Method and system for identifying and presenting important messages
In some aspects, the techniques described herein relate to a method including: receiving, by a computing device, a message corresponding to a user inbox and to be added to the user inbox; applying, by the computing device, prior to adding the message to the user inbox, a message classification model to content of the message to determine one or more classifications corresponding to the message; determining, by the computing device, that the message is an important message based on whether one or more of the classifications is one of a set of predetermined classifications; adding, by the computing device, metadata to the message, the added metadata indicating that the message is an important message; and transmitting, to the user inbox, by the computing device, the message and the added metadata.
1. A method comprising:
receiving, by a computing device, a message corresponding to a user inbox and to be added to the user inbox;
applying, by the computing device, prior to adding the message to the user inbox, a message classification model to content of the message to determine one or more classifications corresponding to the message, the content including body text of the message and the message classification model comprising a multi-label predictive model;
determining, by the computing device, that the message is an important message based on whether one or more of the classifications is one of a set of predetermined classifications;
adding, by the computing device, metadata to the message, the added metadata indicating that the message is an important message; and
transmitting, to the user inbox, by the computing device, the message and the added metadata.
2. The method of claim 1 , wherein the message classification model determines the one or more classifications based on user feedback provided in relation to other messages.
3. The method of claim 1 , wherein the set of predetermined classifications are selected from classifications corresponding to messages in a corpus of messages that have a log-squashed ratio of open counts to delete counts above a predetermined threshold.
4. The method of claim 1 , wherein the set of predetermined classifications include classifications selected by the user.
5. The method of claim 1 , wherein the computing device transmits the message and the added metadata in response to a received query, the query requesting important messages.
6. The method of claim 5 , wherein the query includes query criteria comprising at least one classification corresponding to the message.
7. The method of claim 1 , further comprising:
receiving, by the computing device, feedback from the user corresponding to the message and the added metadata; and
modifying, by the computing device, the set of predetermined classifications based on the feedback.
8. A non-transitory computer-readable storage medium, the computer-readable storage medium including instructions that when executed by a computing device, cause the computing device to:
receive a message corresponding to a user inbox and to be added to the user inbox;
apply, prior to adding the message to the user inbox, a message classification model to content of the message to determine one or more classifications corresponding to the message, the content including body text of the message and the message classification model comprising a multi-label predictive model;
determine that the message is an important message based on whether one or more of the classifications is one of a set of predetermined classifications;
add metadata to the message, the added metadata indicating that the message is an important message; and
transmit, to the user inbox, the message and the added metadata.
9. The computer-readable storage medium of claim 8 , wherein the message classification model determines the one or more classifications based on user feedback provided in relation to other messages.
10. The computer-readable storage medium of claim 8 , wherein the set of predetermined classifications are selected from classifications corresponding to messages in a corpus of messages that have a log-squashed ratio of open counts to delete counts above a predetermined threshold.
11. The computer-readable storage medium of claim 8 , wherein the set of predetermined classifications include classifications selected by the user.
12. The computer-readable storage medium of claim 8 , wherein the computing device transmits the message and the added metadata in response to a received query, the query requesting important messages.
13. The computer-readable storage medium of claim 12 , wherein the query includes query criteria comprising at least one classification corresponding to the message.
14. The computer-readable storage medium of claim 8 , wherein the instructions further configure the computing device to:
receive feedback from the user corresponding to the message and the added metadata; and
modify the set of predetermined classifications based on the feedback.
15. A computing device comprising:
a processor configured to:
receive a message corresponding to a user inbox and to be added to the user inbox;
apply prior to adding the message to the user inbox, a message classification model to content of the message to determine one or more classifications corresponding to the message, the content including body text of the message and the message classification model comprising a multi-label predictive model;
determine that the message is an important message based on whether one or more of the classifications is one of a set of predetermined classifications;
add metadata to the message, the added metadata indicating that the message is an important message; and
transmit, to the user inbox, the message and the added metadata.
16. The computing device of claim 15 , wherein the message classification model determines the one or more classifications based on user feedback provided in relation to other messages.
17. The computing device of claim 15 , wherein the set of predetermined classifications are selected from classifications corresponding to messages in a corpus of messages that have a log-squashed ratio of open counts to delete counts above a predetermined threshold.
18. The computing device of claim 15 , wherein the set of predetermined classifications include classifications selected by the user.
19. The computing device of claim 15 , wherein the processor is further configured to transmit the message and the added metadata in response to a received query, the query requesting important messages, and wherein the query includes query criteria comprising at least one classification corresponding to the message.
20. The computing device of claim 15 , wherein the processor is further configured to:
receive feedback from the user corresponding to the message and the added metadata; and
modify the set of predetermined classifications based on the feedback.