Method for data protection across sharing platforms

1. A method comprising:

during a first time period:

accessing a first corpus of messages sent from a first user account;

correlating sequences of words, in messages of the first corpus of messages, with a first set of behavior signals;

aggregating the first set of behavior signals into a first behavioral model representing combinations of behavior signals, in the first set of behavior signals, characteristic of behavior in messages sent from the first user account; and

associating the first behavioral model with the first user account;

during a second time period succeeding the first time period:

accessing a first message outbound from the first user account to a first recipient account, the first message comprising a first document as an attachment to the first message;

correlating sequences of words, in the first message, with a second set of behavior signals;

accessing a first set of document tags associated with the first document;

retrieving a first data access policy:

associated with a first document tag in the first set of document tags; and

comprising a first threshold at which access to a document associated with the first document tag is restricted; and

in response to detecting a difference between the second set of behavioral signals and the first behavioral model exceeding the first threshold, restricting access, by the first recipient account, to the first document in the first message; and

during a third time period succeeding the second time period:

accessing a second message outbound from the first user account to a second recipient account, the second message comprising the first document as an attachment to the second message;

correlating sequences of words, in the second message, with a third set of behavior signals;

accessing the first set of document tags associated with the first document;

retrieving the first data access policy; and

in response to detecting a difference between the third set of behavioral signals and the first behavioral model falling below the first threshold, permitting transmission of the second message to the second recipient account.

2. The method of claim 1 , further comprising, during the second period:

calculating a risk score for the first message based on the difference between the second set of behavior signals and the first behavioral model;

in response to detecting the risk score exceeding the first threshold, restricting access, by the first recipient account, to the first document in the first message; and

in response to detecting the risk score falling below the first threshold, permitting access, by the first recipient account, to the first document in the first message.

3. The method of claim 1 :

wherein the first data access policy further comprises a second threshold at which a notification to an administrator of a document associated with the first document tag is generated; and

further comprising, in response to detecting the difference between the second set of behavioral signals and the first behavioral model exceeding the second threshold, generating a notification, to an administrator of the first document, comprising an indication of the first message, from the first user account, comprising the first document.

4. The method of claim 1 :

wherein the first data access policy further comprises a rule requiring encryption of an outbound message comprising a document, associated with the first document tag, as an attachment to the outbound message; and

further comprising, in response to detecting the difference between the second set of behavioral signals and the first behavioral model falling below the first threshold, encrypting the first message, to the first recipient account, comprising the first document as an attachment to the first message.

5. The method of claim 1 , further comprising, during the second time period:

aggregating the second set of behavior signals into the first behavioral model representing combinations of behavior signals, in the first set of behavior signals and the second set of behavior signals, characteristic of behavior in messages sent from the first user account; and

associating the first behavioral model with the first user account.

6. The method of claim 1 , wherein accessing a first set of document tags associated with the first document comprises accessing the first document tag representing a first sensitive information signal in the first document.

7. The method of claim 1 :

wherein the first data access policy further comprises a first set of identities permitted to access a document associated with the first document tag, the first set of identities comprising the first user account; and

further comprising, in response to detecting the difference between the second set of behavioral signals and the first behavioral model exceeding the first threshold, modifying the first data access policy excluding the first recipient account from the first set of identities.

8. The method of claim 1 :

wherein the first data access policy comprises a first set of identity characteristics, comprising a first identity characteristic, of a user account permitted to access a document associated with the first document tag;

further comprising, during the second period:

retrieving a second set of identity characteristics, comprising the first identity characteristic, associated with the first user account;

in response to detecting the difference between the second set of behavioral signals and the first behavioral model exceeding the first threshold:

generating a second data access policy comprising a set of identities, comprising the first user account, restricted from accessing a document associated with a document tag representing a sensitive information signal; and

associating the second data access policy with the first document tag; and

in response to detecting the difference between the second set of behavioral signals and the first behavioral model falling below the first threshold, and in response to detecting an intersection between the first set of identity characteristics and the second set of identity characteristics, permitting access, by the first user account, to the first document; and

further comprising, during a fourth period succeeding the second period, restricting access, by the first user account, to the first document based on the second data access policy.

9. The method of claim 1 :

wherein correlating sequences of words in messages of the first corpus of messages, with a first set of behavior signals comprises:

scanning bodies of messages, in the first corpus of messages, for the first set of behavior signals comprising:

financial signals;

sensitive information signals;

action signals;

urgency signals;

deadline signals;

keyword signals; and

syntax signals

wherein aggregating the first set of behavior signals into the first behavioral model comprises:

calculating a first n-dimensional vector representing average frequencies of financial signals, sensitive information signals, action signals, urgency signals, deadline signals, keyword signals, and syntax signals in each message in the first corpus of messages; and

associating the first n-dimensional vector with the first behavioral model;

wherein correlating sequences of words in the first message with the second set of behavior signals comprises:

scanning the first message for the second set of behavior signals comprising:

financial signals;

sensitive information signals;

action signals;

urgency signals;

deadline signals;

keyword signals; and

syntax signals; and

further comprising, during the second time period:

calculating a second n-dimensional vector representing frequencies of financial signals, sensitive information signals, action signals, urgency signals, deadline signals, keyword signals, and syntax signals in the first message; and

in response to detecting a difference between the second n-dimensional vector and the first n-dimensional vector exceeding the first threshold, restricting access, by the first recipient account, to the first document in the first message.

10. The method of claim 9 , further comprising:

recalculating the first n-dimensional vector based on the second n-dimensional vector; and

in response to recalculating the first n-dimensional vector, associating the first n-dimensional vector with the first behavioral model.

11. The method of claim 1 :

further comprising, during the second time period:

detecting a first workflow based on:

the first user account;

the second set of behavior signals;

the first set of document tags; and

the first recipient account; and

wherein retrieving the first set of data access policies comprises retrieving the first set of data access policies:

associated with the first set of document tags; and

corresponding to the first workflow.

12. The method of claim 1 , further comprising:

during the first time period:

accessing a second corpus of messages sent from a group of user accounts comprising the first user account;

correlating sequences of words, in messages of the second corpus of messages, with a fourth set of behavior signals;

aggregating the fourth set of behavior signals into a group behavioral model representing combinations of behavior signals, in the fourth set of behavior signals, characteristic of behavior in messages sent from the group of user accounts; and

associating the group behavioral model with the group of user accounts; and

during the second time period, in response to detecting a difference between the second set of behavioral signals and the group behavioral model exceeding the first threshold, restricting access, by the first recipient account, to the first document in the first message.

13. The method of claim 1 , further comprising:

during the first time period, in response to detecting a first number of messages in the first corpus of messages falling below a second threshold quantity:

accessing a second corpus of messages sent from a group of user accounts, comprising the first user account, each user account in the group of user accounts exhibiting a particular identity characteristic;

correlating sequences of words, in messages of the second corpus of messages, with a fourth set of behavior signals;

aggregating the fourth set of behavior signals into the first behavioral model representing combinations of behavior signals, in a combination of the first set of behavior signals and the fourth set of behavior signals, characteristic of behavior in messages sent from the group of user accounts; and

associating the first behavioral model with the first user account.

14. The method of claim 13 , further comprising:

during a fourth time period succeeding the second time period, in response to detecting a second number of messages in the first corpus of messages exceeding a third threshold quantity:

accessing the first corpus of messages;

correlating sequences of words, in messages of the first corpus of messages, with a fifth set of behavior signals; and

aggregating the fifth set of behavior signals into the first behavioral model representing combinations of behavior signals, in the fifth set of behavior signals, characteristic of behavior in messages sent from the first user account; and

associating the first behavioral model with the first user account.

15. The method of claim 13 , wherein accessing a second corpus of messages sent from a group of user accounts comprises accessing the second corpus of messages sent from the group of user accounts exhibiting the particular identity characteristic comprising finance department membership within an organization.

16. The method of claim 1 , further comprising:

during the first time period:

accessing a first subset of messages, of the first corpus of messages, sent from the first user account to the first recipient account;

correlating sequences of words, in messages of the first subset of messages, with a fourth set of behavior signals;

aggregating the fourth set of behavior signals into a second behavioral model representing combinations of behavior signals, in the fourth set of behavior signals, characteristic of behavior in messages sent from the first user account to the first recipient account; and

associating the second behavioral model with the first user account; and

during the second time period, in response to detecting a difference between the second set of behavior signals and the second behavioral model exceeding the first threshold, restricting access, by the first recipient account, to the first document in the first message.

17. The method of claim 1 , wherein restricting access, by the first recipient account, to the first document in the first message comprises restricting transmission of the first message to the first recipient account.

18. A method comprising:

during a first time period:

accessing a first corpus of messages sent from a first user account;

correlating sequences of words, in messages of the first corpus of messages, with a first set of behavior signals;

aggregating the first set of behavior signals into a first behavioral model representing combinations of behavior signals, in the first set of behavior signals, characteristic of behavior in messages sent from the first user account;

in response to detecting a first number of messages in the first corpus of messages falling below a second threshold quantity:

accessing a second corpus of messages sent from a group of user accounts, comprising the first user account, each user account in the group of user accounts exhibiting a particular identity characteristic;

correlating sequences of words, in messages of the second corpus of messages, with a second set of behavior signals; and

aggregating the second set of behavior signals into the first behavioral model representing combinations of behavior signals, in a combination of the first set of behavior signals and the second set of behavior signals, characteristic of behavior in messages sent from the group of user accounts; and

associating the first behavioral model with the first user account; and

during a second time period succeeding the first time period:

accessing a first message outbound from the first user account to a first recipient account, the first message comprising a first document as an attachment to the first message;

correlating sequences of words, in the first message, with a third set of behavior signals;

accessing a first set of document tags associated with the first document;

retrieving a first data access policy:

associated with a first document tag in the first set of document tags; and

comprising a first threshold at which access to a document associated with the first document tag is restricted; and

in response to detecting a difference between the third set of behavioral signals and the first behavioral model exceeding the first threshold, restricting access, by the first recipient account, to the first document in the first message.

19. A method comprising:

during a first time period:

accessing a first corpus of messages sent from a first user account;

correlating sequences of words, in messages of the first corpus of messages, with a first set of behavior signals;

aggregating the first set of behavior signals into a first behavioral model representing combinations of behavior signals, in the first set of behavior signals, characteristic of behavior in messages sent from the first user account;

associating the first behavioral model with the first user account;

accessing a first subset of messages, of the first corpus of messages, sent from the first user account to the first recipient account;

correlating sequences of words, in messages of the first subset of messages, with a second set of behavior signals;

aggregating the second set of behavior signals into a second behavioral model representing combinations of behavior signals, in the second set of behavior signals, characteristic of behavior in messages sent from the first user account to the first recipient account; and

associating the second behavioral model with the first user account; and

during a second time period succeeding the first time period:

accessing a first message outbound from the first user account to a first recipient account, the first message comprising a first document as an attachment to the first message;

correlating sequences of words, in the first message, with a third set of behavior signals;

accessing a first set of document tags associated with the first document;

retrieving a first data access policy:

associated with a first document tag in the first set of document tags; and

comprising a first threshold at which access to a document associated with the first document tag is restricted; and

in response to detecting a difference between the third set of behavioral signals and the first behavioral model exceeding the first threshold, and in response to detecting a difference between the third set of behavior signals and the second behavioral model exceeding the first threshold, restricting access, by the first recipient account, to the first document in the first message.