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, prior to a message being sent to an intended recipient, the message composed via a user interface of a messaging application, the message identifying the intended recipient and comprising content;
analyzing the message, via the computing device, prior to the message being sent to an intended recipient, and based on the analysis determining a feature vector, the feature vector representing the message;
applying, via the computing device, a classifier to the message using the feature vector;
determining, via the computing device, a set of likely actions of the intended recipient, each action in the set of likely actions being a predicted action the classifier identifies as a future action the intended recipient is likely to take on the message; and
using, via the computing device, the set of likely actions of the intended recipient of the message to make a determination whether to send the message to the recipient.
2. The method of claim 1 , further comprising:
sending, via the computing device, the message to the recipient based on the determination.
3. The method of claim 2 , further comprising:
updating, via the computing device, the classifier using information indicating that the message was sent to the recipient.
4. The method of claim 1 , the information identifying the set of likely actions of the intended recipient of the message comprising an action of forwarding the message.
5. The method of claim 1 , the message identifies multiple intended recipients, the set of likely actions of the intended recipient being determined for each of the multiple intended recipients, and the determination is made for each of the multiple recipients.
6. The method of claim 5 , further comprising:
sending the message to a subset of the multiple intended recipients of the message based on the determination made for each recipient of the subset of recipients.
7. The method of claim 2 , further comprising:
updating, via the computing device, the classifier using information indicating an action taken by the intended recipient to the message.
8. The method of claim 1 , the classifier takes into account activities the intended recipient has taken on other messages.
9. The method of claim 1 , the classifier is directed towards predicting how users might behave in general to the message.
10. The method of claim 1 , further comprising:
using, via the computing device, the feature vector and the classifier to determine a ranking for each action in the set of likely actions.
11. A non-transitory computer-readable storage medium tangibly encoded with computer-executable instructions that when executed by a processor associated with a computing device perform a method comprising:
receiving, prior to a message being sent to an intended recipient, the message composed via a user interface of a messaging application, the message identifying the intended recipient and comprising content;
analyzing the message prior to the message being sent to an intended recipient, and based on the analysis determining a feature vector, the feature vector representing the message;
applying a classifier to the message using the feature vector;
determining a set of likely actions of the intended recipient, each action in the set of likely actions being a predicted action the classifier identifies as a future action the intended recipient is likely to take on the message; and
using the set of likely actions of the intended recipient of the message to make a determination whether to send the message to the recipient.
12. The non-transitory computer-readable storage medium of claim 11 , the method further comprising:
sending the message to the recipient based on the determination.
13. The non-transitory computer-readable storage medium of claim 11 , the classifier takes into account activities the intended recipient has taken on other messages.
14. The non-transitory computer-readable storage medium of claim 11 , the classifier is directed towards predicting how users might behave in general to the message.
15. The non-transitory computer-readable storage medium of claim 11 , the method further comprising:
using the feature vector and the classifier to determine a ranking for each action in the set of likely actions.
16. A computing device comprising:
a processor; and
a non-transitory storage medium for tangibly storing thereon program logic for execution by the processor, the program logic comprising:
receiving logic executed by the processor for receiving, prior to a message being sent to an intended recipient, the message composed via a user interface of a messaging application prior to the message being sent by a sender, the message identifying the intended recipient and comprising content;
analyzing logic executed by the processor for analyzing the message prior to the message being sent to an intended recipient, and based on the analysis determining a feature vector, the feature vector representing the message;
applying logic executed by the processor for applying a classifier to the message using the feature vector;
determining logic executed by the processor for determining a set of likely actions of the intended recipient, each action in the set of likely actions being a predicted action the classifier identifies as a future action the intended recipient is likely to take on the message; and
using logic executed by the processor for using the set of likely actions of the intended recipient of the message to make a determination whether to send the message to the recipient.