Prioritizing messages within a message network
A system and a method are disclosed for recommending electronic messages in a message sharing system. Users can post messages to the message sharing system. These messages from posting users are received by the system and sent to receiving users that have subscribed to the posting users. The receiving users interact with the messages in various ways, such as by sharing the messages with other users. Interaction information is received for each of the electronic messages. The interaction information includes an indication of the number of interactions with the electronic message by receiving users. A score is determined for each electronic message based on the interaction information. Electronic messages are selected for being recommended to a user or a group of users based on the scores. The recommendations are then sent to the users, enabling users to better focus their attention on messages that are likely to be interesting.
1. A system comprising:
one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
receiving a message authored by a posting user in a social messaging system having a plurality of users, wherein the message is a candidate for presentation to a requesting user who requests a plurality of messages to be provided by the social messaging system;
obtaining a number of subscribers for the posting user, the number of subscribers for the posting user representing how many users in the social messaging system have subscribed to receive messages from the posting user;
computing an expected interaction level of subscribers of the posting user for the message according to the number of subscribers for the posting user;
obtaining observed interaction data for the message, the observed interaction data representing interactions with the message by one or more subscribers of the posting user to whom the message has been presented;
computing an observed interaction level of subscribers of the posting user based on the observed interaction data for the message;
computing an interaction level deviation between the observed interaction level of subscribers of the posting user and the expected interaction level of subscribers of the posting user;
computing a score for the message representing how likely it is that subscribers of the posting user in the social messaging system will interact with the message, the score being based on the computed interaction level deviation between the observed interaction level of subscribers of the posting user and the expected interaction level of subscribers of the posting user;
ranking the message relative to a plurality of other messages based on the computed score for the message, the plurality of other messages being candidates for presentation to the requesting user; and
providing one or more highest-ranked messages from among the plurality of other messages and the candidate message, including the candidate message, to a user device for presentation to the requesting user.
2. The system of claim 1 , wherein the computing the expected interaction level of subscribers of the posting user for the message according to the number of subscribers for the posting user comprises providing the number of subscribers for the posting user as input to a function that assigns an expected interaction level based on previous observations of other users interacting with messages posted by a plurality of posting users.
3. The system of claim 2 , wherein the function defines a plurality of ranges and assigns a respective expected number of interactions to each range of the plurality of ranges.
4. The system of claim 1 , wherein the observed interaction data for the message comprises a number of subscribers of the posting user who selected the message after the message was presented.
5. The system of claim 1 , wherein the observed interaction data for the message comprises a number of subscribers of the posting user who shared the message after the message was presented.
6. The system of claim 1 , wherein the observed interaction data for the message comprises a number of subscribers of the posting user who tagged the message as a favorite after the message was presented.
7. The system of claim 1 , wherein computing a score for the message comprises:
applying a first weight to the computed interaction level deviation;
applying a second weight to a particular attribute value of the message; and
computing a weighted sum resulting from applying the first weight to the computed interaction level deviation and applying the second weight to the particular attribute value of the message.
8. The system of claim 7 , wherein the particular attribute value of the message is a number of subscribers to the posting user.
9. A computer-implemented method comprising:
receiving a message authored by a posting user in a social messaging system having a plurality of users, wherein the message is a candidate for presentation to a requesting user who requests a plurality of messages to be provided by the social messaging system;
obtaining a number of subscribers for the posting user, the number of subscribers for the posting user representing how many users in the social messaging system have subscribed to receive messages from the posting user;
computing an expected interaction level of subscribers of the posting user for the message according to the number of subscribers for the posting user;
obtaining observed interaction data for the message, the observed interaction data representing interactions with the message by one or more subscribers of the posting user to whom the message has been presented;
computing an observed interaction level of subscribers of the posting user based on the observed interaction data for the message;
computing an interaction level deviation between the observed interaction level of subscribers of the posting user and the expected interaction level of subscribers of the posting user;
computing a score for the message representing how likely it is that subscribers of the posting user in the social messaging system will interact with the message, the score being based on the computed interaction level deviation between the observed interaction level of subscribers of the posting user and the expected interaction level of subscribers of the posting user;
ranking the message relative to a plurality of other messages based on the computed score for the message, the plurality of other messages being candidates for presentation to the requesting user; and
providing one or more highest-ranked messages from among the plurality of other messages and the candidate message, including the candidate message, to a user device for presentation to the requesting user.
10. The method of claim 9 , wherein the computing the expected interaction level of subscribers of the posting user for the message according to the number of subscribers for the posting user comprises providing the number of subscribers for the posting user as input to a function that assigns an expected interaction level based on previous observations of other users interacting with messages posted by a plurality of posting users.
11. The method of claim 10 , wherein the function defines a plurality of ranges and assigns a respective expected number of interactions to each range of the plurality of ranges.
12. The method of claim 9 , wherein the observed interaction data for the message comprises a number of subscribers of the posting user who selected the message after the message was presented.
13. The method of claim 9 , wherein the observed interaction data for the message comprises a number of subscribers of the posting user who shared the message after the message was presented.
14. The method of claim 9 , wherein the observed interaction data for the message comprises a number of subscribers of the posting user who tagged the message as a favorite after the message was presented.
15. The method of claim 9 , wherein computing a score for the message comprises:
applying a first weight to the computed interaction level deviation;
applying a second weight to a particular attribute value of the message; and
computing a weighted sum resulting from applying the first weight to the computed interaction level deviation and applying the second weight to the particular attribute value of the message.
16. The method of claim 15 , wherein the particular attribute value of the message is a number of subscribers to the posting user.
17. One or more non-transitory computer storage media encoded with computer program instructions that when executed by one or more computers cause the one or more computers to perform operations comprising:
receiving a message authored by a posting user in a social messaging system having a plurality of users, wherein the message is a candidate for presentation to a requesting user who requests a plurality of messages to be provided by the social messaging system;
obtaining a number of subscribers for the posting user, the number of subscribers for the posting user representing how many users in the social messaging system have subscribed to receive messages from the posting user;
computing an expected interaction level of subscribers of the posting user for the message according to the number of subscribers for the posting user;
obtaining observed interaction data for the message, the observed interaction data representing interactions with the message by one or more subscribers of the posting user to whom the message has been presented;
computing an observed interaction level of subscribers of the posting user based on the observed interaction data for the message;
computing an interaction level deviation between the observed interaction level of subscribers of the posting user and the expected interaction level of subscribers of the posting user;
computing a score for the message representing how likely it is that subscribers of the posting user in the social messaging system will interact with the message, the score being based on the computed interaction level deviation between the observed interaction level of subscribers of the posting user and the expected interaction level of subscribers of the posting user;
ranking the message relative to a plurality of other messages based on the computed score for the message, the plurality of other messages being candidates for presentation to the requesting user; and
providing one or more highest-ranked messages from among the plurality of other messages and the candidate message, including the candidate message, to a user device for presentation to the requesting user.
18. The one or more non-transitory computer storage media of claim 17 , wherein the computing the expected interaction level of subscribers of the posting user for the message according to the number of subscribers for the posting user comprises providing the number of subscribers for the posting user as input to a function that assigns an expected interaction level based on previous observations of other users interacting with messages posted by a plurality of posting users.
19. The one or more non-transitory computer storage media of claim 18 , wherein the function defines a plurality of ranges and assigns a respective expected number of interactions to each range of the plurality of ranges.
20. The one or more non-transitory computer storage media of claim 17 , wherein the observed interaction data for the message comprises a number of subscribers of the posting user who selected the message after the message was presented.