Message content cleansing

1. An electronic computing device comprising:

a processing unit; and

system memory, the system memory including instructions which, when executed by the processing unit, cause the electronic computing device to:

monitor a message to determine whether a response is required based on compliance with regulatory email compliance rules, the monitoring of the message including to:

use one or more machine learning algorithms to automatically identify one or more parts of the message that are not relevant to monitoring based on the compliance with the regulatory email compliance rules, wherein the one or more machine learning algorithms comprise:

one or more unsupervised machine learning algorithms used to identify noise items within the message by developing rules to look for patterns; and

one or more supervised machine learning algorithms used to identify a location for the noise items within the message;

automatically exclude the one or more parts of the message that are not relevant from the monitoring of the message based on the compliance with the regulatory email compliance rules; and

identify content in the message that is not excluded from the monitoring of the message and that can require a response based on the compliance with the regulatory email compliance rules; and

when the content that can require a response based on the compliance with the regulatory email compliance rules is discovered in a section of the message other than the one or more parts of the message that are not relevant, automatically designate the message for further analysis.

2. The electronic computing device of claim 1 , wherein the processing unit further causes the electronic computing devices to create rules to detect the one or more parts of the message that are not relevant.

3. The electronic computing device of claim 2 , wherein the rules can be adjusted based on an analysis of the one or more messages.

4. The electronic computing device of claim 1 , wherein automatically identify one or more parts of the message that are not relevant to the monitoring of the message based on the compliance with the regulatory email compliance rules comprises identifying a uniform resource locator (URL).

5. The electronic computing device of claim 1 , wherein automatically identify one or more parts of the message that are not relevant to the monitoring of the message based on the compliance with the regulatory email compliance rules comprises identifying one or more echoes of earlier messages in the message.

6. The electronic computing device of claim 5 , wherein the message is an email message and identifying one or more echoes of earlier messages in the message comprises identifying one or more symbols indicating replies and forwards in the email message.

7. The electronic computing device of claim 1 , wherein the unsupervised machine learning algorithms include one or more of n-gram and k-means clustering and wherein the one or more supervised machine learning algorithms include a Naïve Bayes classifier.

8. The electronic computing device of claim 1 , further comprising automatically validating the one or more parts of the message that are not excluded from the monitoring of the message to verify that there are not any additional parts of the message that should be excluded from the monitoring.

9. The electronic computing device of claim 1 , further comprising automatically identifying a start and end position in the message for each part of the message identified not to be relevant.

10. The electronic computing device of claim 1 , wherein identify one or more keywords or phrases in a part of the message that is not excluded from the monitoring of the message further comprise identifying text indicative of a customer complaint or employee misconduct.

11. The electronic computing device of claim 1 , wherein the one or more parts of the message that are not relevant to the monitoring are identified using a rules-based review of the message.

12. A method implemented on an electronic computing device for monitoring email messages based on email compliance regulations, the method comprising:

on the electronic computing device, using one or more machine learning algorithms, automatically identifying one or more parts of an email message that are not relevant to the monitoring of the email message for violations based on the email compliance regulations, wherein the one or more machine learning algorithm includes:

one or more unsupervised machine learning algorithms used to identify noise items within the email message by developing rules to look for patterns; and

one or more supervised machine learning algorithms used to identify a location for the noise items within the email message;

automatically excluding the one or more parts of the email message that are not relevant from the monitoring of the email message based on the email compliance regulations;

identifying one or more keywords or phrases in a part of the email message that is not excluded from the monitoring of the email message; and

checking the email message for the one more keywords or phrases; and

when one or more of the keywords or phrases are discovered in the part of the email message that is not excluded from the monitoring of the email message, automatically designate the email message for further analysis.

13. The method of claim 12 , further comprising creating rules to detect the one or more parts of the message that are not relevant.

14. The method of claim 12 , wherein automatically identifying one or more parts of the email message that are not relevant to the monitoring of the email message based on the email compliance regulations comprises identifying one or more of a signature block, a uniform resource locator (URL) and a disclaimer in the email message.

15. The method of claim 12 , wherein automatically identifying one or more parts of the email message that are not relevant to the monitoring of the email message based on the email compliance regulations comprises identifying one or more echoes of earlier email messages in the email message.

16. The method of claim 15 , wherein identifying one or more echoes of earlier email messages in the email message comprises identifying one or more symbols indicating replies and forwards in the email message.

17. The method of claim 12 , wherein the one or more machine learning algorithms comprise one or more of n-gram clustering, k-means clustering and a Naïve Bayes classifier.

18. The method of claim 12 , wherein the one or more machine learning algorithms uses organization specific information, including a signature block and a disclaimer, to identify the one or more parts of the message that is not relevant to the monitoring.

19. An electronic computing device comprising:

a processing unit; and

system memory, the system memory including instructions which, when executed by the processing unit, cause the electronic computing device to:

monitor an email message to determine whether a response is required based on email compliance regulations, the monitoring of the email message comprising:

automatically identify one or more parts of the email message that are not relevant to the monitoring of the email message based on the email compliance regulations, the one or more parts including one or more of a signature block, a uniform resource locator (URL), a disclaimer and one or more echoes of previous emails in the email message, the one or more parts being identified using one or more machine learning algorithms, the one or more machine learning algorithms including:

one or more unsupervised learning algorithms used to identify noise items within the email message by developing rules to look for patterns; and

one or more supervised learning algorithms used to identify a location for the noise items within the email message;

automatically exclude the one or more parts of the email message that are not relevant from the monitoring of the email message based on the email compliance regulations;

automatically identify one or more keywords or phrases in a part of the email message that is not excluded from the monitoring of the email message, the one or more keywords or phrases indicating possible action that needs to be taken to comply with the email compliance regulations;

automatically identify a start and end position in the email message for each part of the email message identified to be relevant; and

check the email message for the one more keywords or phrases; and

when one or more of the keywords or phrases are discovered in the part of the email message that is not excluded from the monitoring of the email message, automatically designate the email message for further analysis.