Social media monitoring system and method

1. A computer software system of social media account monitoring, comprising:

a processor and a memory operatively coupled with the processor, such that the processor is configured to execute program code to implement a plurality of filters, the program code comprising:

a sampling scheduler configured to prioritize sampling based on capacity for processing posts, wherein the sampling scheduler directs post data to input at least a plurality of the plurality of filters, including processed post data from a previous filter, with the sampling scheduler configured to selectively provide instructions to skip one or more of the plurality of filters based on target rates of posts to sample,

wherein the plurality of filters include:

one or more firehose filters configured to perform a content search of all social media posts from one or more sources to determine a firehose post data output, including social media post data, to send to the sampling scheduler;

a single post filter configured to read post data from a single post input to process one or more features from a single social media post object to determine relevant social media posts to include in a single post data output to send to the sampling scheduler;

a contextual filter configured to read post data from a contextual input to gather and process additional data beyond data processed by the single post filter to augment classification and to include in a contextual post data output to send to the sampling scheduler; and

a target population filter configured to read post data from a population input to determine if one or more users are to be identified for a target social media user population such that one or more identified users with one or more probability scores are stored in a target candidate pool with other identified users to identify user relevance for additional monitoring or review.

2. The system of claim 1 , wherein the target population filter is executed prior to execution of the single post filter.

3. The system of claim 1 , wherein the target population filter is executed after the execution of the contextual filter.

4. The system of claim 1 , wherein the content search of the firehose filter is configured to perform keyword or phrase searches.

5. The system of claim 1 , further including a triage process.

6. The system of claim 5 , wherein the triage process includes one or more of a triage queue, a triage database, and an action database.

7. The system of claim 1 , further including a clustering filter configured to filter out non-relevant posts.

8. The system of claim 1 , wherein the single post filter includes one or more of a single post feature extractor, a single post classifier, and a single post classifier learning algorithm.

9. The system of claim 1 , wherein the contextual filter includes one or more of a contextual data harvester, a contextual feature extractor, a contextual classifier, and a contextual classifier learning algorithm.

10. The system of claim 1 , wherein the target population filter includes one or more of a non-relational user classifier learning algorithm, a non-relational user classifier, and a non-relational user feature extractor.

11. The system of claim 1 , further including a candidate pool update feature including one or more of a new candidate algorithm, a relational user feature extractor, and a relational user classifier learning algorithm.

12. A software method, having a plurality of filters, for social media account monitoring, comprising:

executing a sampling scheduler configured to prioritize sampling based on capacity for processing posts, wherein the sampling scheduler directs post data to input at least a plurality of the plurality of filters, including processed post data from a previous filter, with the sampling scheduler configured to selectively provide instructions to skip one or more of the plurality of filters based on target rates of posts to sample,

selectively executing one or more firehouse filters configured to perform a content search all social media posts from one or more sources to determine a firehose post data output, including social media post data, to send to the sampling scheduler;

selectively executing a single post filter configured to read post data from a single post input to process one or more features from a single social media post object to determine relevant social media posts to include in a single post data output to send to the sampling scheduler;

selectively executing a contextual filter configured to read post data from a contextual input to gather and process additional data beyond data processed by the single post filter to augment classification and to include in a contextual post data output to send to the sampling scheduler; and

selectively executing a target population filter configured to read post data from a population input to determine if one or more users are to be identified for a target social media user population such that one or more identified users with one or more probability scores are stored in a target candidate pool with other identified users to identify user relevance for additional monitoring or review.

13. The method of claim 12 , wherein the target population filter is executed prior to execution of the single post filter.

14. The method of claim 12 , wherein the target population filter is executed after the execution of the contextual filter.

15. The method of claim 12 , wherein the content search of the firehose filter is configured to perform keyword or phrase searches.

16. The method of claim 12 , further including a triage process having one or more of a triage queue, a triage database, and an action database.

17. The method of claim 12 , wherein the single post filter includes one or more of a single post feature extractor, a single post classifier, and a single post classifier learning algorithm.

18. The method of claim 12 , wherein the contextual filter includes one or more of a contextual data harvester, a contextual feature extractor, a contextual classifier, and a contextual classifier learning algorithm.

19. The method of claim 12 , further including executing a clustering filter configured to filter out non-relevant posts.

20. A non-transitory computer readable medium having stored thereon program code having a plurality of filters and configured for social media account monitoring, comprising:

a sampling scheduler configured to prioritize sampling based on capacity for processing posts, wherein the sampling scheduler directs post data to input at least a plurality of the plurality of filters, including processed post data from a previous filter, with the sampling scheduler configured to selectively provide instructions to skip one or more of the plurality of filters based on target rates of posts to sample,

the plurality of filters including:

one or more firehose filters configured to perform a content search of all social media posts from one or more sources to determine a firehose post data output, including social media post data, to send to the sampling scheduler;

a single post filter configured to read post data from a single post input to process one or more features from a single social media post object to determine relevant social media posts to include in a single post data output to send to the sampling scheduler;

a contextual filter configured to read post data from a contextual input to gather and process additional data beyond data processed by the single post filter to augment classification and to include in a contextual post data output to send to the sampling scheduler; and

a target population filter configured to read post data from a population input to determine if one or more users are to be identified for a target social media user population such that one or more identified users with one or more probability scores are stored in a target candidate pool with other identified users to identify user relevance for additional monitoring or review.