Behavioral targeting system

1. A method for behavioral targeting comprising:

receiving, by a processor, a plurality of events generated by a user;

classifying, by the processor, each of the plurality of user events into an event category;

storing, by the processor, the plurality of user events and associated event categories to an event log at predetermined time intervals;

selecting, by the processor as each of the plurality of events is stored, a model from a plurality of models based on a target objective, each of the plurality of models is associated with a different event category, the selected model comprising rules used for behavioral targeting processing, the target objective comprising brand advertising and direct response advertising, the behavioral targeting processing comprising applying the rules of the selected model to process the plurality of events to generate user scores for the target objective;

retrieving, by the processor, the plurality of stored user events from the event log wherein the predetermined time intervals span a pre-defined time period;

performing, by the processor, the behavioral targeting processing in a batch process comprising generating a plurality of short-term raw user interest scores and a plurality of long-term raw user interest scores by applying the rules of the selected model;

converting, by the processor, the plurality of short-term raw user interest scores and the plurality of long-term raw user interest scores to respective mapped short-term and long-term user interest scores;

generating, by the processor, a combined score of the user for the respective event category associated with selected model by combining the plurality of mapped short-term user interest scores and the plurality of mapped long-term user interest scores; and

communicating, by the processor to an advertising server, the combined score for enabling serving advertisements customized to the user.

2. The method of claim 1 , wherein the respective mapped user interest scores being indicative of a tendency of the user for a particular action.

3. The method of claim 1 , wherein the behavioral targeting processing comprises behavioral processing and dimension processing.

4. The method of claim 3 , wherein the dimension processing comprises recency, intensity and frequency processing.

5. The method of claim 4 , wherein each model comprises respective weights for the recency, the intensity and the frequency processing based on the respective category and an event type.

6. A computing system for behavioral targeting comprising:

at least one hardware processor; a storage medium for tangibly storing thereon program logic for execution by the processor, the program logic comprising:

event receiving logic, executed by the processor, for receiving a plurality of events generated by a user;

classifying logic, executed by the processor, for classifying each of the plurality of user events into an event category;

storing logic, executed by the processor, for storing the plurality of user events and associated event categories to an event log at predetermined time intervals;

model selecting logic, executed by the processor, for selecting, as each of the plurality of events is stored, a model from a plurality of models based on a target objective, each of the plurality of models is associated with a different event category, the selected model comprising rules used for behavioral targeting processing, the target objective comprising brand advertising and direct response advertising, the behavioral targeting processing comprising applying the rules of the selected model to process the plurality of events to generate user scores for the target objective;

retrieving logic, executed by the processor, for retrieving the plurality of stored user events from the event log wherein the predetermined time intervals span a pre-defined time period;

behavioral targeting processing logic, executed by the processor, for performing the behavioral targeting processing in a batch process comprising generating a plurality of short-term raw user interest scores and a plurality of long-term raw user interest scores by applying the rules of the selected model;

converting logic, executed by the processor, for converting the plurality of short-term raw user interest scores and the plurality of long-term raw user interest scores to respective mapped short-term and long-term user interest scores;

generating logic, executed by the processor, for generating a combined score of the user for the respective event category associated with the selected model by combining the plurality of mapped short-term user interest scores and the plurality of mapped long-term user interest scores; and

communicating logic, executed by the processor, for communicating the combined score for enabling serving advertisements customized to the user.

7. The system of claim 6 ,

wherein the behavioral targeting processing logic, executed by the processor, for performing the behavioral targeting processing in a batch process comprising generating the plurality of long-term raw user interest scores for a plurality of categories into which the plurality of received events are classified, each of the plurality of long-term raw user interest scores corresponding to each of the predetermined time intervals at which the plurality of user events are retrieved.

8. A non-transitory computer readable storage medium for behavioral targeting, comprising instructions, which when executed by a processor cause the processor to:

receive a plurality of events generated by a user;

classify each of the plurality of user events into an event category;

storing the plurality of user events and associated event categories to an event log at predetermined time intervals;

select, as each of the plurality of events is stored, a model from a plurality of models based on a target objective, each of the plurality of models is associated with a different event category, the selected model comprising rules used for behavioral targeting processing, the target objective comprising brand advertising and direct response advertising, the behavioral targeting processing comprising applying the rules of the selected model to process the plurality of events to generate user scores for the target objective;

retrieve the plurality of stored user events from the event log wherein the predetermined time intervals span a pre-defined time period;

perform the behavioral targeting processing in a batch process comprising generating a plurality of short-term raw user interest scores and a plurality of long-term raw user interest scores by applying the rules of the selected model;

convert the plurality of short-term raw user interest scores and the plurality of long-term raw user interest scores to respective mapped short-term and long-term user interest scores;

generate a combined score of the user for the respective event category associated with the selected model by combining the plurality of mapped short-term user interest scores and the plurality of mapped long-term user interest scores; and

communicate to an advertising server, the combined score for enabling serving advertisements customized to the user.

9. The non-transitory computer readable storage medium of claim system of claim 8 comprising instructions for the processor to perform the behavioral targeting processing in a batch process comprising generating the plurality of long-term raw user interest scores for a plurality of categories into which the plurality of received events are classified, each of the plurality of long-term raw user interest scores corresponding to each of the predetermined time intervals at which the plurality of user events are retrieved.