Systems and methods for generating representative models
Systems and methods are presented herein for generating representative models and assigning members of a task-facilitation service to representatives based on corresponding representative models. The task-facilitation service can transmit a set of queries that when received cause a computing device to generate a set of responses. The task-facilitation service may generate a feature vector that corresponds to the set of responses. The feature vector may be used to generate a representative model that corresponds to a user of the computing device. The representative model may be usable to establish communications one or more members of the task-facilitation service. The task-facilitation service may determine a correspondence between the representative model and one or more user models that correspond to the one or more members. The task-facilitation service may receive a selection of a particular user model and facilitate a communication to a client device associated with the particular user model.
1 . A method comprising:
transmitting a set of queries for a first computing device;
generating a training dataset using a set of responses to the set of queries and information associated with the first computing device;
augmenting the training dataset using a set of responses derived from a second computing device;
generating a quality metric by a machine-learning model using the augmented training dataset, wherein the quality metric is a prediction of a connection quality between the first computing device and a client device associated with a particular user model, and wherein the quality metric is based on a comparison of features associated with the first computing device and features of the particular user model;
facilitating a connection based on the quality metric, wherein the connection is between the client device associated with the particular user model and the first computing device;
generating a proposal for an executable task, wherein the proposal is generated based on a task request received over the connection between the client device and the first computing device;
generating an updated instance of the training dataset by appending the proposal to the augmented training dataset and by removing a portion of the training dataset associated with the second computing device, and wherein the updated instance of the training dataset includes only data derived from the first computing device;
determining that a training metric associated with the first computing device exceeds a threshold, wherein the training metric corresponds to a percentage of the updated instance of the training dataset that is associated with the first computing device;
retraining the machine-learning model based on the connection and the updated instance of the training dataset, wherein retraining the machine-learning model is in response to the training metric exceeding the threshold;
generating an updated quality metric using the machine-learning model and the updated instance of the training dataset, wherein the updated quality metric is a prediction of a connection quality between the client device and a third computing device; and
facilitating a connection between the client device and the third computing device based on updated quality metric.
2 . The method of claim 1 , wherein establishing the connection between the client device and the first computing device comprises:
determining, based on the particular user model, content of a first communication when establishing the connection;
selecting, based on the particular user model a communication protocol to establish the connection; and
establishing the connection between the client device and the first computing device by transmitting the first communication using the communication protocol.
3 . The method of claim 1 , further comprising:
updating a representative model in response to establishing a connection between the client device and the first computing device.
4 . The method of claim 3 , wherein the representative model includes an identification of each user model connected to the representative model.
5 . The method of claim 3 , wherein the representative model includes an identification of tasks executed on behalf of a user associated with the particular user model.
6 . The method of claim 3 , wherein the representative model is a machine-learning model and wherein the representative model is trained using a feature vector derived from historical interactions with the first computing device.
7 . The method of claim 3 , further comprising:
receiving, from the first computing device, an identification of a new task;
identifying a previously executed task in the representative model;
generating, in response to receiving the new task, a proposal based on the previously executed task, wherein the proposal is an implementation of the new task; and
facilitating a transmission to the first computing device that includes an identification of the proposal, wherein the proposal, when authorized by the first computing device, causes an execution of the new task.
8 . A system comprising:
one or more processors; and
a non-transitory computer-readable storage medium storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations including:
transmitting a set of queries for a first computing device;
generating a training dataset using a set of responses to the set of queries and information associated with the first computing device;
augmenting the training dataset using a set of responses derived from a second computing device;
generating a quality metric by a machine-learning model using the augmented training dataset, wherein the quality metric is a prediction of a connection quality between the first computing device and a client device associated with a particular user model, and wherein the quality metric is based on a comparison of features associated with the first computing device and features of the particular user model;
facilitating a connection based on the quality metric, wherein the connection is between the client device and the first computing device;
generating a proposal for an executable task, wherein the proposal is generated based on a task request received over the connection between the client device and the first computing device;
generating an updated instance of the training dataset by appending the proposal to the augmented training dataset and by removing a portion of the training dataset associated with the second computing device, and wherein the updated instance of the training dataset includes only data derived from the first computing device;
determining that a training metric associated with the first computing device exceeds a threshold, wherein the training metric corresponds to a percentage of the updated instance of the training dataset that is associated with the first computing device;
retraining the machine-learning model based on the connection and the updated instance of the training dataset, wherein retraining the machine-learning model is in response to the training metric exceeding the threshold;
generating an updated quality metric using the machine-learning model and the updated instance of the training dataset, wherein the updated quality metric is a prediction of a connection quality between the client device and a third computing device; and
facilitating a connection between the client device and the third computing device based on updated quality metric.
9 . The system of claim 8 , wherein establishing the connection between the client device and the first computing device comprises:
determining, based on the particular user model, content of a first communication when establishing the connection;
selecting, based on the particular user model a communication protocol to establish the connection; and
establishing the connection between the client device and the first computing device by transmitting the first communication using the communication protocol.
10 . The system of claim 8 , wherein the operations further include:
updating a representative model in response to establishing a connection between the client device and the first computing device.
11 . The system of claim 10 , wherein the representative model includes an identification of each user model connected to the representative model.
12 . The system of claim 10 , wherein the representative model includes an identification of tasks executed on behalf of a user associated with the particular user model.
13 . The system of claim 10 , wherein the representative model is a machine-learning model and wherein the representative model is trained using a feature vector derived from historical interactions with the first computing device.
14 . The system of claim 10 , wherein the operations further include:
receiving, from the first computing device, an identification of a new task;
identifying a previously executed task in the representative model;
generating, in response to receiving the new task, a proposal based on the previously executed task, wherein the proposal is an implementation of the new task; and
facilitating a transmission to the first computing device that includes an identification of the proposal, wherein the proposal, when authorized by the first computing device, causes an execution of the new task.
15 . A non-transitory computer-readable storage medium that stores instructions that, when executed by one or more processors, cause the one or more processors to perform operations including:
transmitting a set of queries for a first computing device;
generating a training dataset using a set of responses to the set of queries and information associated with the first computing device;
augmenting the training dataset using a set of responses derived from a second computing device;
generating a quality metric by a machine-learning model using the augmented training dataset, wherein the quality metric is a prediction of a connection quality between the first computing device and a client device associated with a particular user model, and wherein the quality metric is based on a comparison of features associated with the first computing device and features of the particular user model;
facilitating a connection based on the quality metric, wherein the connection is between the client device and the first computing device;
generating a proposal for an executable task, wherein the proposal is generated based on a task request received over the connection between the client device and the first computing device;
generating an updated instance of the training dataset by appending the proposal to the augmented training dataset and by removing a portion of the training dataset associated with the second computing device, and wherein the updated instance of the training dataset includes only data derived from the first computing device;
determining that a training metric associated with the first computing device exceeds a threshold, wherein the training metric corresponds to a percentage of the updated instance of the training dataset that is associated with the first computing device;
retraining the machine-learning model based on the connection and the updated instance of the training dataset, wherein retraining the machine-learning model is in response to the training metric exceeding the threshold;
generating an updated quality metric using the machine-learning model and the updated instance of the training dataset, wherein the updated quality metric is a prediction of a connection quality between the client device and a third computing device; and
facilitating a connection between the client device and the third computing device based on updated quality metric.
16 . The non-transitory computer-readable storage medium of claim 15 , wherein establishing the connection between the client device and the first computing device comprises:
determining, based on the particular user model, content of a first communication when establishing the connection;
selecting, based on the particular user model a communication protocol to establish the connection; and
establishing the connection between the client device and the first computing device by transmitting the first communication using the communication protocol.
17 . The non-transitory computer-readable storage medium of claim 15 , wherein the operations further include:
updating a representative model in response to establishing a connection between the client device and the first computing device.
18 . The non-transitory computer-readable storage medium of claim 17 , wherein the representative model includes an identification of each user model connected to the representative model.
19 . The non-transitory computer-readable storage medium of claim 17 , wherein the representative model includes an identification of tasks executed on behalf of a user associated with the particular user model.
20 . The non-transitory computer-readable storage medium of claim 17 , wherein the representative model is a machine-learning model and wherein the representative model is trained using a feature vector derived from historical interactions with the first computing device.