Mining email inboxes for suggesting actions
Embodiments are directed towards automatically learning user behavioral patterns when interacting with messages and based on the learned patterns, suggesting one or more predicted actions that a user might take in response to receiving subsequent message. One or more classifiers are trained and employed to predict one or more actions that a user might take in response to receiving the message. In one embodiment, the one or more predicted actions are provided suggested to the user as an action the user might take on the message. Messages may be rank ordered within a given suggested action based on a confidence level of the prediction.
1. A method comprising:
receiving, at a computing device, a message in an inbox of a user, said message sent from a sender and comprising content;
analyzing, via the computing device, said message, and based on said analysis, determining a feature vector, said feature vector representing said content of said message;
determining, via the computing device, a type of said message based on said feature vector; applying, via the computing device, a classifier to said message, said classifier being specific to said type of message;
determining, via the computing device, a set of actions based on said application of the classifier, each action being a predicted action said classifier identifies as a future action said user will take on said message;
automatically causing display, via the computing device, within a message display portion of the inbox, of an interface object, the interface object being displayed adjacent to said message within said message display portion, the interface object enabling selection of said set of actions without opening said message;
receiving, by the computing device, a selection of an action from said set of actions via said interface object, wherein the selection of the action comprises replying to said message or forwarding said message;
in response to the selection of the action, causing said action to be performed on said message,
and
identifying and displaying a set of candidate recipients based on a type of the selected action.
2. The method of claim 1 , wherein said classifier is a vertical classifier that accounts for activities the user has taken on other messages of said type.
3. The method of claim 2 , further comprising: determining that the user is not a new user, and has not received messages from the sender previously, and employing based on the determination, said vertical classifier.
4. The method of claim 1 , wherein said classifier is a horizontal classifier that accounts for activities other users have taken on other messages of said type.
5. The method of claim 4 , further comprising: determining whether the user is a new user, and if that is true, then employing said horizontal classifier.
6. The method of claim 1 , wherein said classifier is a vertical pair-wise classifier that accounts for activities the user has taken on other messages from said sender.
7. The method of claim 6 , further comprising: determining whether the user has received messages from said sender previously, and if that is true, then employing said vertical pair-wise classifier.
8. The method of claim 1 , wherein each action in said set of actions is weighted based on a level of confidence determined by said classifier.
9. The method of claim 1 , wherein at least one of the actions in the set of actions is directed towards preserving a conversation associated with the message.
10. A non-transitory computer-readable storage medium tangibly encoded with computer-executable instructions, that when executed by a computing device, perform a method comprising:
receiving, at the computing device, a message in an inbox of a user, said message sent from a sender and comprising content;
analyzing, via the computing device, said message, and based on said analysis, determining a feature vector, said feature vector representing said content of said message;
determining, via the computing device, a type of said message based on said feature vector; applying, via the computing device, a classifier to said message, said classifier being specific to said type of message;
determining, via the computing device, a set of actions based on said application of the classifier, each action being a predicted action said classifier identifies as a future action said user will take on said message,
automatically causing display, via the computing device, within a message display portion of the inbox, of an interface object, the interface object being displayed adjacent to said message within said message display portion, the interface object enabling selection of said set of actions without opening said message;
receiving, by the computing device, a selection of an action from said set of actions via said interface object, wherein the selection of the action comprises replying to said message or forwarding said message;
in response to the selection of the action, causing said action to be performed on said message, and
identifying and displaying a set of candidate recipients based on a type of the selected action.
11. The non-transitory computer-readable storage medium of claim 10 , wherein said classifier is a vertical classifier that accounts for activities the user has taken on other messages of said type.
12. The non-transitory computer-readable storage medium of claim 11 , further comprising: determining that the user is not a new user, and has not received messages from the sender previously, and employing based on the determination, said vertical classifier.
13. The non-transitory computer-readable storage medium of claim 10 , wherein said classifier is a horizontal classifier that accounts for activities other users have taken on other messages of said type.
14. The non-transitory computer-readable storage medium of claim 13 , further comprising: determining whether the user is a new user, and if that is true, then employing said horizontal classifier.
15. The non-transitory computer-readable storage medium of claim 10 , wherein said classifier is a vertical pair-wise classifier that accounts for activities the user has taken on other messages from said sender.
16. The non-transitory computer-readable storage medium of claim 15 , further comprising: 5 determining whether the user has received messages from said sender previously, and if that is true, then employing said vertical pair-wise classifier.
17. The non-transitory computer-readable storage medium of claim 10 , wherein each action in said set of actions is weighted based on a level of confidence determined by said classifier.
18. A computing device comprising:
a processor; and
a non-transitory computer-readable storage medium for tangibly storing thereon program logic for execution by the processor, the program logic comprising:
logic executed by the processor for receiving, at the computing device, a message in an inbox of a user, said message sent from a sender and comprising content;
logic executed by the processor for analyzing, via the computing device, said message, and based on said analysis, determining a feature vector, said feature vector representing said content of said message;
logic executed by the processor for determining, via the computing device, a type of said message based on said feature vector; logic executed by the processor for applying, via the computing device, a classifier to said message, said classifier being specific to said type of message;
logic executed by the processor for determining, via the computing device, a set of actions based on said application of the classifier, each action being a predicted action said classifier identifies as a future action said user will take on said message,
logic executed by the processor for automatically causing display, via the computing device, within a message display portion of the inbox, of an interface object, the interface object being displayed adjacent to said message within said message display portion, the interface object enabling selection of said set of actions without opening said message:
logic executed by the processor for receiving, by the computing device, a selection of an action from said set of actions via said interface object, wherein the selection of the action comprises replying to said message or forwarding said message; and
logic executed by the processor for causing said action to be performed on said message, in response to the selection of the action, and
identifying and displaying a set of candidate recipients based on a type of the selected action.