Automatic poll generation based on user communications
A method, computer system, and computer program product are provided for automatically generating polls. A message of a plurality of messages is received corresponding to a conversation between a plurality of users. One or more candidate polls are generated using a natural language processing model and determining a poll type for each of the one or more candidate polls based on the message and a context of the conversation. It is determined that at least one candidate poll of the one or more candidate polls is relevant according to the context of the conversation. In response to determining that the at least one candidate poll is relevant, the poll of the poll type is generated based on the message for presentation to the plurality of users.
1. A method comprising:
receiving a message of a plurality of messages corresponding to a conversation between a plurality of users;
generating a plurality of candidate polls using a natural language processing model and determining a poll type for each candidate poll of the plurality of candidate polls based on the message and a context of the conversation;
determining that at least one candidate poll of the plurality of candidate polls is relevant according to the context of the conversation and based on the poll type; and
in response to determining that the at least one candidate poll is relevant, generating a poll of the poll type based on the message for presentation to the plurality of users.
2. The method of claim 1 , wherein the plurality of messages are obtained from one or more sources selected from a group consisting of: an email message, a short messaging service message, a chat message, and text generated based on an audio conversation.
3. The method of claim 1 , wherein the context comprises a portion of the conversation that corresponds to a selected time window within a time of the message in the conversation.
4. The method of claim 1 , wherein the poll type of the poll is selected from a group consisting of: an open-text entry poll, a word cloud poll, a ranking poll, a rating poll and a multiple-choice poll.
5. The method of claim 1 , wherein the poll is further generated based on one or more selected from a group consisting of: a title or topic of the conversation, an identity of one or more of the plurality of users, and previous polls and corresponding options.
6. The method of claim 1 , wherein determining the poll type is based on a poll type classifier that is trained using a corpus that includes example polls and corresponding messages used to generate each example poll.
7. The method of claim 1 , wherein determining that the at least one candidate poll is relevant is based on a poll relevance classifier that is trained using examples of previous polls that are each labeled with a number of received votes.
8. The method of claim 1 , wherein one or more of the plurality of users manually provides one or more additional options to the poll, and wherein the one or more additional options are used to retrain the natural language processing model.
9. A system comprising:
one or more computer processors;
one or more computer readable storage media; and
program instructions stored on the one or more computer readable storage media for execution by at least one of the one or more computer processors, the program instructions comprising instructions to:
receive a message of a plurality of messages corresponding to a conversation between a plurality of users;
generate a plurality of candidate polls using a natural language processing model and determining a poll type for each candidate poll of the plurality of candidate polls based on the message and a context of the conversation;
determine that at least one candidate poll of the plurality of candidate polls is relevant according to the context of the conversation and based on the poll type; and
in response to determining that the at least one candidate poll is relevant, generate a poll of the poll type based on the message for presentation to the plurality of users.
10. The system of claim 9 , wherein the plurality of messages are obtained from one or more sources selected from a group consisting of: an email message, a short messaging service message, a chat message, and text generated based on an audio conversation.
11. The system of claim 9 , wherein the context comprises a portion of the conversation that corresponds to a selected time window within a time of the message in the conversation.
12. The system of claim 9 , wherein the poll type of the poll is selected from a group consisting of: an open-text entry poll, a word cloud poll, a ranking poll, a rating poll and a multiple-choice poll.
13. The system of claim 9 , wherein the poll is further generated based on one or more selected from a group consisting of: a title or topic of the conversation, an identity of one or more of the plurality of users, and previous polls and corresponding options.
14. The system of claim 9 , wherein determining the poll type is based on a poll type classifier that is trained using a corpus that includes example polls and corresponding messages used to generate each example poll.
15. The system of claim 9 , wherein determining that the at least one candidate poll is relevant is based on a poll relevance classifier that is trained using examples of previous polls that are each labeled with a number of received votes.
16. The system of claim 9 , wherein one or more of the plurality of users manually provides one or more additional options to the poll, and wherein the one or more additional options are used to retrain the natural language processing model.
17. One or more non-transitory computer readable storage media having program instructions embodied therewith, the program instructions executable by a computer to cause the computer to perform operations including:
receive a message of a plurality of messages corresponding to a conversation between a plurality of users;
generate a plurality of candidate polls using a natural language processing model and determining a poll type for each candidate poll of the plurality of candidate polls based on the message and a context of the conversation;
determine that at least one candidate poll of the plurality of candidate polls is relevant according to the context of the conversation and based on the poll type; and
in response to determining that the at least one candidate poll is relevant, generate a poll of the poll type based on the message for presentation to the plurality of users.
18. The one or more non-transitory computer readable storage media of claim 17 , wherein the plurality of messages are obtained from one or more sources selected from a group consisting of: an email message, a short messaging service message, a chat message, and text generated based on an audio conversation.
19. The one or more non-transitory computer readable storage media of claim 17 , wherein the context comprises a portion of the conversation that corresponds to a selected time window within a time of the message in the conversation.
20. The one or more non-transitory computer readable storage media of claim 17 , wherein the poll type of the poll is selected from a group consisting of: an open-text entry poll, a word cloud poll, a ranking poll, a rating poll and a multiple-choice poll.