Automated Learning For Speech-Based Applications

Systems and methods for modifying a computer-based speech recognition system. A speech utterance is processed with the computer-based speech recognition system using a set of internal representations, which may comprise parameters for recognizing speech in a speech utterance, such as parameters of an acoustic model and/or a language model. The computer-based speech recognition system may perform a first task in response to the processed speech utterance. The utterance may also be provided to a human who performs a second task based on the utterance. Data indicative of the first task, performed by the computer system, is compared to data indicative of a second task, performed by the human in response to the speech utterance. Based on the comparison, the set of internal representations may be updated or modified to improve the speech recognition performance and capabilities of the speech recognition system. 115-. (canceled)16. A method comprising:receiving, by a computer-based speech recognition system, input including one or more words or phrases, the computer-based speech recognition system comprising at least one processor and at least one memory device;determining, by the computer-based speech recognition system, to provide the input to a human;receiving a response from the human that identifies a first task for the input;performing the first task that is identified by the human for the input;processing the input to identify a second task for the input, the processing using a set of internal representations of the computer-based speech recognition system;comparing the first task identified by the human with the second task identified by the computer-based recognition system to determine one or more differences between the first task and the second task; andmodifying the set of internal representations of the computer-based speech recognition system based at least in part on the one or more differences between the first task and the second task to create a modified ...

Conversation User Interface

A conversation user interface enables users to better understand their interactions with computing devices, particularly when speech input is involved. The conversation user interface conveys a visual representation of a conversation between the computing device, or virtual assistant thereon, and a user. The conversation user interface presents a series of dialog representations that show input from a user (verbal or otherwise) and responses from the device or virtual assistant. Associated with one or more of the dialog representations are one or more graphical elements to convey assumptions made to interpret the user input and derive an associated response. The conversation user interface enables the user to see the assumptions upon which the response was based, and to optionally change the assumption(s). Upon change of an assumption, the conversation GUI is refreshed to present a modified dialog representation of a new response derived from the altered set of assumptions. 1. One or more computer-readable media storing computer-executable instructions that , when executed on one or more processors , cause the one or more processors to perform acts comprising:causing display of a conversation user interface in conjunction with a site of a service provider;receiving input from a user while the user engages in a session on the site of the service provider, the user input comprising one of audio input, keypad input, or touch input;representing the user input in the conversation user interface;determining a response to the user input;representing the response in the conversation user interface;enabling the user to interact with the conversation user interface to ascertain how the response was determined and to modify assumptions used to determine the response;determining a revised response based on the modified assumptions; andrepresenting the modified response in the conversation user interface.2. One or more computer-readable media as recited in claim 1 , wherein the ...

Human-to-human Conversation Analysis

Customer support, and other types of activities in which there is a dialogue between two humans can generate large volumes of conversation records. Automated analysis of these records can provide information about high-level features of, for example, the workings of a customer service department. Analysis of these conversations between a customer and a customer-support agent may also allow identification of customer support activities that can be provided by virtual agents instead of actual human agents. The analysis may evaluate conversations in terms of complexity, duration, and sentiment of the participants. Additionally, the conversations may also be analyzed to identify the existence of selected concepts or keywords. Workflow characteristics, the extent to which the conversation represents a multi-step process intended to accomplish a task, may also be determined for the conversations. Characteristics of individual conversations may be combined to obtain generalized or representative features for a set of a conversation records.

Data Clustering System, Methods, and Techniques

Data having some similarities and some dissimilarities may be clustered or grouped according to the similarities and dissimilarities. The data may be clustered using agglomerative clustering techniques. The clusters may be used as suggestions for generating groups where a user may demonstrate certain criteria for grouping. The system may learn from the criteria and extrapolate the groupings to readily sort data into appropriate groups. The system may be easily refined as the user gains an understanding of the data. 1. One or more computer-readable storage media storing computer-readable instructions that , when executed , instruct one or more processors to perform operations comprising:receiving a set of data provisionally grouped into clusters;displaying a representation of a portion the provisionally grouped data;receiving an indication from a user that the portion of the provisionally grouped data belongs in a first group or an indication from the user that the portion of the provisionally grouped data belongs in a second group, the indication based at least in part on a trigger of the user, an experience of the user, an understanding of the user, a recognition of the user, or a combination thereof;clustering the set of data based at least in part by emulating the user based at least in part on the indication from the user.2. The one or more computer-readable storage media of claim 1 , wherein the indication from a user comprises an indication that the portion of the provisionally grouped data belongs in a subgroup of the first group or second group claim 1 , the subgroup not the same as one of the clusters.3. One or more computer-readable storage media storing computer-readable instructions that claim 1 , when executed claim 1 , instruct one or more processors to perform operations comprising:comparing a first model to a second model with respect to a portion of a data set;when the first model agrees with the second model, validating the first model;when the ...

Conversation User Interface

A conversation user interface enables users to better understand their interactions with computing devices, particularly when speech input is involved. The conversation user interface conveys a visual representation of a conversation between the computing device, or virtual assistant thereon, and a user. The conversation user interface presents a series of dialog representations that show input from a user (verbal or otherwise) and responses from the device or virtual assistant. Associated with one or more of the dialog representations are one or more graphical elements to convey assumptions made to interpret the user input and derive an associated response. The conversation user interface enables the user to see the assumptions upon which the response was based, and to optionally change the assumption(s). Upon change of an assumption, the conversation GUI is refreshed to present a modified dialog representation of a new response derived from the altered set of assumptions. 1. One or more computer-readable media storing computer-executable instructions that , when executed on one or more processors , cause the one or more processors to perform acts comprising:causing display of a conversation user interface in conjunction with a site of a service provider;receiving input from a user while the user engages in a session on the site of the service provider, the user input comprising one of audio input, keypad input, or touch input;representing the user input in the conversation user interface;determining a response to the user input;representing the response in the conversation user interface;enabling the user to interact with the conversation user interface to ascertain how the response was determined and to modify assumptions used to determine the response;determining a revised response based on the modified assumptions; andrepresenting the modified response in the conversation user interface.2. One or more computer-readable media as recited in claim 1 , wherein the ...

HUMAN-TO-HUMAN CONVERSATION ANALYSIS

Customer support, and other types of activities in which there is a dialogue between two humans can generate large volumes of conversation records. Automated analysis of these records can provide information about high-level features of, for example, the workings of a customer service department. Analysis of these conversations between a customer and a customer-support agent may also allow identification of customer support activities that can be provided by virtual agents instead of actual human agents. The analysis may evaluate conversations in terms of complexity, duration, and sentiment of the participants. Additionally, the conversations may also be analyzed to identify the existence of selected concepts or keywords. Workflow characteristics, the extent to which the conversation represents a multi-step process intended to accomplish a task, may also be determined for the conversations. Characteristics of individual conversations may be combined to obtain generalized or representative features for a set of a conversation records. 1. A method of automatically analyzing dialog between humans , the method comprising:accessing a corpus of human-to-human conversations;scoring, using one or more processing units, individual conversations in the corpus according to one or more metrics, the one or more metrics being indicative of at least one of duration, sentiment, complexity and workflow of the individual conversation; andcombining the one or more metrics for each individual conversation of the corpus to generate a representative value of the one or more metrics for the corpus.2. The method of claim 1 , wherein claim 1 , the corpus comprises over 1 claim 1 ,000 individual conversations.3. The method of claim 1 , wherein claim 1 , for an individual conversation claim 1 , one of the humans is a customer and the other one of the humans is a customer service.4. The method of claim 3 , wherein representative vale is indicative of difficulty in replacing the customer ...

METHOD AND SYSTEM FOR GENERATING AND CORRECTING CLASSIFICATION MODELS

Data having some similarities and some dissimilarities may be clustered or grouped according to the similarities and dissimilarities. The data may be clustered using agglomerative clustering techniques. The clusters may be used as suggestions for generating groups where a user may demonstrate certain criteria for grouping. The system may learn from the criteria and extrapolate the groupings to readily sort data into appropriate groups. The system may be easily refined as the user gains an understanding of the data. 1. One or more computer-readable storage media storing computer-readable instructions that , when executed , instruct one or more processors to perform operations comprising:receiving a set of data provisionally grouped into clusters;displaying a representation of a portion the provisionally grouped data;receiving an indication from a user that the portion of the provisionally grouped data belongs in a first group or an indication from the user that the portion of the provisionally grouped data belongs in a second group, the indication based at least in part on a trigger of the user, an experience of the user, an understanding of the user, a recognition of the user, or a combination thereof; andclustering the set of data based at least in part by emulating the user based at least in part on the indication from the user.2. The one or more computer-readable storage media of claim 1 , wherein the indication from a user comprises an indication that the portion of the provisionally grouped data belongs in a subgroup of the first group or second group claim 1 , the subgroup not the same as one of the clusters.3. One or more computer-readable storage media storing computer-readable instructions that claim 1 , when executed claim 1 , instruct one or more processors to perform operations comprising:receiving data;clustering the data based at least in part on a clustering algorithm to create clustered data comprising data associated with a cluster and data not ...

METHOD AND SYSTEM FOR GENERATING AND CORRECTING CLASSIFICATION MODELS

Data having some similarities and some dissimilarities may be clustered or grouped according to the similarities and dissimilarities. The data may be clustered using agglomerative clustering techniques. The clusters may be used as suggestions for generating groups where a user may demonstrate certain criteria for grouping. The system may learn from the criteria and extrapolate the groupings to readily sort data into appropriate groups. The system may be easily refined as the user gains an understanding of the data. 1. One or more non-transitory computer-readable storage media storing computer-readable instructions that , when executed , instruct one or more processors to perform operations comprising:comparing, using a comparator, (1) a first score using test results from a first classification model based on a data set of natural language and (2) a second score using test results from a second classification model based on the data set of natural language;when the first score agrees with the second score within a threshold range, validating, by the one or more processors, the first classification model;when the first score is different from the second score with respect to the threshold range, providing, to a user device, a clarification question for a user;when an indication is received from the user, in response to the clarification question, that the first classification model is correct, updating, by the one or more processors, the second classification model based on the indication; andwhen an indication is received from the user, in response to the clarification question, that the second classification model is correct, updating, by the one or more processors, the first classification model based on the indication.2. The one or more non-transitory computer-readable storage media of claim 1 , wherein the first classification model comprises a symbolic language model; and the second classification model comprises a statistical language model.3. The one or more ...

Automated Learning For Speech-Based Applications

Systems and methods for modifying a computer-based speech recognition system. A speech utterance is processed with the computer-based speech recognition system using a set of internal representations, which may comprise parameters for recognizing speech in a speech utterance, such as parameters of an acoustic model and/or a language model. The computer-based speech recognition system may perform a first task in response to the processed speech utterance. The utterance may also be provided to a human who performs a second task based on the utterance. Data indicative of the first task, performed by the computer system, is compared to data indicative of a second task, performed by the human in response to the speech utterance. Based on the comparison, the set of internal representations may be updated or modified to improve the speech recognition performance and capabilities of the speech recognition system. 115-. (canceled)16. A method for checking the performance of an updated computer-based speech recognition system , comprising:receiving, by the computer-based speech recognition system, a speech input and a first task associated with the speech input, the first task being determined by processing the speech input using an original set of internal representations for the computer-based speech recognition system, the original set of internal representations comprising one or more parameters for recognizing speech in the speech input, the computer-based speech recognition system comprising at least one processor and at least one memory device;comparing the first task with a second task associated with the speech input, the second task being identified by a human in response to hearing the speech input;based at least in part on the comparison, modifying the original set of internal representations to create a modified set of internal representations;processing, using the modified set of internal representations, the speech input to identify a third task for the speech ...

Human-to-human conversation analysis

Customer support, and other types of activities in which there is a dialogue between two humans can generate large volumes of conversation records. Automated analysis of these records can provide information about high-level features of, for example, the workings of a customer service department. Analysis of these conversations between a customer and a customer-support agent may also allow identification of customer support activities that can be provided by virtual agents instead of actual human agents. The analysis may evaluate conversations in terms of complexity, duration, and sentiment of the participants. Additionally, the conversations may also be analyzed to identify the existence of selected concepts or keywords. Workflow characteristics, the extent to which the conversation represents a multi-step process intended to accomplish a task, may also be determined for the conversations. Characteristics of individual conversations may be combined to obtain generalized or representative features for a set of a conversation records.

Automated learning for speech-based applications

Systems and methods for modifying a computer-based speech recognition system. A speech utterance is processed with the computer-based speech recognition system using a set of internal representations, which may comprise parameters for recognizing speech in a speech utterance, such as parameters of an acoustic model and/or a language model. The computer-based speech recognition system may perform a first task in response to the processed speech utterance. The utterance may also be provided to a human who performs a second task based on the utterance. Data indicative of the first task, performed by the computer system, is compared to data indicative of a second task, performed by the human in response to the speech utterance. Based on the comparison, the set of internal representations may be updated or modified to improve the speech recognition performance and capabilities of the speech recognition system.

Enhanced multi-modal communication

Systems and methods for handling dual modality communication between at least one user device and at least one server. The modalities comprise audio modalities and mechanical motion modalities. The server may be simultaneously connected to the user device via a data network and a voice network and simultaneously receive audio-based input and mechanical motion-based input.

Multistream network feature processing for a distributed speech recognition system

A distributed voice recognition system and method for obtaining acoustic features and speech activity at multiple frequencies by extracting high frequency components thereof on a device, such as a subscriber station and transmitting them to a network server having multiple stream processing capability, including cepstral feature processing, MLP nonlinear transformation processing, and multiband temporal pattern architecture processing. The features received at the network server are processed using all three streams, wherein each of the three streams provide benefits not available in the other two, thereby enhancing feature interpretation. Feature extraction and feature interpretation may operate at multiple frequencies, including but not limited to 8 kHz, 11 kHz, and 16 kHz.

Distributed voice recognition system utilizing multistream feature processing

A distributed voice recognition system and method for obtaining acoustic features and speech activity at multiple frequencies by extracting high frequency components thereof on a device, such as a subscriber station, and transmitting them to a network server having multiple stream processing capability, including cepstral feature processing, MLP nonlinear transformation processing, and multiband temporal pattern architecture processing. The features received at the network server are processed using all three streams, wherein each of the three streams provide benefits not available in the other two, thereby enhancing feature interpretation. Feature extraction and feature interpretation may operate at multiple frequencies, including but not limited to 8 kHz, 11 kHz, and 16 kHz.

Human-to-human conversation analysis

Customer support, and other types of activities in which there is a dialogue between two humans can generate large volumes of conversation records. Automated analysis of these records can provide information about high-level features of, for example, the workings of a customer service department. Analysis of these conversations between a customer and a customer-support agent may also allow identification of customer support activities that can be provided by virtual agents instead of actual human agents. The analysis may evaluate conversations in terms of complexity, duration, and sentiment of the participants. Additionally, the conversations may also be analyzed to identify the existence of selected concepts or keywords. Workflow characteristics, the extent to which the conversation represents a multi-step process intended to accomplish a task, may also be determined for the conversations. Characteristics of individual conversations may be combined to obtain generalized or representative features for a set of a conversation records.

Automated assistant for customer service representatives

Systems and methods for handling information communicated by voice. The method may comprise: (i) receiving a call from a caller, the call comprising utterances from the caller; (ii) verbally communicating information to the caller through a customer service representative, the agent interacting with a display; (iii) processing the utterances with a computing device; (iv) determining content of the utterances; and (v) displaying information on the display based on the content.

Data clustering system and methods

Verteiltes allzweck-steuerungssystem für operationssäle

General purpose distributed operating room control system

The present invention pertains to control systems and provides a run time configurable control system for selecting and operating one of a plurality of operating room devices from a single input source, the system comprising a master controller (10) having a voice control interface (32) and a means for routing control signals (44). The system additionally may include a plurality of slave controllers (14) to provide expandability of the system. The system also includes an output means (86) for generating messages to the user relating to the status of the control system in general and the status of the devices connected to the control system.

General purpose distributed operating room control system

The present invention pertains to control systems and provides a run time configurable control system for selecting and operating one of a plurality of operating room devices from a single input source, the system comprising a master controller (10) having a voice control interface (32) and a means for routing control signals (44). The system additionally may include a plurality of slave controllers (14) to provide expandability of the system. The system also includes an output means (86) for generating messages to the user relating to the status of the control system in general and the status of the devices connected to the control system.

Data clustering system and methods

Data having some similarities and some dissimilarities may be clustered or grouped according to the similarities and dissimilarities. The data may be clustered using agglomerative clustering techniques. The clusters may be used as suggestions for generating groups where a user may demonstrate certain criteria for grouping. The system may learn from the criteria and extrapolate the groupings to readily sort data into appropriate groups. The system may be easily refined as the user gains an understanding of the data.