Systems and methods for sentiment analysis of message posting in group conversations
The present disclosure provides, among other things, methods and systems of managing communications, the methods and systems including: receiving a first group communication from a first group; determining, based on the first group communication, a first group sentiment; receiving a first communication with a request to send the first communication to the first group; determining, based on the first communication, a first communication sentiment; comparing the first group sentiment with the first communication sentiment; and based on the comparing, performing an action on the request to send the first communication.
1. A method of managing communications, the method comprising:
receiving a group communication from a group;
determining, based on the group communication, a group sentiment;
receiving a communication with a request to send the communication to the group;
determining, based on the communication, a communication sentiment;
comparing the group sentiment with the communication sentiment; and
based on the comparison of the group sentiment with the communication sentiment, performing an action on the request to send the communication,
wherein the action comprises sending a notification to a sender of the communication and
wherein the notification comprises a result of the comparison of the group sentiment with the communication sentiment, a recommendation to not send the communication to the group and an option for the sender to override the recommendation and send the communication to the group before a timer that indicates how long the group sentiment will prevail in the group expires.
2. The method of claim 1 , wherein if the comparison of the group sentiment with the communication sentiment finds an agreement between the group sentiment and the communication sentiment, then the action comprises forwarding the communication to the group, and wherein if the comparison of the group sentiment with the communication sentiment finds a disagreement between the group sentiment and the communication sentiment, then the action comprises not forwarding the communication to the group and notifying a sender of the communication.
3. The method of claim 1 , wherein the group sentiment comprises a type of sentiment and a weight score.
4. The method of claim 3 , further comprising:
setting a current sentiment of the group based on the determining the group sentiment; and
setting the timer for the current sentiment in the group, wherein an amount of the timer is based on the weight score.
5. The method of claim 1 , wherein, before the determining the group sentiment, the group has a current sentiment, and wherein the determining the group sentiment further comprises:
receiving a new group communication in the group;
determining a sentiment of the new group communication; and
comparing the current sentiment to the sentiment of the new group communication to determine the group sentiment.
6. The method of claim 1 , wherein the group sentiment comprises a weight score, wherein the comparison of the group sentiment with the communication sentiment finds a disagreement between the communication sentiment and the group sentiment, and wherein the action is to forward the communication to the group based on the weight score.
7. The method of claim 1 , further comprising:
determining a confidence score for the communication sentiment;
determining a weight score for the communication sentiment; and
determining a time score for the communication sentiment, wherein the time score is based on the weight score.
8. The method of claim 7 , wherein the comparison of the group sentiment with the communication sentiment determines that the communication sentiment disagrees with the group sentiment, and if the weight score is above a threshold, then the action comprises posting the communication to the group.
9. The method of claim 1 , further comprising:
determining a weight score for the communication sentiment; and
comparing the weight score to a threshold before the performing the action.
10. The method of claim 1 , wherein the group is a social media group, wherein the social media group has members of the social media group, and wherein a sender of the request to send the communication is a member of the members.
11. The method of claim 1 , wherein a text string is associated with the communication and an image is associated with the communication, and wherein the determining the communication sentiment comprises a first sentiment analysis of the text string and a second sentiment analysis of the image.
12. The method of claim 9 , wherein at least one of the determining the group sentiment and the determining the communication sentiment comprises an aspect based sentiment analysis.
13. The method of claim 1 , further comprising:
storing a property of the group communication in a database of sentiment decisions;
enabling a machine learning process to analyze the database of sentiment decisions; and
updating a data model to automatically perform, based on an analysis performed by the machine learning process, an analysis of sentiment properties comprising the property of the communication.
14. A communication system, comprising:
a processor; and
a computer memory storing data thereon that enables the processor to:
receive a group communication from a group;
determine, based on the group communication, a type of group for the group, wherein the type of group comprises a policy;
receive a communication with a request to post the communication to the group;
determine, based on the communication, a communication sentiment;
compare the policy with the communication sentiment; and
based on the comparison of the policy with the communication sentiment, perform an action on the request to post the communication,
wherein the action comprises sending a notification to a sender of the communication, and
wherein the notification comprises a result of the comparison of the policy with the communication sentiment, a recommendation to not send the communication to the group, and an option for the sender to override the recommendation and send the communication to the group before a timer that indicates how long the policy will prevail in the group expires.
15. The communication system of claim 14 , wherein the policy comprises a rule about a basic sentiment, wherein the communication sentiment has a relationship to the basic sentiment, and wherein the action is determined by the rule and the relationship.
16. The communication system of claim 14 , wherein the type of group of the group is determined by a machine learning process.
17. The communication system of claim 14 , wherein the determining the communication sentiment comprises an aspect based sentiment analysis using a machine learning process.
18. A system, comprising:
a microprocessor and a computer-readable medium coupled thereto, wherein the microprocessor receives and executes instructions from the computer-readable medium and wherein the instructions cause the microprocessor to:
receive a group communication from a group;
determine, based on the group communication, a group type;
receive a communication with a request to post the communication to the group;
determine, based on the communication, a communication type;
compare the group type with the communication type of the communication of the request; and
based on the comparison of the group type with the communication type, perform an action on the request to post the communication,
wherein the action comprises sending a notification to a sender of the communication and
wherein the notification comprises a result of the comparison of the group type with the communication type, a recommendation to not send the communication to the group and an option for the sender to override the recommendation and send the communication to the group before a timer that indicates how long the group type will prevail in the group expires.
19. The system of claim 18 , wherein the group type is determined by a machine learning process.
20. The system of claim 18 , wherein the determining the group type comprises an aspect based sentiment analysis using a machine learning process.