Determining conversational structure from speech
Embodiments are directed to organizing conversations. Words may be provided from a conversation stream. Each word may be mapped to a graph model based on characteristics of each word. The graph model may be partitioned based on one or more attributes of a nodes and edges included in the graph model such that nodes associated with relationship strength that exceeds a threshold value may be grouped together. Sentence models may be generated based on sentences included in the conversation stream. Combined models may be generated based on the sentence models and the graph such that each sentence model may be associated with one or more partitions of the graph model. A conversation digest may be generated based on the combined model such that the conversation digest identifies one or more dominant portions of the conversation that include key subject matter.
1. A method for organizing conversations over a network using one or more network computers that include one or more processors that perform actions, comprising:
employing a conversation stream of words to generate a graph model for a conversation of one or more of words;
generating one or more sentence models based on the graph model, wherein the one or more sentence models provide one or more key phrases and one or more sentences based on the one or more words;
generating a combined model based on the one or more sentence models and the graph model to provide one or more associations between one or more portions of the conversation stream with the one or more key phrases and one or more partitions of the graph model; and
determining one or more of topics or sub-topics in the conversation based on the one or more associations.
2. The method of claim 1 , further comprising:
employing the combined model to determine an amount of dominance by each portion of the conversation stream based on a strength of the one or more associations between the one or more key phrases and the one or more partitions of the graph model.
3. The method of claim 1 , wherein determining the one or more of topics or sub-topics in the conversation, further comprises:
determining one or more context switches in the conversation based on an amount of dominance by each portion of the conversation stream, wherein the one or more context switches are employed to identify the one or more of topics or sub-topics.
4. The method of claim 1 , further comprising:
categorizing the one or more partitions by one or more types based on one or more of characteristics or relationships for each word that is included in a partition.
5. The method of claim 1 , further comprising:
determining one or more values for a phrase for one or more portions of each sentence;
modifying the one or more values for each sentence based on one or more of a repetition of each word, a degree of each word, or a length of each sentence; and
employing the one or more modified values to order one or more key words for the conversation.
6. The method of claim 1 , further comprising:
determining a transition from a first portion to a second portion of the conversation stream based on an amount of dominance associated with the first portion and another amount of dominance associated with the second portion.
7. The method of claim 1 , wherein the conversation stream of words, further comprises, providing the one or more words from one or more of a live conversation, a recorded conversation, a live video, a recorded video, or a transcription of the conversation; and
converting speech in an audio portion of the live conversation, the recorded conversation, the live video, or the recorded video into text based on natural language processing.
8. A system for organizing conversation information over a network, comprising:
a network computer, comprising:
a memory that stores at least instructions; and
one or more processors that execute instructions that enable performance of actions, including:
employing a conversation stream of words to generate a graph model for a conversation of one or more of words;
generating one or more sentence models based on the graph model, wherein the one or more sentence models provide one or more key phrases and one or more sentences based on the one or more words;
generating a combined model based on the one or more sentence models and the graph model to provide one or more associations between one or more portions of the conversation stream with the one or more key phrases and one or more partitions of the graph model; and
determining one or more of topics or sub-topics in the conversation based on the one or more associations.
9. The system of claim 8 , further comprising:
employing the combined model to determine an amount of dominance by each portion of the conversation stream based on a strength of the one or more associations between the one or more key phrases and the one or more partitions of the graph model.
10. The system of claim 8 , wherein determining the one or more of topics or sub-topics in the conversation, further comprises:
determining one or more context switches in the conversation based on an amount of dominance by each portion of the conversation stream, wherein the one or more context switches are employed to identify the one or more of topics or sub-topics.
11. The system of claim 8 , further comprising:
categorizing the one or more partitions by one or more types based on one or more of characteristics or relationships for each word that is included in a partition.
12. The system of claim 8 , further comprising:
determining one or more values for a phrase for one or more portions of each sentence;
modifying the one or more values for each sentence based on one or more of a repetition of each word, a degree of each word, or a length of each sentence; and
employing the one or more modified values to order one or more key words for the conversation.
13. The system of claim 8 , further comprising:
determining a transition from a first portion to a second portion of the conversation stream based on an amount of dominance associated with the first portion and another amount of dominance associated with the second portion.
14. The system of claim 8 , wherein the conversation stream of words, further comprises, providing the one or more words from one or more of a live conversation, a recorded conversation, a live video, a recorded video, or a transcription of the conversation; and
converting speech in an audio portion of the live conversation, the recorded conversation, the live video, or the recorded video into text based on natural language processing.
15. A processor readable non-transitory storage media that includes instructions for organizing conversation information over a network, wherein execution of the instructions by one or more hardware processors enable performance of actions, comprising:
employing a conversation stream of words to generate a graph model for a conversation of one or more of words;
generating one or more sentence models based on the graph model, wherein the one or more sentence models provide one or more key phrases and one or more sentences based on the one or more words;
generating a combined model based on the one or more sentence models and the graph model to provide one or more associations between one or more portions of the conversation stream with the one or more key phrases and one or more partitions of the graph model; and
determining one or more of topics or sub-topics in the conversation based on the one or more associations.
16. The processor readable non-transitory storage media of claim 15 , further comprising:
employing the combined model to determine an amount of dominance by each portion of the conversation stream based on a strength of the one or more associations between the one or more key phrases and the one or more partitions of the graph model.
17. The processor readable non-transitory storage media of claim 15 , wherein determining the one or more of topics or sub-topics in the conversation, further comprises:
determining one or more context switches in the conversation based on an amount of dominance by each portion of the conversation stream, wherein the one or more context switches are employed to identify the one or more of topics or sub-topics.
18. The processor readable non-transitory storage media of claim 15 , further comprising:
categorizing the one or more partitions by one or more types based on one or more of characteristics or relationships for each word that is included in a partition.
19. The processor readable non-transitory storage media of claim 15 , further comprising:
determining one or more values for a phrase for one or more portions of each sentence;
modifying the one or more values for each sentence based on one or more of a repetition of each word, a degree of each word, or a length of each sentence; and
employing the one or more modified values to order one or more key words for the conversation.
20. The processor readable non-transitory storage media of claim 15 , further comprising:
determining a transition from a first portion to a second portion of the conversation stream based on an amount of dominance associated with the first portion and another amount of dominance associated with the second portion.