Selecting user accounts in social network to answer question
Messages generated by user accounts in a social networking application over a period of time are processed to determine the subjects and topic associated with the messages, as well as the geographical locations of the users associated with the user accounts, and the times when the users associated with the accounts are most active. The determined, subject, time, and location information is used to create a model that may be used to predict whether a user in the social networking application is willing, available, and has the knowledge or topical affinity to answer a question proposed by another user in the social networking application based on a subject, time, and/or location associated with the question. When a user enters a question, the user may be presented with a list of their friends or contacts ranked according to the probabilities generated by the model.
1. A method comprising:
receiving a plurality of user account identifiers by a computing device;
receiving a plurality of messages, wherein each message is associated with a user account and a time by the computing device;
receiving a question by the computing device, wherein the question is associated with a time;
determining a subset of the plurality of user account identifiers based on the messages associated with the user accounts, the times associated with the messages, the question, and the time associated with the question by the computing device; and
providing the determined subset by the computing device.
2. The method of claim 1 , further comprising:
ranking the user account identifiers of the subset of the plurality of user account identifiers; and
providing the determined subset according to the ranking.
3. The method of claim 1 , wherein determining the subset of the plurality of user account identifiers based on the messages associated with the user accounts, the times associated with the messages, the question, and the time associated with the question comprises:
determining one or more subjects associated with the question;
determining one or more subjects associated with each of the user accounts based on the messages associated with each user account; and
determining the subset of the plurality of user account identifiers based on the times associated with the messages, the question, the time associated with the question, the one or more subjects associated with each of the user accounts, and the one or more subjects associated with the question.
4. The method of claim 1 , wherein determining the subset of the plurality of user account identifiers based on the messages associated with the user accounts, the times associated with the messages, the question, and the time associated with the question comprises:
determining the time associated with the question;
determining a probability that each of the identified user accounts is available at the determined time based on the times associated with the messages associated with each user account; and
determining the subset of the plurality of user account identifiers based on the messages associated with the user accounts, the probability determined for each identified user account, and the question.
5. The method of claim 1 , wherein each of the messages is associated with a location, and determining the subset of the plurality of user account identifiers based on the messages associated with the user accounts, the times associated with the messages, the question, and the time associated with the question comprises:
determining a location associated with the question;
determining a probable location for each identified user account at the time associated with the question based on the locations and times associated with the messages associated with each user account; and
determining the subset of the plurality of user account identifiers based on the messages associated with the user accounts, the question, the location associated with the question, and the determined probable location for each identified user account.
6. The method of claim 1 , wherein the question is associated with a user account identifier, and determining the subset of the plurality of user account identifiers based on the messages associated with the user accounts, the times associated with the messages, the question, and the time associated with the question comprises:
determining a social affinity between the user account associated with the question and each of the user accounts associated with the messages; and
determining the subset of the plurality of user account identifiers based on the messages associated with the user accounts, the times associated with the messages, the question, the time associated with the question, and the determined social affinity between the user account associated with the question and each of the user accounts associated with the messages.
7. The method of claim 1 , further comprising:
receiving an indication of selection of one or more of the user account identifiers in the determined subset; and
providing the question to the user accounts associated with the one or more selected user account identifiers.
8. The method of claim 1 , wherein the user account identifiers are identifiers of user accounts in one or more social networking applications.
9. A method comprising:
receiving a plurality of messages by a computing device, wherein each message is associated with a user account of a plurality of user accounts and a time;
generating a recommendation model based on the plurality of messages and the time associated with each message by the computing device;
receiving a question by the computing device, wherein the question is associated with a time; and
based on the received question, the time associated with the question, and the generated recommendation model, recommending one or more user accounts of the plurality of user accounts to answer the received question by the computing device.
10. The method of claim 9 , wherein generating the recommendation model comprises:
determining one or more subjects of each message of the received plurality of messages; and
generating the recommendation model based on the determined one or more subjects, the plurality of messages, and the time associated with each message.
11. The method of claim 9 , wherein generating the recommendation model comprises:
determining a location associated with each message of the plurality of received messages; and
generating the recommendation model based on the determined locations, the plurality of messages, and the time associated with each message.
12. The method of claim 9 , wherein generating the recommendation model comprises:
determining a social affinity between a user account associated with the question and the user accounts associated with each message of the plurality of received messages; and
generating the recommendation model based on the determined affinities, the plurality of messages, and the time associated with each message.
13. The method of claim 9 , wherein the user accounts are user account accounts in one or more social networking applications.
14. A system comprising:
a computing device;
a plurality of user account identifiers;
a plurality of messages, wherein each message is associated with a time and a user account identified by the plurality of user account identifiers; and
a recommendation engine adapted to:
generate one or more models based on the plurality of messages and the times associated with the messages of the plurality of messages;
receive a question, wherein the question is associated with a time;
determine a subset of a plurality of user account identifiers based on the one or more models, the received question, and the time associated with the received question; and
provide the determined subset.
15. The system of claim 14 , wherein each message is associated with a location, and the recommendation engine is further adapted to:
generate the one or more models further based on the location, and the user account associated with each message;
determine a location associated with the received question; and
determine the subset of the plurality of user accounts based on the one or more models, the received question, and the time and location associated with the received question.
16. The system of claim 14 , wherein the recommendation engine is further adapted to:
determine a social affinity between pairs of user accounts of the plurality of user accounts;
generate the one or more models further based on the determined social affinity between the pairs of user accounts;
determine a user account associated with the received question; and
determine the subset of the plurality of user accounts based on the one or more models, the received question, the time associated with the received question, and the user account associated with the received question.
17. The system of claim 14 , wherein each message is associated with a subject, and the recommendation engine is further adapted to:
generate the one or more models further based on the subject associated with each message;
determine a subject associated with the received question; and
determine the subset of the plurality of user accounts based on the one or more models, the received question, the time associated with the received question, and the subject associated with the received question.