Tracking specialized concepts, topics, and activities in conversations
Embodiments are directed to organizing conversation information. A tracker vocabulary may be provided to a universal model to predict a generalized vocabulary associated with the tracker vocabulary. A tracker model may be generated based on the portions of the universal model activated by the tracker vocabulary such that a remainder of the universal model may be excluded from the tracker model. Portions of a conversation stream may be provided to the tracker model. A match score may be generated based on the track model and the portions of the conversation stream such that the match score predicts if the portions of the conversation stream may be in the generalized vocabulary predicted for the tracker vocabulary. Tracker metrics may be collected based on the portions of the conversation and the match scores such that the tracker metrics may be included in reports or notifications.
1. A method for organizing conversation information over a network using one or more network computers that include one or more processors that are operative to execute instructions, wherein the execution of the instructions enable performance of actions, comprising:
providing a vocabulary to a first model to predict a generalized vocabulary associated with the vocabulary, wherein the first model is trained to predict the generalized vocabulary that corresponds to the provided vocabulary;
generating a second model based on one or more portions of the first model that are activated by the provided vocabulary, wherein a remainder portion of the first model is excluded from the second model;
generating one or more scores based on the second model and one or more portions of a conversation, wherein the one or more scores predict the one or more portions of the conversation that are in the generalized vocabulary predicted for the provided vocabulary; and
providing one or more reports or notifications for the one or more portions of the conversation stream and the one or more scores.
2. The method of claim 1 , wherein the second model is configured to generate a score when a topic, a semantic meaning, a usage a sentiment, or an action of the conversation matches one or more of the topic, the semantic meaning, the usage, the sentiment, or the action associated the provided vocabulary.
3. The method of claim 1 , wherein the second model is configured to generate a score when one or more of a sound, a gesture or a movement matches one or more of the sound, the gesture, or the movement of the provided vocabulary.
4. The method of claim 1 , further comprising:
collecting one or more metrics based on the one or more portions of the conversation and the one or more scores, wherein the one or more metrics are included in the one or more of reports or notifications.
5. The method of claim 1 , further comprising:
employing a sliding window to store one or more words for each portion of a real time stream of the conversation, wherein a size of the sliding window to store an amount of words is dynamically adjusted to improve predictions that the one or more portions of the conversation are in the generalized vocabulary.
6. The method of claim 1 , wherein the conversation further comprises:
employing a sliding window to store non-text communication for each portion of a real time stream of the conversation that includes one or more of audio, video, image, or spatial information, wherein a size of the sliding window to store one or more of an amount of audio, video, image or spatial information is dynamically adjusted based on one or more of a time duration, an amount of data, or a video protocol.
7. The method of claim 1 , further comprising:
providing a processor readable non-transitory storage media to store the instructions for organizing conversation information over the network that are executed by the one or more processors.
8. A network computer for organizing conversation information over a network, comprising:
a memory that stores at least instructions; and
one or more processors that execute instructions, wherein the executed instructions enable performance of actions, including:
providing a vocabulary to a first model to predict a generalized vocabulary associated with the vocabulary, wherein the first model is trained to predict the generalized vocabulary that corresponds to the provided vocabulary;
generating a second model based on one or more portions of the first model that are activated by the provided vocabulary, wherein a remainder portion of the first model is excluded from the second model;
generating one or more scores based on the second model and one or more portions of a conversation, wherein the one or more scores predict the one or more portions of the conversation that are in the generalized vocabulary predicted for the provided vocabulary; and
providing one or more reports or notifications for the one or more portions of the conversation stream and the one or more scores.
9. The network computer of claim 8 , wherein the second model is configured to generate a score when a topic, a semantic meaning, a usage a sentiment, or an action of the conversation matches one or more of the topic, the semantic meaning, the usage, the sentiment, or the action associated the provided vocabulary.
10. The network computer of claim 8 , wherein the second model is configured to generate a score when one or more of a sound, a gesture or a movement matches one or more of the sound, the gesture, or the movement of the provided vocabulary.
11. The network computer of claim 8 , further comprising:
collecting one or more metrics based on the one or more portions of the conversation and the one or more scores, wherein the one or more metrics are included in the one or more of reports or notifications.
12. The network computer of claim 8 , further comprising:
employing a sliding window to store one or more words for each portion of a real time stream of the conversation, wherein a size of the sliding window to store an amount of words is dynamically adjusted to improve predictions that the one or more portions of the conversation are in the generalized vocabulary.
13. The network computer of claim 8 , wherein the conversation further comprises:
employing a sliding window to store non-text communication for each portion of a real time stream of the conversation that includes one or more of audio, video, image, or spatial information, wherein a size of the sliding window to store one or more of an amount of audio, video, image or spatial information is dynamically adjusted based on one or more of a time duration, an amount of data, or a video protocol.
14. The network computer of claim 8 , further comprising:
determining one or more network nodes in the first model that were activated to predict the generalized vocabulary based on one or more activation functions associated with the one or more network nodes.
15. 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:
providing a vocabulary to a first model to predict a generalized vocabulary associated with the vocabulary, wherein the first model is trained to predict the generalized vocabulary that corresponds to the provided vocabulary;
generating a second model based on one or more portions of the first model that are activated by the provided vocabulary, wherein a remainder portion of the first model is excluded from the second model;
generating one or more scores based on the second model and one or more portions of a conversation, wherein the one or more scores predict the one or more portions of the conversation that are in the generalized vocabulary predicted for the provided vocabulary; and
providing one or more reports or notifications for the one or more portions of the conversation stream and the one or more scores.
16. The system of claim 15 , wherein the second model is configured to generate a score when a topic, a semantic meaning, a usage a sentiment, or an action of the conversation matches one or more of the topic, the semantic meaning, the usage, the sentiment, or the action associated the provided vocabulary.
17. The system of claim 15 , wherein the second model is configured to generate a score when one or more of a sound, a gesture or a movement matches one or more of the sound, the gesture, or the movement of the provided vocabulary.
18. The system of claim 15 , further comprising:
collecting one or more metrics based on the one or more portions of the conversation and the one or more scores, wherein the one or more metrics are included in the one or more of reports or notifications.
19. The system of claim 15 , further comprising:
employing a sliding window to store one or more words for each portion of a real time stream of the conversation, wherein a size of the sliding window to store an amount of words is dynamically adjusted to improve predictions that the one or more portions of the conversation are in the generalized vocabulary.
20. The system of claim 15 , wherein the conversation further comprises:
employing a sliding window to store non-text communication for each portion of a real time stream of the conversation that includes one or more of audio, video, image, or spatial information, wherein a size of the sliding window to store one or more of an amount of audio, video, image or spatial information is dynamically adjusted based on one or more of a time duration, an amount of data, or a video protocol.