Personalized delivery time optimization

Patent metadata

Field	Value
Publication number	US-9967226-B2
Application number	US-201615289836-A
Country	US
Kind code	B2
Filing date	Oct 10, 2016
Priority date	Jun 30, 2014
Publication date	May 8, 2018
Grant date	May 8, 2018

Abstract

Techniques for optimizing a delivery time for the delivery of messages are described. According to various embodiments, a system determines, for each of a plurality of time intervals, a likelihood of a particular member of an online social network service performing a particular member user action on a particular message content item during the corresponding time interval. The plurality of time intervals are then ranked, based on the determined likelihoods corresponding to the plurality of time intervals. Thereafter, a particular time interval is identified from among the plurality of time intervals that is associated with a highest ranking. The particular time interval is then classified as an optimum personalized message delivery time for the particular member.

First claim

What is claimed is:

1. A computer-implemented method comprising:
determining by a machine including a memory and at least one processor, for each of a plurality of time intervals, a likelihood of a particular member of an online social network service performing a particular member user action on a particular message content item during the corresponding time interval;
ranking the plurality of time intervals, based on the determined likelihoods corresponding to the plurality of time intervals;
identifying a particular time interval from among the plurality of time intervals that is associated with a highest ranking; and
classifying the particular time interval as an optimum personalized message delivery time for the particular member;
wherein the determining comprises:
accessing, via one or more data sources, data including:
email content data describing a particular email content item and member email Interaction data describing the particular member's interactions with various email content; and
geolocation data indicating a current location of a client device of the particular member at a time interval corresponding to the particular member's interaction with the various email content;
encoding the data accessed from the one or more data sources into one or more feature vectors, and assembling the one or more feature vectors to thereby generate an assembled feature vector; and
performing prediction modeling, based on the assembled feature vector and a trained prediction model, to predict the likelihood of the particular member performing the particular user action on the particular email content item.

2. The method of claim 1 , further comprising:
determining a time zone associated with the current location of the client device of the particular member; and
encoding the time zone into the assembled feature vector.

3. The method of claim 1 , wherein the email content data specifies an email type, and wherein the email type is at least one of a network connection update e-mail, a news update e-mail, a jobs update e-mail, an influencer post update e-mail, a company update e-mail, a group update e-mail, a university update e-mail, and a digest e-mail.

4. The method of claim 1 , wherein the member email interaction data indicates a quantity of various email types transmitted to the particular member, a quantity of clicks submitted by the particular member in conjunction with the various email types, and a quantity of email unsubscribe requests submitted by the particular member in conjunction with the various email types.

5. The method of claim 1 , further comprising:
accessing, via one or more data sources, member site interaction data describing the particular member's interaction with various features or content of an online social network service; and
encoding the member site interaction data into the assembled feature vector.

6. The method of claim 1 , further comprising:
accessing, via one or more data sources, member profile data describing the particular member; and
encoding the member profile data into the assembled feature vector.

7. The method of claim 1 , wherein the prediction model is any one of a logistic regression model, a Naïve Bayes model, a support vector machines (SVM) model, a decision trees model, and a neural network model.

8. The method of claim 1 , wherein the prediction modeling comprises a training operation to refine coefficients of a logistic regression model, based on training set data comprising the assembled feature vector.

9. A system comprising:
a processors; and
a memory comprising instructions that, when executed the processor, cause the system to:
determine, for each of a plurality of time intervals, a likelihood of a particular member of an online social network service performing a particular member user action on a particular message content item during the corresponding time interval;
rank the plurality of time intervals, based on the determined likelihoods corresponding to the plurality of time intervals;
identify a particular time interval from among the plurality of time intervals that is associated with highest ranking; and
classify the particular time interval as an optimum personalized message delivery time for the particular member;
wherein to determine the likelihood, the memory comprises instructions that, when executed by the processor, cause the system to:
access, via one or more data sources, data including:
email content data describing a particular email content item and member email interaction data describing the particular members interactions with various email content; and
geolocation data indicating a current location of a client device of the particular member at a time interval corresponding to the particular member's interactions with the various email content;

encode the data accessed from the one or more data sources into one or more feature vectors, and assembling the one or more feature vectors to thereby generate an assembled feature vector; and
perform prediction modeling, based on the assembled feature vector and a trained prediction model, to predict the likelihood of the particular member performing the particular user action on the particular email content item.

10. The system of claim 9 , further comprising instructions that, when executed by the processor, cause the system to:
determine a time zone associated with the current location of the client device of the particular member; and
encode the time zone into the assembled feature vector.

11. The system of claim 9 , wherein the email content data specifies an email type, and wherein the email is at least one of a network connection update e-mail, a news update e-mail, a jobs update e-mail, an influencer post update e-mail, a company update e-mail, a group update e-mail, a university update e-mail, and a digest e-mail.

12. The system of claim 9 , wherein the member email interaction data indicates a quantity of various email types transmitted to the particular member, a quantity of clicks submitted by the particular member in conjunction with the various email types, and a quantity of email unsubscribe requests submitted by the particular member in conjunction with the various email types.

13. The system of claim 9 , further comprising instructions that, when executed by the processor, cause the system to:
access, via one or more data sources, member site interaction data describing the particular member's interaction with various features or content of an online social network service; and
encode the member site interaction data into the assembled feature vector.

14. The system of claim 9 , further comprising instructions that, when executed by the processor, cause the system to:
access, via one or more data sources, member profile data describing the particular member; and
encode the member profile data into the assembled feature vector.

15. The system of claim 9 , wherein the prediction model is any one of a logistic regression model, a Naïve Bayes model, a support vector machines (SVM) model, a decision trees model, and a neural network model.

16. The system of claim 9 , wherein the prediction modeling comprises a training operation to refine coefficients of a logistics regression model, based on training set data comprising the assembled feature vector.

17. A non-transitory machine-readable storage medium comprising instructions that, when executed by one or more processors of a machine, cause the machine to perform operations comprising:
determining, for each of a plurality of time intervals, a li

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Vijay Ravi Kiran Holur
• Arai Benjamin
• Hull Mark
• Irmak Utku
• Khincha Pramod Chand
• Shah Samir M
• Yan Ji
• Yuan Lawrence

Classifications

• G06Q10/40

  Business processes related to social networking or social networking services · CPC title

• H04L63/102

  Entity profiles · CPC title

• G06F17/3053

  Physics · mapped topic

• H04L51/20

  Electricity · mapped topic

• H04L67/306

  User profiles · CPC title