Techniques for device configuration using messaging history information

1. A computer-implemented method, comprising:

determining a promotion contact list for a user account comprising a subset of a contact list for the user account containing contacts with whom a user of the user account has not sent a message to within a defined recency;

determining a predicted bidirectional communication interest for each contact on the promotion contact list for the user account;

determining a user predicted likelihood of future activeness for the user account comprising a prediction that the user account will have a defined level of messaging activity in a defined period of time;

determining a contact predicted likelihood of future activeness for each contact on the promotion contact list comprising a prediction that the contact will have a defined level of messaging activity in a defined period of time;

determining a ranking weight for each contact on the promotion contact list based on the predicted bidirectional communication interest for each contact on the promotion contact list, the user predicted likelihood of future activeness and the contact predicted likelihood of future activeness; and

ordering the promotion contact list for display based on the determined ranking weight for each contact on the promotion contact list.

2. The method of claim 1 , wherein the ranking weight for each contact on the promotion contact list weighs contacts with a higher contact predicted likelihood of future activeness more heavily where the user predicted likelihood of future activeness is higher, wherein the ranking weight for each contact on the promotion contact list weighs contacts with a higher contact predicted likelihood of future activeness less heavily where the user predicted likelihood of future activeness is lower.

3. The method of claim 1 , wherein contacts with a higher predicted communication interest have a higher ranking weight.

4. The method of claim 1 , wherein the user predicted likelihood of future activeness is based on a number of threads for the user account, a number of sends for the user account, and a number of highly-active threads for the user account, wherein the contact predicted likelihood of future activeness on the communication system for each contact is based on a number of threads for each contact, a number of sends for each contact, and a number of highly-active threads for each contact.

5. The method of claim 1 , wherein the user predicted likelihood of future activeness is a function of a number of threads for the user account, a number of sends for the user account, and a number of highly-active threads for the user account, wherein the contact predicted likelihood of future activeness on the communication system for each contact is a function of a number of threads for each contact, a number of sends for each contact, and a number of highly-active threads for each contact.

6. The method of claim 5 , the function determined based on a linear regression of an historical data set for the communication system.

7. The method of claim 1 , the ranking weight reduced for one or more previously-viewed contacts on the promotion contact list.

8. The method of claim 1 , further comprising:

associating a contact of the promotion contact list with an event badge, the event badge selected based on event information for the contact.

9. An apparatus, comprising:

a processor circuit on a device; and

memory containing instructions that, when executed, cause the processor to:

determine a promotion contact list for the user account comprising a subset of a contact list for the user account containing contacts with whom a user of the user account has not sent a message to within a defined recency;

determine a bidirectional predicted communication interest for each contact on the promotion contact list;

determine a user predicted likelihood of future activeness for the user account comprising a prediction that the user account will have a defined level of messaging activity in a defined period of time;

determine a contact predicted likelihood of future activeness for each contact on the promotion contact list comprising a prediction that the contact will have a defined level of messaging activity in a defined period of time;

determine a ranking weight for each contact on the promotion contact list based on the predicted bidirectional communication interest for each contact on the promotion contact list, the user predicted likelihood of future activeness and the contact predicted likelihood of future activeness; and

transmit an ordered promotion contact list to the client device ordered for display based on a determined ranking weight for each contact on the ordered promotion contact list.

10. The apparatus of claim 9 , wherein the ranking weight for each contact on the promotion contact list weighs contacts with a higher contact predicted likelihood of future activeness more heavily where the user predicted likelihood of future activeness is higher, wherein the ranking weight for each contact on the promotion contact list weighs contacts with a higher contact predicted likelihood of future activeness less heavily where the user predicted likelihood of future activeness is lower, wherein contacts with a higher predicted communication interest have a higher ranking weight.

11. The apparatus of claim 9 , wherein the user predicted likelihood of future activeness is based on a number of threads for the user account, a number of sends for the user account, and a number of highly-active threads for the user account, wherein the contact predicted likelihood of future activeness on the communication system for each contact is based on a number of threads for each contact, a number of sends for each contact, and a number of highly-active threads for each contact.

12. The apparatus of claim 9 , wherein the user predicted likelihood of future activeness is a function of a number of threads for the user account, a number of sends for the user account, and a number of highly-active threads for the user account, wherein the contact predicted likelihood of future activeness on the communication system for each contact is a function of a number of threads for each contact, a number of sends for each contact, and a number of highly-active threads for each contact, the function determined based on a linear regression of an historical data set for the communication system.

13. The apparatus of claim 9 , the ranking weight reduced for one or more previously-viewed contacts of the promotion contact list.

14. The apparatus of claim 9 , the instructions further causing the processor to:

associate a contact of the promotion contact list with an event badge, the event badge selected based on event information for the contact.

15. At least one computer-readable storage medium comprising instructions that, when executed, cause a system to:

determine a promotion contact list for a user account comprising a subset of a contact list for the user account containing contacts with whom a user of the user account has not send a message to within a defined recency;

determine a predicted bidirectional communication interest for each contact on the promotion contact list;

determine a user predicted likelihood of future activeness comprising a prediction that the user account will have a defined level of messaging activity in a defined period of time;

determine a contact predicted likelihood of future activeness for each contact on the promotion contact list comprising a prediction that the contact will have a defined level of messaging activity in a defined period of time;

determine a ranking weight for each contact on the promotion contact list based on the predicted bidirectional communication interest for each contact on the promotion contact list, the user predicted likelihood of future activeness, and the contact predicted likelihood of future activeness; and

order the promotion contact list for display based on the determined ranking weight for each contact on the promotion contact list.

16. The computer-readable storage medium of claim 15 , wherein the ranking weight for each contact on the promotion contact list weighs contacts with a higher contact predicted likelihood of future activeness more heavily where the user predicted likelihood of future activeness is higher, wherein the ranking weight for each contact on the promotion contact list weighs contacts with a higher contact predicted likelihood of future activeness less heavily where the user predicted likelihood of future activeness is lower, wherein contacts with a higher predicted communication interest have a higher ranking weight.

17. The computer-readable storage medium of claim 15 , wherein the user predicted likelihood of future activeness is based on a number of threads for the user account, a number of sends for the user account, and a number of highly-active threads for the user account, wherein the contact predicted likelihood of future activeness on the communication system for each contact is based on a number of threads for each contact, a number of sends for each contact, and a number of highly-active threads for each contact.

18. The computer-readable storage medium of claim 15 , wherein the user predicted likelihood of future activeness is a function of a number of threads for the user account, a number of sends for the user account, and a number of highly-active threads for the user account, wherein the contact predicted likelihood of future activeness on the communication system for each contact is a function of a number of threads for each contact, a number of sends for each contact, and a number of highly-active threads for each contact, the function determined based on a linear regression of an historical data set for the communication system.

19. The computer-readable storage medium of claim 15 , the ranking weight reduced for one or more previously-viewed contacts of the promotion contact list.

20. The computer-readable storage medium of claim 15 , comprising further instructions that, when executed, cause a system to:

associate a contact of the promotion contact list with an event badge, the event badge selected based on event information for the contact.