Conversation facilitation system for mitigating loneliness
The present disclosure describes a conversation facilitation system for facilitating conversation-based social interactions to improve senior health, one or more operations and functions being efficiently achieved via this system comprise: receiving a dialog act of a conversation, applying natural language understanding (NLU) processing on the dialog act, computing a conversation metric, and generating a result of the conversation to conclude the conversation based on the conversation metric.
1. A computer-implemented method comprising:
selecting, with one or more processors, a senior at risk of loneliness to be a participant in a conversation between a plurality of human participants that includes the senior at risk of loneliness and at least one other human participant;
setting up, with the one or more processors, the conversation of the plurality of human participants;
receiving, with the one or more processors, a dialog act of the conversation between the plurality of human participants that includes the senior at risk of loneliness and at least one other human participant;
applying, with the one or more processors, natural language understanding processing on the dialog act;
tracking a topic of the conversation;
tracking a mood of at least one participant of the conversation; and
generating a conversation prompt, in response to the topic and mood, to intervene in the conversation and facilitate mitigation of loneliness in the senior at risk of loneliness; and
generating, with the one or more processors, a result of the conversation to conclude the conversation based on the conversation metric.
2. The method of claim 1 , wherein applying the natural language understanding processing on the dialog act comprises:
performing named entity recognition training on the dialog act; and
performing topic classifier training on the dialog act.
3. The method of claim 2 , wherein performing the named entity recognition training on the dialog act comprises training a long short term memory model to assign an entity included in the dialog act with a tag.
4. The method of claim 2 , wherein performing the named entity recognition training on the dialog act comprises computing a category score based on biasing a personal relationship known to a participant of the conversation.
5. The method of claim 2 , wherein performing the topic classifier training on the dialog act is based on a web search for a set of documents that reflects a topic of the conversation.
6. The method of claim 1 , further comprising applying sentiment analysis on the dialog act and using the sentiment analysis to track the mood of the conversation.
7. The method of claim 1 , further comprising proactively initiating the conversation to mitigate loneliness in a senior at risk of loneliness based upon detecting a risk factor for loneliness in the senior.
8. The method of claim 7 , further comprising generating, prior to the conversation of the plurality of human participants, a set of questions to structure the conversation to mitigate loneliness.
9. The method of claim 1 , wherein the dialog act is a time-stamped data element of the conversation that contains at least one of audio data, video data, or text data that has been recognized.
10. A system comprising:
a processor; and
a memory storing one or more instructions that, when executed, cause the processor to implement operations including:
selecting a senior at risk of loneliness to participate in a conversation that includes a plurality of human participants that includes the senior at risk of loneliness and at least one other human participant;
setting up the conversation;
receiving a dialog act of a conversation between the plurality of human participants that includes the senior at risk of loneliness and at least one other human participant;
applying natural language understanding processing on the dialog act;
computing a conversation metric; tracking a topic of the conversation;
tracking a mood of at least one participant of the conversation; and
generating a conversation prompt, in response to the topic and mood, to intervene in the conversation and facilitate mitigation of loneliness in the senior at risk of loneliness; and
generating a result of the conversation to conclude the conversation based on the conversation metric.
11. The system of claim 10 , wherein applying the natural language understanding processing on the dialog act comprises:
performing named entity recognition training on the dialog act; and
performing topic classifier training on the dialog act.
12. The system of claim 11 , wherein performing the named entity recognition training on the dialog act comprises training a long short term memory model to assign an entity included in the dialog act with a tag.
13. The system of claim 11 , wherein performing the named entity recognition training on the dialog act comprises computing a category score based on biasing a personal relationship known to a participant of the conversation.
14. The system of claim 11 , wherein performing the topic classifier training on the dialog act is based on a web search for a set of documents that reflects a topic of the conversation.
15. The system of claim 10 , further comprising applying sentiment analysis on the dialog act and using the sentiment analysis to track the mood of the conversation.
16. The system of claim 10 , further comprising proactively initiating the conversation to mitigate loneliness in a senior at risk of loneliness based upon detecting a risk factor for loneliness in the senior.
17. The system of claim 16 , further comprising generating, prior to the conversation of the plurality of human participants, a set of questions to structure the conversation to mitigate loneliness.
18. The system of claim 10 , wherein the dialog act is a time-stamped data element of the conversation that contains at least one of audio data, video data, or text data that has been recognized.
19. The method of claim 1 , further comprising generating, prior to the conversation of the plurality of human participants, at least one question to structure the conversation.
20. The method of claim 1 , further comprising generating, prior to the conversation of the plurality of human participants, at least one question to structure the conversation.