The present disclosure relates generally to network diagnostics, and more specifically to techniques for determining the health of customer premises equipment (CPE) devices. In certain examples, a scoring system can determine a health score for a CPE device, the health score indicating the health of the CPE device. The health score is determined by receiving, analyzing, and integrating information from different sources, including sources in addition to CPE diagnostic data for the CPE device at the current time. Such sources can include CPE diagnostic data for the CPE device at other times, CPE diagnostic data for other CPE devices related to the CPE device (e.g., in the same household or surrounding area as the CPE device), technical specifications of the CPE device (e.g., make, model and firmware of the CPE device), or per-equalization parameters obtained by the CPE device.

A server can receive a plurality of records at a databases such that each record is associated with a phone call and includes at least one request generated based on a transcript of the phone call. The server can generate a training dataset based on the plurality of records. The server can further train a binary classification model using the training dataset. Next, the server can receive a live transcript of a phone call in progress. The server can generate at least one live request based on the live transcript using a natural language processing module of the server. The server can provide the at least one live request to the binary classification model as input to generate a prediction. Lastly, the server can transmit the prediction to an entity receiving the phone call in progress. The prediction can cause a transfer of the call to a chatbot.

One disclosed system includes dialer device with a dialer identifier that initiates a call with a receiver device with a receiver identifier over a voice channel. A database stores: interaction data in association with both the receiver identifier and the dialer identifier; a content quality metric in association with the interaction data; and a network topology in which the receiver identifier and the dialer identifier are associated with nodes and are connected by an edge. A server is programmed to: update, based on the content quality metric and the network topology, a rankings associated with the receiver identifier and the dialer identifier; and one of: surface the ranking associated with the dialer identifier and the ranking associated with the receiver identifier to an external searching agent using an external data channel and without using any voice channel; or rank the dialer identifier for an internal searching agent using the ranking.

The present invention relates to a service system supporting a voice call using a digital assistant device, a method thereof, and a non-transitory computer readable medium having a computer program recorded thereon, and more particularly, to a service system which performs a call between a digital assistant device and a terminal corresponding to a call opponent through a voice command of a user and supports call conversion between the digital assistant device and a user terminal through the voice command of the user, a method thereof, and a non-transitory computer readable medium having a computer program recorded thereon.

The present disclosure relates generally to network diagnostics, and more specifically to techniques for determining the health of customer premises equipment (CPE) devices. In certain examples, a scoring system can determine a health score for a CPE device, the health score indicating the health of the CPE device. The health score is determined by receiving, analyzing, and integrating information from different sources, including sources in addition to CPE diagnostic data for the CPE device at the current time. Such sources can include CPE diagnostic data for the CPE device at other times, CPE diagnostic data for other CPE devices related to the CPE device (e.g., in the same household or surrounding area as the CPE device), technical specifications of the CPE device (e.g., make, model and firmware of the CPE device), or per-equalization parameters obtained by the CPE device.

Techniques are provided for evaluating multiple machine learning models to identify issues with a communication. One method comprises applying an audio signal associated with a communication to at least two of: (i) a trigger word analysis module that evaluates contextual information to determine if a trigger word is detected in the audio signal; (ii) an audio activity pattern analysis module that determines if a silence pattern anomaly is detected; and (iii) a communication application analysis module that evaluates features provided by a communication application relative to applicable thresholds; and combining results of the at least two of the trigger word analysis module, the audio activity pattern analysis module and the communication application analysis module to identify a communication issue. The combining may evaluate an accuracy of the trigger word analysis module, the audio activity pattern analysis module and/or the communication application analysis module to combine the results.

Techniques are provided for evaluating multiple machine learning models to identify issues with a communication. One method comprises applying an audio signal associated with a communication to at least two of: (i) a trigger word analysis module that evaluates contextual information to determine if a trigger word is detected in the audio signal; (ii) an audio activity pattern analysis module that determines if a silence pattern anomaly is detected; and (iii) a communication application analysis module that evaluates features provided by a communication application relative to applicable thresholds; and combining results of the at least two of the trigger word analysis module, the audio activity pattern analysis module and the communication application analysis module to identify a communication issue. The combining may evaluate an accuracy of the trigger word analysis module, the audio activity pattern analysis module and/or the communication application analysis module to combine the results.

According to one embodiment, a method, computer system, and computer program product for detecting and repairing gaps within a call is provided. The present invention may include requesting a personal discourse gap recovery model (PDGRM) for participants in a call, where the PDGRM may be a machine learning model that models a user's speech patterns based on word collocation, dictionary, and speaking rate. The present invention may further comprise detecting one or more gaps in the call based on network connectivity, inserting an audio delay into the call, constructing repaired dialogue to fill at least one of the one or more gaps in the call based on the PDGRM, and inserting the repaired dialogue into the call.

The present disclosure provides a method of measuring the effectiveness of an intervention in a hybrid fiber-metal access network. The effectiveness measure is determined in accordance with the improvement in the attenuation and the maximum achievable data rate. The effectiveness measure is used to determine whether a further network intervention is required.

One disclosed system includes dialer device with a dialer identifier that initiates a call with a receiver device with a receiver identifier over a voice channel. A database stores: interaction data in association with both the receiver identifier and the dialer identifier; a content quality metric in association with the interaction data; and a network topology in which the receiver identifier and the dialer identifier are associated with nodes and are connected by an edge. A server is programmed to: update, based on the content quality metric and the network topology, a rankings associated with the receiver identifier and the dialer identifier; and one of: surface the ranking associated with the dialer identifier and the ranking associated with the receiver identifier to an external searching agent using an external data channel and without using any voice channel; or rank the dialer identifier for an internal searching agent using the ranking.