Creating a machine learning policy based on express indicators

1. A method for creating a machine learning policy based on express indicators comprising:

receiving, by a scanner, at least one electronic message, the at least one electronic message outbound from at least one sender to at least one recipient;

determining, by the scanner, at least one classification type that applies to the at least one electronic message, the at least one classification type determined based on at least one or more express indications determined automatically prior to sending the at least one electronic message to the recipient, wherein determining the at least one classification type comprises scanning the at least one electronic message by a plurality of scanners associated with the scanner, each scanner configured to scan the at least one electronic message for at least one respective classification type of a plurality of classification types;

providing, by the scanner to a machine learning trainer, the at least one electronic message and at least one identification of the at least one classification type that applies to the at least one electronic message, the machine learning trainer adapted to determine a machine learning policy that associates attributes of the at least one electronic message with the at least one classification type;

receiving, by an enforcer, a second electronic message, the second electronic message comprising one or more of the attributes that the machine learning policy associates with the at least one classification type; and

enforcing, by the enforcer, handling the second electronic message in the manner that complies with compliance criteria associated with the at least one classification type, the enforcing based on the machine learning policy indicating that the at least one classification type applies to the second electronic message.

2. The method of claim 1 , wherein the express indication comprises a flag configured automatically by the scanner.

3. The method of claim 1 , wherein the express indication comprises a keyword that the scanner associates with enabling the classification.

4. The method of claim 1 , wherein the express indication comprises administrator feedback.

5. The method of claim 1 , further comprising:

determining that at least a threshold number of electronic messages have been provided to the machine learning trainer; and

enabling the machine learning policy based on the threshold number of electronic messages having been provided to the machine learning trainer.

6. The method of claim 1 , wherein the classification indicates whether at least one of the following applies to the electronic message: encryption, quarantine, archive, brand.

7. The method of claim 1 , wherein the classification indicates that best mode of delivery encryption applies to the electronic message.

8. The method of claim 1 , wherein the machine learning policy corresponds to one of a plurality of machine learning policies determined by the machine learning trainer, and wherein the method further comprises indicating to the machine learning trainer which of the machine learning policies to train with the electronic message.

9. The method of claim 1 , further comprising:

in response to determining that one or more senders have one or more characteristics in common, indicating to the machine learning trainer to train the same machine learning policy based on electronic messages received from different senders.

10. A non-transitory computer-readable medium comprising instructions to:

determine a classification that applies to an electronic message, the classification is an indication transmitted with the electronic message;

provide a machine learning trainer with the electronic message and an indication of the classification that applies to the electronic message, the message learning trainer adapted to determine a machine learning policy that associates attributes of the electronic message with the classification;

receive, by a scanner, at least a first electronic message, the first electronic message outbound from a sender to a recipient;

determine, by the scanner, at least a first classification type that applies to the first electronic message, the first classification type determined based on an express indication which is automatically determined prior to sending the first electronic message to the recipient, the express indication indicating that the first electronic message be handled in a manner that complies with a compliance criteria, wherein determining the first classification type comprises scanning the first electronic message by a plurality of scanners associated with the scanner, wherein each scanner is configured to scan the first electronic message for at least one respective classification type of a plurality of classification types;

provide, by the scanner to the machine learning trainer, the first electronic message and an indication of the at least one classification type that apply to the first electronic message, the machine learning trainer adapted to determine the machine learning policy that associates all the attributes of the first electronic message with the plurality of classification types;

receive, by an enforcer, at least a second electronic message, the second electronic message comprising the attributes that the machine learning policy associates with the first classification type; and

enforce, by the enforcer, handling the second electronic message in the manner that complies with the compliance criteria associated with the classification types associated with the first electronic message, the enforcing based on the machine learning policy indicating that the classification types associated with the first electronic message apply to the second electronic message.

11. The non-transitory computer-readable medium comprising of claim 10 , wherein the express indication comprises at least one of: a flag configured by a user to be automatically applied, a keyword that the scanner associates with enabling the classification, or administrator feedback.

12. The non-transitory computer-readable medium comprising of claim 10 , wherein the classification indicates whether at least one of the following applies to the electronic message: encryption, quarantine, archive, brand.

13. The non-transitory computer-readable medium comprising of claim 10 , wherein the machine learning policy corresponds to one of a plurality of machine learning policies determined by the machine learning trainer, and wherein the instructions are further adapted to indicate to the machine learning trainer which of the machine learning policies to train with the electronic message.

14. The non-transitory computer-readable medium comprising of claim 10 , further configured to indicate to the machine learning trainer to train the same machine learning policy based on electronic messages received from different users in response to determining that different users have one or more characteristics in common.