Summarizing conversational speech
Embodiments are directed to summarizing conversational speech. Conversation segments may be provided based on a conversation stream and segmentation models. Summarization models may be determined based on characteristics of the conversation segments. Summarization information may be generated for each of the conversation segments based on the summarization models such that the summarization information includes a text-based summarization of the conversation segment. Summarization profiles may be generated for the conversation segments based on the summarization information such that each summarization profile is associated with quality scores. Summarization models may be modified based on the summarization profiles and the associated quality scores such that the summarization profiles are updated based on the modified summarization models. Modified summarization models and the updated summarization profiles may be employed to provide reports to a user.
1. A method for managing conversation information over a network, wherein one or more processors are configured to execute instructions that are configured to cause performance of the method, comprising:
determining one or more summarization models based on one or more characteristics of one or more conversation segments of a conversation stream;
generating summarization information for the one or more the conversation segments based on the one or more summarization models;
generating one or more summarization profiles for the one or more conversation segments based on the summarization information;
modifying the one or more summarization models based on the one or more summarization profiles; and
updating the one or more summarization profiles based on the one or more modified summarization models, wherein the one or more updated summarization profiles and the one or more modified summarization models are employed to provide one or more reports to a user.
2. The method of claim 1 , further comprising:
generating one or more summarization results for the one or more summarization models based on one or more user preference indicators provided by one or more selected users; and
determining a rank for the one or more summarization models based on the one or more summarization results.
3. The method of claim 2 , further comprising:
generating one or more user interfaces for the one or more selected users to score the one or more summarization models based on the one or more summarization results.
4. The method of claim 1 , wherein generating the summarization information further comprises:
generating a text summarization of the conversation segment based on one or more of natural language processing models.
5. The method of claim 1 , further comprising
generating one or more conversation digests of conversation streams based on one or more characteristics of audio conversations, wherein the one or more conversation digests provide are arranged to represent the contextual structure of a conversation as a sequence of one or more of a topic, speaker, or a connection.
6. The method of claim 1 , wherein modifying the one or more summarization models, further comprises:
associating one or more quality scores with each summarization profile that is used to modify the one or more summarization models.
7. The method of claim 1 , wherein generating the summarization information for each of the conversation segments, further comprises:
determining one or more conversation types based on the one or more conversation segments, wherein the one or more conversation types include one or more of text from an email, text from a chat session, a two-person telephone call, a group meeting, or a presentation.
8. A network computer for managing conversation information over a network, comprising:
a memory that stores at least instructions; and
one or more processors that execute instructions that are configured to cause performance of actions, including:
determining one or more summarization models based on one or more characteristics of one or more conversation segments of a conversation stream;
generating summarization information for the one or more the conversation segments based on the one or more summarization models;
generating one or more summarization profiles for the one or more conversation segments based on the summarization information;
modifying the one or more summarization models based on the one or more summarization profiles; and
updating the one or more summarization profiles based on the one or more modified summarization models, wherein the one or more updated summarization profiles and the one or more modified summarization models are employed to provide one or more reports to a user.
9. The network computer of claim 8 , further comprising:
generating one or more summarization results for the one or more summarization models based on one or more user preference indicators provided by one or more selected users; and
determining a rank for the one or more summarization models based on the one or more summarization results.
10. The network computer of claim 8 , further comprising:
generating one or more user interfaces for the one or more selected users to score the one or more summarization models based on the one or more summarization results.
11. The network computer of claim 8 , wherein generating the summarization information further comprises:
generating a text summarization of the conversation segment based on one or more of natural language processing models.
12. The network computer of claim 8 , further comprising
generating one or more conversation digests of conversation streams based on one or more characteristics of audio conversations, wherein the one or more conversation digests provide are arranged to represent the contextual structure of a conversation as a sequence of one or more of a topic, speaker, or a connection.
13. The network computer of claim 8 , wherein modifying the one or more summarization models, further comprises:
associating one or more quality scores with each summarization profile that is used to modify the one or more summarization models.
14. The network computer of claim 8 , wherein generating the summarization information for each of the conversation segments, further comprises:
determining one or more conversation types based on the one or more conversation segments, wherein the one or more conversation types include one or more of text from an email, text from a chat session, a two-person telephone call, a group meeting, or a presentation.
15. A processor readable non-transitory storage media that includes instructions for managing conversation information over a network, wherein one or more hardware processors execute the instructions that are configured to cause performance of actions, comprising:
determining one or more summarization models based on one or more characteristics of one or more conversation segments of a conversation stream;
generating summarization information for the one or more the conversation segments based on the one or more summarization models;
generating one or more summarization profiles for the one or more conversation segments based on the summarization information;
modifying the one or more summarization models based on the one or more summarization profiles; and
updating the one or more summarization profiles based on the one or more modified summarization models, wherein the one or more updated summarization profiles and the one or more modified summarization models are employed to provide one or more reports to a user.
16. The media of claim 15 , further comprising:
generating one or more summarization results for the one or more summarization models based on one or more user preference indicators provided by one or more selected users; and
determining a rank for the one or more summarization models based on the one or more summarization results.
17. The media of claim 15 , further comprising:
generating one or more user interfaces for the one or more selected users to score the one or more summarization models based on the one or more summarization results.
18. The media of claim 15 , wherein generating the summarization information further comprises:
generating a text summarization of the conversation segment based on one or more of natural language processing models.
19. The media of claim 15 , further comprising
generating one or more conversation digests of conversation streams based on one or more characteristics of audio conversations, wherein the one or more conversation digests provide are arranged to represent the contextual structure of a conversation as a sequence of one or more of a topic, speaker, or a connection.
20. The media of claim 15 , wherein generating the summarization information for each of the conversation segments, further comprises:
determining one or more conversation types based on the one or more conversation segments, wherein the one or more conversation types include one or more of text from an email, text from a chat session, a two-person telephone call, a group meeting, or a presentation.