequencer in descending order of priority.

8. The system of claim 7, wherein to process the click event packets the sequencer applies filters thereto, including at least a defrauding filter, and aggregates the data packets into a time window (N′

).

9. The system of claim 1, wherein each of the collector and sequencer comprise a memory, wherein at least one of the collector and the sequencer expose the data packets in their respective memories using a remote call that enables application programming interfaces (APIs) of remote processes to monitor the click event packets and memory usage for statistical purposes.

10. The system of claim 1, further comprising:

a query application programming interface (API) to query the database on certain criteria to provide access to the system by remote clients, wherein the query API includes a fault tolerance that can retrieve click event packets even if subsets of partitions are unavailable.

11. The system of claim 1, wherein the real-time listener exposes a STAT command through a control protocol interface, and in response to the STAT command, the real-time listener returns an internal status for a running listener process.

12. A method of processing real-time advertisement (ad) clicks in a real-time click system, comprising:

listening by a plurality of real-time listeners for click event packets emitted by an ad server when an ad is clicked by web users;

storing the click event packets in a memory buffer of each listener in real-time;

pulling the click event packets from the memory buffers by a plurality of backend components at a predetermined time interval;

determining a partition number associated with each of the click event packets;

processing the data packets for a predetermined partition number; and

storing the data packets in a database corresponding to the predetermined partitioned number, wherein the data packets become accessible by remote agents for querying.

13. The method of claim 12, wherein the backend component comprises a collector to pull the click event packets from the plurality of memory buffers, and a sequencer to receive the click event packets from the collector having a partition number corresponding thereto, wherein the sequencer processes the data packets and stores the data packets in the database.

14. The method of claim 13, further comprising:

partitioning the click event packets into data packets of discrete time periods (N) for each of a plurality of advertisers and based on the predetermined partition number.

15. The method of claim 14, further comprising:

updating a priority of the data packets in a priority queue memory of the respective sequencers, wherein the data packets of each advertiser receives a priority based on a pluggable implementation of at least one business rule, and wherein the data packets are processed and stored in the database in descending order of priority.

16. The method of claim 15, wherein the at least one business rule comprises a rule that tracks at least one of a number of clicks and a cost of the number of clicks, the method further comprising:

giving a higher priority to the data packets of advertisers that at least one of receive more click events and incur higher costs for the user clicks.

17. The method of claim 16, further comprising:

updating a budget balance of an advertiser based on the cost of the number of clicks; and

turning off the advertiser account if the cost of the number of clicks exceeds the budget balance during a pre-determined time period.

18. The method of claim 14, wherein processing the data packets comprises:

executing click defrauding by the sequencer as defined by a series of pluggable defrauding filters; and

aggregating the data packets into a time window (N′

) for each advertiser account.

19. The method of claim 14, further comprising:

enabling remote clients to query the data packets in the database on certain criteria through a query application programming interface (API).

20. The method of claim 14, wherein the collector includes a memory, the method further comprising:

exposing the data packets in the memories of at least one of the collectors and the sequencers using a remote call that enables application programming interfaces (APIs) of remote processes to monitor the data packets and memory usage for statistical purposes.

21. The method of claim 14, wherein the predetermined partition number is determined through:

taking a hash code of each of a plurality of account identifications associated with the data packets; and

applying a modulator operator on the result of each hash code to result the partition number for each data packet.

22. A method of processing real-time advertisement (ad) clicks in a real-time click system, comprising:

receiving into memory a plurality of data packets including attributes of captured click events, wherein the data packets are pre-partitioned into discrete time periods (N) for each of a plurality of advertiser click groupings, and wherein the data packets belong to at least one of a plurality of predetermined partition numbers;

updating a priority of the data packets in a priority queue memory, wherein each click grouping receives a priority based on a pluggable implementation of at least one business rule;

processing the data packets for a predetermined partition number in descending order of the priority; and

saving the processed data packets to a database.

23. The method of claim 22, wherein processing the data packets comprises at least one of:

discarding a certain number (M) of time intervals in the past from a current time period; and

aggregating clicks at a campaign level by summing raw click charge amounts for each time period N for each click grouping.

24. The method of claim 22, further comprising:

executing click-through protection that includes removal of clicks from the data packets through at least one of;

filtering the click events to remove duplicate click events;

tracking internet protocol (IP) addresses of searchers by using a series of pluggable defrauding filters; and

blacklisting IP addresses that meet criteria of the series of pluggable defrauding filters.

25. The method of claim 24, further comprising:

enabling remote clients to query the click events data packets in the database on certain criteria through a remote application interface.