CN111611351B - Control method and device for online customer service session and electronic equipment - Google Patents

Control method and device for online customer service session and electronic equipment Download PDF

Info

Publication number
CN111611351B
CN111611351B CN201910138876.5A CN201910138876A CN111611351B CN 111611351 B CN111611351 B CN 111611351B CN 201910138876 A CN201910138876 A CN 201910138876A CN 111611351 B CN111611351 B CN 111611351B
Authority
CN
China
Prior art keywords
session
customer service
probability
message
training sample
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201910138876.5A
Other languages
Chinese (zh)
Other versions
CN111611351A (en
Inventor
陈永强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Didi Infinity Technology and Development Co Ltd
Original Assignee
Beijing Didi Infinity Technology and Development Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Didi Infinity Technology and Development Co Ltd filed Critical Beijing Didi Infinity Technology and Development Co Ltd
Priority to CN201910138876.5A priority Critical patent/CN111611351B/en
Publication of CN111611351A publication Critical patent/CN111611351A/en
Application granted granted Critical
Publication of CN111611351B publication Critical patent/CN111611351B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Computer And Data Communications (AREA)

Abstract

The application provides a control method, a device and electronic equipment for an online customer service session, wherein the method comprises the following steps: monitoring a customer service session on line; if the service provider does not receive the message of the user within the preset time after sending the current message, acquiring session information corresponding to the customer service session; the session information comprises session content and/or session statistical parameters in the customer service session; inputting session information into a pre-trained probability prediction model to obtain the continuous probability of customer service session; the online state of the customer service session is controlled based on the continuation probability. According to the embodiment of the application, the probability of user response can be predicted according to the session information of the customer service session through the probability prediction model, and the online state of the session is controlled based on the probability; the method can avoid the problem of users from disconnecting the session without solving the problem of users, and can also avoid the time waste of customer service after the problem of users is solved and waiting for the response of users, thereby improving the working efficiency of the customer service while guaranteeing the experience of the users.

Description

Control method and device for online customer service session and electronic equipment
Technical Field
The application relates to the technical field of artificial intelligence, in particular to a control method and device for online customer service session and electronic equipment.
Background
The online customer service of the service providing system can provide various services for users, timely solve the questions of the users and meet the demands of the users. In order to avoid the waste of customer service resources caused by the fact that a certain user occupies online customer service for a long time, in the related technology, if the user does not respond and does not actively disconnect the session any more, a timeout disconnection mode is generally adopted to automatically disconnect the session, but the mode may not solve the requirements of the user, in addition, the timeout disconnection waiting time is long, the time of manual customer service is still wasted, so that the user quantity that the manual customer service can serve is reduced, and the working efficiency of the manual customer service is reduced.
Disclosure of Invention
Accordingly, an object of the embodiments of the present application is to provide a method, an apparatus, and an electronic device for controlling an online customer service session, so as to avoid disconnecting the session when a user problem is not solved, and also avoid wasting time for waiting for a user to respond after the user problem is solved, thereby improving the working efficiency of the customer service while ensuring the user experience.
According to one aspect of the present application, an electronic device is provided that may include one or more storage media and one or more processors in communication with the storage media. One or more storage media store machine-readable instructions executable by a processor. When the electronic device is in operation, the processor and the storage medium communicate over the bus, and the processor executes the machine-readable instructions to perform one or more of the following:
A method for controlling an online customer service session, the method comprising: monitoring a customer service session on line; the customer service session is a session between a user and a service provider; if the service provider does not receive the message of the user within the preset time after sending the current message, acquiring session information corresponding to the customer service session; the session information comprises session content and/or session statistical parameters in the customer service session; inputting session information into a pre-trained probability prediction model to obtain the continuous probability of customer service session; the online state of the customer service session is controlled based on the continuation probability.
In some embodiments, the step of obtaining session information corresponding to the customer service session if the service provider does not receive the message of the user within a preset time period after sending the current message includes: starting a timer when the service provider sends a current message; the timing duration of the timer is preset duration; monitoring whether a message of a user is received within the timing duration of the timer; if not, obtaining the session information corresponding to the customer service session.
In some embodiments, the session content includes: based on the current message of the service provider, the current message and at least one part of messages which are continuous before the current message; the session statistics parameters include: the interval duration between two adjacent messages in the customer service session, the duration of the current message from the current time, the number of messages of the user, and the number of messages of the service provider.
In some embodiments, the probability prediction model is trained by: extracting session data from a preset session database; determining a plurality of training samples based on the session data; each training sample comprises a session identifier, and session content, a session closing reason and session statistical parameters corresponding to the session identifier; training a preset initial model based on a training sample to obtain a probability prediction model; wherein the initial model comprises: convolutional networks, recursive networks, and loss functions.
In some embodiments, the step of determining a plurality of training samples based on session data includes: dividing the session data into a plurality of pieces according to the session identification in the extracted session data; each session data corresponds to a session identifier; and extracting a session identifier contained in each session data, wherein the session identifier corresponds to session content, a session closing reason and session statistical parameters, and obtaining a training sample corresponding to each session data.
In some embodiments, after obtaining the training samples corresponding to each session data, the method further includes: for each training sample, judging whether the training sample contains the information of the user and the information of the session service provider; if not, the training sample is rejected.
In some embodiments, the step of training the preset initial model based on the training sample includes: determining a current training sample from a plurality of training samples, inputting session content of the current training sample into a convolution network of a preset initial model, and outputting a feature matrix corresponding to the session content; inputting the feature matrix and the session statistical parameters into a recursive network of the initial model, and outputting feature values corresponding to the current training samples; inputting the characteristic value into a loss function of the initial model, and outputting a prediction result of a session closing reason of the current training sample; updating the prediction accuracy of the initial model according to the session closing reason of the current training sample and the prediction result of the session closing reason; judging whether to terminate training based on preset conditions, wherein the preset conditions comprise: the prediction accuracy is greater than or equal to a preset accuracy threshold, and/or the number of current training samples input to the convolutional network is equal to a preset iteration number threshold; if not, determining the next current training sample from the plurality of training samples, and continuing to train the convolutional network, the recursive network and the loss function; and if so, taking the initial model trained by the current training sample as a probability prediction model.
In some embodiments, the convolutional network comprises a 3D convolutional neural network; the step of inputting the session content of the current training sample into the convolution network of the preset initial model and outputting the feature matrix corresponding to the session content includes: and mapping the conversation content of the current training sample through the 3D convolutional neural network to obtain a feature matrix corresponding to the conversation content.
In some embodiments, the recursive network includes a gating loop unit network with a preset number of layers; the gating circulation unit network with the preset layer number is sequentially connected; the number of layers of the gating circulation unit network corresponds to the number of messages in the session content; the step of inputting the feature matrix and the session statistical parameter into the recursive network of the initial model and outputting the feature value corresponding to the current training sample comprises the following steps: inputting the feature matrix of the message corresponding to the first layer gating circulation unit network in the feature matrix and session statistics parameters of the session content into the first layer gating circulation unit network to obtain an output result of the first layer gating circulation unit network; for each layer of gating and circulating unit network except the first layer of gating and circulating unit network, inputting the output result of the previous layer of gating and circulating unit network of the layer of gating and circulating unit network, the feature matrix of the message corresponding to the layer of gating and circulating unit network and the session statistical parameter of the session content into the layer of gating and circulating unit network until the last layer of gating and circulating unit network; and taking the output result of the last layer of gating circulating unit network as a characteristic value corresponding to the current training sample.
In some embodiments, the loss function comprises a softmax function; the number of the characteristic values corresponding to the current training samples is a plurality of; the step of inputting the feature value into the loss function of the initial model and outputting the prediction result of the session closing reason of the current training sample includes: calculating an exponential function value of each characteristic value through a softmax function; determining the probability of each characteristic value according to the index function value of each characteristic value and the sum of the index function values of each characteristic value of the current training sample; and outputting the probability of the appointed characteristic value in each characteristic value, and determining the probability of the appointed characteristic value and the session closing reason corresponding to the appointed characteristic value as the prediction result of the session closing reason of the current training sample.
In some embodiments, the step of controlling the online status of the customer service session based on the continuation probability includes: if the continuation probability is smaller than a preset probability threshold, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
In some embodiments, the step of controlling the online status of the customer service session based on the continuation probability includes: if the continuation probability is equal to or higher than a preset probability threshold, monitoring whether a message of the user is received within a preset timeout time threshold; if not, the customer service session is disconnected or the service provider is instructed to disconnect the customer service session.
In some embodiments, the step of controlling the online status of the customer service session based on the continuation probability includes: if the continuation probability is equal to or higher than a preset probability threshold, monitoring whether a message of the user is received within a preset timeout time threshold; if not, predicting new continuing probability of the customer service session through the probability prediction model continuously; and when the new continuation probability is lower than a preset probability threshold value, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
According to another aspect of the present application, there is also provided a control device for an online customer service session, the device including: the session monitoring module is used for monitoring the currently online customer service session; the customer service session is a session between a user and a service provider; the information acquisition module is used for acquiring session information corresponding to the customer service session if the service provider does not receive the message of the user within a preset time period after sending the current message; the session information comprises session content and/or session statistical parameters in the customer service session; the probability prediction module is used for inputting session information into a pre-trained probability prediction model to obtain the continuous probability of the customer service session; and the session control module is used for controlling the online state of the customer service session based on the continuation probability.
In some embodiments, the information obtaining module is configured to: starting a timer when the service provider sends a current message; the timing duration of the timer is preset duration; monitoring whether a message of a user is received within the timing duration of the timer; if not, obtaining the session information corresponding to the customer service session.
In some embodiments, the session content includes: based on the current message of the service provider, the current message and at least one part of messages which are continuous before the current message; the session statistics parameters include: the interval duration between two adjacent messages in the customer service session, the duration of the current message from the current time, the number of messages of the user, and the number of messages of the service provider.
In some embodiments, the apparatus further comprises a model training module for: extracting session data from a preset session database; determining a plurality of training samples based on the session data; each training sample comprises a session identifier, and session content, a session closing reason and session statistical parameters corresponding to the session identifier; training a preset initial model based on a training sample to obtain a probability prediction model; wherein the initial model comprises: convolutional networks, recursive networks, and loss functions.
In some embodiments, the model training module is configured to: dividing the session data into a plurality of pieces according to the session identification in the extracted session data; each session data corresponds to a session identifier; and extracting a session identifier contained in each session data, wherein the session identifier corresponds to session content, a session closing reason and session statistical parameters, and obtaining a training sample corresponding to each session data.
In some embodiments, the apparatus further comprises: the message judging module is used for judging whether the training samples contain the message of the user and the message of the session service provider for each training sample; and the sample removing module is used for removing the training sample if the training sample does not contain the user message and the session service provider message.
In some embodiments, the model training module is configured to: determining a current training sample from a plurality of training samples, inputting session content of the current training sample into a convolution network of a preset initial model, and outputting a feature matrix corresponding to the session content; inputting the feature matrix and the session statistical parameters into a recursive network of the initial model, and outputting feature values corresponding to the current training samples; inputting the characteristic value into a loss function of the initial model, and outputting a prediction result of a session closing reason of the current training sample; updating the prediction accuracy of the initial model according to the session closing reason of the current training sample and the prediction result of the session closing reason; judging whether to terminate training based on preset conditions, wherein the preset conditions comprise: the prediction accuracy is greater than or equal to a preset accuracy threshold, and/or the number of current training samples input to the convolutional network is equal to a preset iteration number threshold; if not, determining the next current training sample from the plurality of training samples, and continuing to train the convolutional network, the recursive network and the loss function; and if so, taking the initial model trained by the current training sample as a probability prediction model.
In some embodiments, the convolutional network comprises a 3D convolutional neural network; the model training module is used for: and mapping the conversation content of the current training sample through the 3D convolutional neural network to obtain a feature matrix corresponding to the conversation content.
In some embodiments, the recursive network includes a gating loop unit network with a preset number of layers; the gating circulation unit network with the preset layer number is sequentially connected; the number of layers of the gating circulation unit network corresponds to the number of messages in the session content; the model training module is used for: inputting the feature matrix of the message corresponding to the first layer gating circulation unit network in the feature matrix and session statistics parameters of the session content into the first layer gating circulation unit network to obtain an output result of the first layer gating circulation unit network; for each layer of gating and circulating unit network except the first layer of gating and circulating unit network, inputting the output result of the previous layer of gating and circulating unit network of the layer of gating and circulating unit network, the feature matrix of the message corresponding to the layer of gating and circulating unit network and the session statistical parameter of the session content into the layer of gating and circulating unit network until the last layer of gating and circulating unit network; and taking the output result of the last layer of gating circulating unit network as a characteristic value corresponding to the current training sample.
In some embodiments, the loss function comprises a softmax function; the number of the characteristic values corresponding to the current training samples is a plurality of; the model training module is used for: calculating an exponential function value of each characteristic value through a softmax function; determining the probability of each characteristic value according to the index function value of each characteristic value and the sum of the index function values of each characteristic value of the current training sample; and outputting the probability of the appointed characteristic value in each characteristic value, and determining the probability of the appointed characteristic value and the session closing reason corresponding to the appointed characteristic value as the prediction result of the session closing reason of the current training sample.
In some embodiments, the session control module is configured to: if the continuation probability is smaller than a preset probability threshold, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
In some embodiments, the session control module is configured to: if the continuation probability is equal to or higher than a preset probability threshold, monitoring whether a message of the user is received within a preset timeout time threshold; if not, the customer service session is disconnected or the service provider is instructed to disconnect the customer service session.
In some embodiments, the session control module is configured to: if the continuation probability is equal to or higher than a preset probability threshold, monitoring whether a message of the user is received within a preset timeout time threshold; if not, predicting new continuing probability of the customer service session through the probability prediction model continuously; and when the new continuation probability is lower than a preset probability threshold value, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
According to another aspect of the present application, there is also provided an electronic apparatus including: the system comprises a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, when the electronic device is running, the processor and the storage medium are communicated through the bus, and the processor executes the machine-readable instructions to execute the steps of the control method of the online customer service session.
According to another aspect of the present application, there is also provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the control method of an online customer service session as described above.
Based on any aspect, if the service provider does not receive the message of the user within the preset time after sending the current message in the current online customer service session, acquiring session information corresponding to the customer service session; and then inputting the session information into a probability prediction model to obtain the continuous probability of the customer service session, and controlling the online state of the customer service session based on the continuous probability. In the mode, the probability prediction model can predict the probability of user response according to the session information of the customer service session, namely the continuing probability of the customer service session, and the online state of the session is controlled based on the continuing probability; compared with the overtime disconnection mode in the related art, the method can avoid the session from being disconnected without solving the user problem, can also avoid the time waste of customer service after the user problem is solved and wait for the user to respond, and improves the working efficiency of the customer service while guaranteeing the user experience.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered limiting the scope, and that other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 illustrates a block diagram of a service providing system provided by an embodiment of the present application;
FIG. 2 shows a schematic diagram of exemplary hardware and software components of an electronic device provided by embodiments of the present application;
FIG. 3 is a flowchart illustrating a method for controlling an online customer service session according to an embodiment of the present disclosure;
FIG. 4 is a flowchart of a method for training a probabilistic predictive model provided by an embodiment of the present application;
FIG. 5 illustrates a flowchart of another method for training a probabilistic predictive model provided by an embodiment of the present application;
FIG. 6 is a flowchart illustrating another method for controlling an online customer service session according to an embodiment of the present application;
fig. 7 is a schematic structural diagram of a control device for an online customer service session according to an embodiment of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it should be understood that the accompanying drawings in the present application are only for the purpose of illustration and description, and are not intended to limit the protection scope of the present application. In addition, it should be understood that the schematic drawings are not drawn to scale. A flowchart, as used in this application, illustrates operations implemented according to some embodiments of the present application. It should be understood that the operations of the flow diagrams may be implemented out of order and that steps without logical context may be performed in reverse order or concurrently. Moreover, one or more other operations may be added to the flow diagrams and one or more operations may be removed from the flow diagrams as directed by those skilled in the art.
In addition, the described embodiments are only some, but not all, of the embodiments of the present application. The components of the embodiments of the present application, which are generally described and illustrated in the figures herein, may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present application, as provided in the accompanying drawings, is not intended to limit the scope of the application, as claimed, but is merely representative of selected embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, are intended to be within the scope of the present application.
In order to enable one skilled in the art to use the present disclosure, the following embodiments are presented in connection with a specific application scenario "rental car service". It will be apparent to those having ordinary skill in the art that the general principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the present application. Although the present application is primarily described in terms of a rental car service, it should be understood that this is but one exemplary embodiment. The present application may also be applied to other service systems, such as a system for sending and/or receiving express, a service system for business of both parties, etc.
It should be noted that the term "comprising" will be used in the embodiments of the present application to indicate the presence of the features stated hereinafter, but not to exclude the addition of other features.
Fig. 1 is a block diagram of a service providing system 100 according to some embodiments of the present application. For example, the service providing system 100 may be an online transport service platform for a transport service such as a taxi, a ride service, a express, a carpool, a bus service, a driver rental, or a class service, or any combination thereof. The service providing system 100 may include one or more of a server 110, a network 120, a user terminal 130, a service provider terminal 140, and a database 150, and a processor executing instruction operations may be included in the server 110.
In some embodiments, the server 110 may be a single server or a group of servers. The server farm may be centralized or distributed (e.g., server 110 may be a distributed system). In some embodiments, the server 110 may be local or remote to the terminal. For example, the server 110 may access information and/or data stored in the user terminal 130, the service provider terminal 140, or the database 150, or any combination thereof, via the network 120. As another example, the server 110 may be directly connected to at least one of the user terminal 130, the service provider terminal 140, and the database 150 to access stored information and/or data. In some embodiments, server 110 may be implemented on a cloud platform. In some embodiments, server 110 may be implemented on an electronic device 200 having one or more of the components shown in fig. 2 herein. In some embodiments, server 110 may include a processor. The processor may process information and/or data related to the service request to perform one or more functions described herein.
Network 120 may be used for the exchange of information and/or data. In some embodiments, one or more components in the service providing system 100 (e.g., the server 110, the user terminal 130, the service provider terminal 140, and the database 150) may send information and/or data to other components. For example, the server 110 may obtain a service request from the user terminal 130 via the network 120. In some embodiments, network 120 may be any type of wired or wireless network, or a combination thereof. By way of example only, the network 120 may include a wired network, a wireless network, a fiber optic network, a telecommunications network, and the like. In some embodiments, network 120 may include one or more network access points. For example, network 120 may include wired or wireless network access points, such as base stations and/or network switching nodes, through which one or more components of service providing system 100 may connect to network 120 to exchange data and/or information.
In some embodiments, "user" and "user terminal" may be used interchangeably, and "service provider" and "service provider terminal" may be used interchangeably. In some embodiments, the user terminal 130 may include a mobile device, a tablet computer, a laptop computer, or a built-in device in a motor vehicle, or the like. In some embodiments, the user terminal 130 may be a device with positioning technology for locating the user and/or the position of the user terminal.
In some embodiments, the service provider terminal 140 may be a similar or identical device as the user terminal 130. In some embodiments, the service provider terminal 140 may be a device with positioning technology for locating the location of the service provider and/or service provider terminal. In some embodiments, the user terminal 130 and/or the service provider terminal 140 may communicate with other positioning devices to determine the location of the user, the user terminal 130, the service provider, or the service provider terminal 140, or any combination thereof. In some embodiments, the user terminal 130 and/or the service provider terminal 140 may send the positioning information to the server 110.
Database 150 may store data and/or instructions. In some embodiments, database 150 may store data obtained from user terminal 130 and/or service provider terminal 140. Database 150 may store data and/or instructions for the exemplary methods described herein. The database 150 may include mass Memory, removable Memory, volatile Read-write Memory, or Read-Only Memory (ROM), etc. Database 150 may be implemented on a cloud platform.
In some embodiments, database 150 may be connected to network 120 to communicate with one or more components in service providing system 100 (e.g., server 110, user terminal 130, service provider terminal 140, etc.). One or more components in service providing system 100 may access data or instructions stored in database 150 via network 120. In some embodiments, database 150 may be directly connected to one or more components in service providing system 100 (e.g., server 110, user terminal 130, service provider terminal 140, etc.); alternatively, database 150 may be part of server 110.
Fig. 2 shows a schematic diagram of exemplary hardware and software components of an electronic device 200 of a server 110, a user terminal 130, a service provider terminal 140, which may implement the concepts of the present application, according to some embodiments of the present application. For example, a processor may be used on the electronic device 200 and to perform the functions herein.
The electronic device 200 may be a general purpose computer or a special purpose computer, both of which may be used to implement the control method of the online customer service session of the present application. Although only one computer is shown, the functionality described herein may be implemented in a distributed fashion across multiple similar platforms for convenience to balance processing loads.
For example, the electronic device 200 may include a network port 210 connected to a network, one or more processors 220 for executing program instructions, a communication bus 230, and various forms of storage media 240, such as magnetic disk, ROM, or RAM, or any combination thereof. By way of example, the computer platform may also include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The methods of the present application may be implemented in accordance with these program instructions. The electronic device 200 also includes an Input/Output (I/O) interface 250 between the computer and other Input/Output devices (e.g., keyboard, display screen).
For ease of illustration, only one processor is depicted in the electronic device 200. It should be noted, however, that the electronic device 200 in the present application may also include multiple processors, and thus steps performed by one processor described in the present application may also be performed jointly by multiple processors or separately. For example, if the processor of the electronic device 200 performs steps a and B, it should be understood that steps a and B may also be performed by two different processors together or performed separately in one processor. For example, the first processor performs step a, the second processor performs step B, or the first processor and the second processor together perform steps a and B.
Based on the above description of the electronic device, the embodiment of the application first describes a control method of an online customer service session, as shown in fig. 3, where the method includes the following steps:
step S302, monitoring a customer service session on line; wherein the customer service session is a session between a user and a service provider;
the service provider can specifically provide online customer service, manual customer service and the like for the service providing system; the service providing system can be a car renting system, a shopping system, an education system, an administrative system and the like; a user logging into the system may enjoy a corresponding service and thus the user may also be referred to as a service user. In the process of using the service provided by the service providing system, users often generate some questions or disputes, and at the moment, the users need to connect with the service provider to solve the questions or disputes. The user can click on the on-line customer service or manual customer service related button of the webpage or the application program webpage of the service providing system, and the service provider can be connected.
When users connect with service providers, if the number of service providers is small, the number of users connecting with the service providers is large, and the users may need to be queued to connect with the service providers to enter customer service sessions with the service providers. When both the user and the service provider enter a session, the customer service session may begin. In actual implementation, a session window may be set on the web page or the application page for the user and the service provider to conduct a session. A customer service session typically contains all messages to the session disconnection after the user has been connected to the service provider, and a customer service session typically corresponds to a unique session identification. During the conversation process, the system can monitor the current online customer service conversation in real time in the whole course, namely monitor the message sent by the user and the service provider (of course, the message sent by the system can be contained), and wait for the time of the message replying by the other party after the message is sent by one party; the system may also monitor the message replied by the user only after the service provider sends the message, and wait for the time for the user to reply to the message, etc.
Step S304, if the service provider does not receive the user' S message within the preset time after sending the current message, obtaining the session information corresponding to the customer service session; the session information comprises session content and/or session statistical parameters in the customer service session;
when the service provider sends out the current message, the user sometimes cannot reply to the message in time, and at the moment, two possibilities exist, namely, the user's question is solved, the session is ended, but the user does not actively disconnect the session; another is that the user may not have viewed the message sent by the service provider or the message sent by the service provider does not solve the user's question, but the user is busy with something else and does not get to reply to the message.
In the related art, if the service provider does not receive the user's message for a long time after sending the current message, the system will disconnect the session by itself; however, the possible reason that the user does not reply to the message is not considered in the method, so that on one hand, the doubt of the user is not solved, the user experience is affected, and on the other hand, the service provider is left empty for a long time, so that the working efficiency of the service provider is reduced; based on this, in this embodiment, the probability that the user's query is solved and the session should be ended is inferred through the messages that both parties have sent in the current customer service session, and then, based on the probability, it is determined whether to disconnect the session.
Specifically, in the step S304, after the service provider sends the current message, the service provider may count time, and when the count time reaches the preset duration, the message of the user is still not received, that is, session information corresponding to the customer service session is obtained. The preset time length can be set according to actual requirements, such as 30 seconds, 60 seconds and the like; the preset duration is generally shorter than the timeout duration used in the timeout disconnection mode in the related art, so as to save the time of the service provider. The obtained session information corresponding to the customer service session may only include one of session content or session statistics parameters in the customer service session, or may also include both session content and session statistics parameters in the customer service session. The session content in the customer service session may be all messages sent by the user and the service provider in the customer service session, or may be part of messages, where the part of messages may be messages sent by the user in the customer service session, messages sent by the service provider, or messages in specified positions in the sequence after being ordered according to the time of sending the messages, and so on. The session statistics parameters may be statistics information related to a customer service session, such as waiting time for sending each message, time for sending a current message by the service provider, and time from the current time, and may also include statistics information related to a user in the customer service session, such as the number of times the user uses online customer service.
Step S306, inputting session information into a pre-trained probability prediction model to obtain the continuous probability of customer service session;
the probabilistic predictive model may be implemented by a variety of machine learning tools such as neural networks, decision tree models, and the like. In the training process, the probability prediction model can acquire training samples from a database of customer service sessions, wherein the training samples generally comprise positive samples and negative samples, and the positive samples can be extracted from sessions disconnected in a time-out manner; the negative sample can be extracted from the session actively disconnected by the user; through the learning training of the positive sample and the negative sample, the probability prediction model can predict the continuation probability of the current customer service session based on the session information, and the continuation probability can be understood as the probability of the user replying to the message after the service provider sends the message.
Step S308, controlling the online state of the customer service session based on the continuation probability.
The online state of the customer service session may be online or disconnected; if the continuation probability is higher, the question of the user is not solved based on the session message, and the possibility that the user sends the message is higher, and at the moment, the session can be kept in an online state; if the continuation probability is low, it is assumed that the user's question is resolved based on the session information, and the user is highly likely to no longer send a message, and at this time, the customer service session may be disconnected. In addition, the system can disconnect the customer service session directly, and also can send prompt information to the service provider, and the service provider can determine whether to disconnect the session according to the actual situation. In actual implementation, a probability threshold or probability interval can be set, and the predicted continuous probability is compared with the probability threshold or probability interval, so that the online state of the customer service session is controlled.
In the method for controlling the online customer service session, if the service provider monitors that the message of the user is not received within the preset time after the service provider sends the current message in the online customer service session, session information corresponding to the customer service session is obtained; and then inputting the session information into a probability prediction model to obtain the continuous probability of the customer service session, and controlling the online state of the customer service session based on the continuous probability. In the mode, the probability prediction model can predict the probability of user response according to the session information of the customer service session, namely the continuing probability of the customer service session, and the online state of the session is controlled based on the continuing probability; compared with the overtime disconnection mode in the related art, the method can avoid the session from being disconnected without solving the user problem, can also avoid the time waste of customer service after the user problem is solved and wait for the user to respond, and improves the working efficiency of the customer service while guaranteeing the user experience.
In the above embodiment, it is proposed that the probability prediction model predicts the continuous probability of the customer service session, and then controls the online state of the customer service session based on the continuous probability; in this embodiment, a training manner of the probability prediction model is described with emphasis.
FIG. 4 is a schematic diagram of a training method of a probability prediction model; the method comprises the following steps:
step S402, extracting session data from a preset session database;
the session database may be a hive form database, although the session database may be implemented by other Hadoop architectures or databases (or data repositories) of other architectures. The session database stores session data generated when a user in the service providing system performs a session with a service provider. In the process of extracting the session data, a screening field can be set, and the session data meeting the conditions can be screened from the session database according to the screening field. For example, the screening field may contain a user identification (also referred to as a user ID), a session identification (also referred to as a session ID), a cause of session closure (including normal closure and timeout disconnect), each message type in a session (including voice, text, pictures, etc.); the time of sending each message in the session, the source of the message (including service provider, user, system), the content of the message, etc.
Step S404, determining a plurality of training samples based on session data; each training sample comprises a session identifier, and session content, a session closing reason and session statistical parameters corresponding to the session identifier;
The extracted session data includes session data corresponding to a plurality of sessions, and each session corresponds to a unique session identifier. A training sample is usually determined by session data corresponding to a session, and specifically, the session identifier, the session content, the reason for closing the session, the session statistical parameter, and the like may be extracted from session data of a session. The session content is a message sent by a user and a service provider in the session; the session content may contain all of the messages in the session or only a portion of the messages. For example, after ordering the messages in the session according to time, only the last messages are used as session content, and the number of the messages contained in the session content can be preset. The session statistics parameters may specifically be a time difference between adjacent messages after each message in the session is ordered according to time, a time difference between a last message and session hang-up time, a number of messages sent by a user, a number of messages sent by a service provider, and the like.
In addition, the determined training samples can be divided into positive samples and negative samples according to the session closing reason; the session closing reason of the positive sample is overtime disconnection, and the session closing reason of the negative sample is normal closing. In actual implementation, the number of positive samples and negative samples can be predetermined, and the ratio of the positive samples to the negative samples is set within a reasonable range, for example, the ratio of the number of positive samples to the number of negative samples is 1: 1. 4:1, etc.
Step S406, training a preset initial model based on a training sample to obtain a probability prediction model; wherein the initial model comprises: convolutional networks, recursive networks, and loss functions.
In the training process, training samples can be input into an initial model one by one, so that the initial model outputs a prediction result, the prediction result can comprise the probability of normal closing and the probability of overtime disconnection corresponding to the training samples, the prediction result is compared with the session closing reason in the training samples, and then parameters in the initial model are adjusted to achieve the purpose of training the model.
Considering that most of training samples are text data, the initial model comprises a convolution network, a recursion network and a loss function; wherein the convolutional network can convert text data into vectors or matrices capable of participating in computation; the recursion network can consider the time sequence of each message in the sample data, so as to extract the related characteristic data for predicting whether the user replies again later; the loss function may calculate a probability that the training sample is normally off and a probability that the timeout is open based on the feature data.
The model training mode can obtain a probability prediction model, and the continuous probability of the customer service session can be predicted and obtained based on the probability prediction model. In order to further improve the accuracy and stability of the probability prediction model, another model training method is provided in this embodiment, and the method further describes a specific training mode of the model and a data cleaning process of training data; as shown in fig. 5, the method comprises the steps of:
Step S502, extracting session data from a preset session database;
step S504, dividing the session data into a plurality of pieces according to the session identification in the extracted session data; each session data corresponds to a session identifier;
because the extracted session data relates to a plurality of sessions and each session corresponds to a unique session identifier, the session data can be divided based on the unique session identifier, so that the session data corresponding to each session is obtained.
Step S506, extracting a session identifier contained in each session data, wherein the session identifier corresponds to session content, a session closing reason and session statistical parameters, and obtaining a training sample corresponding to each session data.
The session data corresponding to each divided session generally includes a large amount of information, especially information irrelevant to probability prediction, such as a user account; in order to make the training data more concise and efficient, session content, a session closing reason and session statistical parameters related to probability prediction need to be extracted from each session data, and the extracted information forms a training sample.
Step S508, judging whether the training samples contain the user information and the session service provider information for each training sample; if not, the training sample is rejected. Until all training samples are traversed, a plurality of final training samples are obtained.
After the training samples corresponding to each session data are obtained, a plurality of training samples can be finally obtained; among these training samples, there still exist training samples that have no significant effect on model training, which may also be referred to as ineffective samples; for example, there are only samples of messages sent by the user, only samples of messages sent by the service provider; because there is no user and service provider message interaction in these samples, model training is not favored, and thus these samples need to be culled.
In actual implementation, the message sending sources of all the messages in the session content can be identified one by one, and whether the training sample contains the messages of the user and the messages of the session service provider is determined; if the identified messaging sources of the respective messages include the user and the session service provider, the training samples need not be culled; if the identified messaging source of each message includes only the user or the session service provider, the training sample is culled.
Step S510, determining a first current training sample from the plurality of training samples;
step S512, inputting the session content of the current training sample into a convolution network of a preset initial model, and outputting a feature matrix corresponding to the session content;
When in actual implementation, a plurality of training samples can be ordered according to a preset rule, then the training samples are determined to be current training samples one by one according to the arrangement sequence, and the current training samples are input into a convolution network; of course, the current training sample may also be randomly determined from a plurality of training samples. The convolution network can comprise a plurality of convolution layers, and each layer of convolution layer carries out convolution calculation on the conversation content through the corresponding convolution check, so that the conversation content in a text form is converted into a feature matrix; compared with the session content in a text form, the feature matrix is more convenient to participate in calculation and is beneficial to the feature extraction processing of the subsequent network.
Specifically, the convolutional network may be implemented by a 3D convolutional neural network that facilitates extracting correlations between neighboring words, or between neighboring messages; when the convolutional network in the model is a 3D convolutional neural network, mapping processing can be carried out on the conversation content of the current training sample through the 3D convolutional neural network, and a feature matrix corresponding to the conversation content is obtained. In actual implementation, mapping each message in the conversation content through a 3D convolutional neural network to obtain a feature matrix corresponding to each message; the process of mapping a message may also be referred to as text casting a message.
Step S514, inputting the feature matrix and the session statistical parameters into a recursive network of the initial model, and outputting feature values corresponding to the current training samples;
the recurrent network may be an LSTM (Long Short-Term Memory) network, a GRU (Gated Recurrent Unit, gated loop unit) network, a structural recurrent neural network, etc.; the recursive network generally has a tree hierarchy and network nodes, and can recursively process input information; when processing a text language, the recursive network can parse sentences in the text language, thereby extracting language meanings of the text language. The feature value corresponding to the current training sample output by the recursive network generally represents the meaning of the session content in the current training sample.
For further understanding, the following description will take, as an example, a recursive network including a network of gated loop units with a preset number of layers; the gating circulation unit network with the preset layer number is sequentially connected; the number of layers of the gating circulation unit network corresponds to the number of messages in the session content; for example, if the session content of the training sample includes five messages, the recursive network includes five layers of gated loop unit networks, and the five layers of gated loop unit networks are sequentially connected; after the five messages are sequenced according to time sequence, the first layer of gating circulation unit network processes the first message, the second layer of gating circulation unit network processes the second message until the fifth layer of gating circulation unit network processes the fifth message, and a final characteristic value is output; in these gating cell networks, the input data of the current layer of gating cell network generally includes the output data of the previous layer of gating cell network, and also includes the message corresponding to the current layer of gating cell network and other related data.
Based on the gating cycle cell network, the step S514 may be further implemented by the following steps 02-06:
step 02, inputting a feature matrix of a message corresponding to the first layer gating circulation unit network in the feature matrix and session statistics parameters of session content into the first layer gating circulation unit network to obtain an output result of the first layer gating circulation unit network;
specifically, the feature matrix of each message in the feature matrix may be ordered according to the time sequence of each message in the session content; as the number of the messages corresponds to the number of layers of the gating cycle unit network; therefore, the feature matrix corresponding to one message can be obtained from the feature matrices according to the arrangement sequence; inputting the acquired characteristic rectangle and the session statistical parameter of the session content into a first layer of gating circulation unit network for training; typically, the first layer of gating cyclic unit network performs convolution calculation on the input data, and the output result of the gating cyclic unit network is typically a feature matrix or feature vector obtained after the convolution calculation on the input data. The session statistical parameters input to the first layer gating cyclic unit network may be all session statistical parameters corresponding to the session content, or may be part of session statistical parameters associated with the message corresponding to the input feature matrix.
Step 04, for each layer of gating and circulating unit network except the first layer of gating and circulating unit network, inputting the output result of the previous layer of gating and circulating unit network of the layer of gating and circulating unit network, the feature matrix of the message corresponding to the layer of gating and circulating unit network and the session statistical parameter of the session content into the layer of gating and circulating unit network until the last layer of gating and circulating unit network;
for each layer of gating and circulating unit network except the first layer of gating and circulating unit network, the gating and circulating unit network of the previous layer is connected before the gating and circulating unit network; the input data of the gating circulation unit network comprises an output result of a previous layer of gating circulation unit network, a characteristic matrix of a message corresponding to the layer of gating circulation unit network and session statistical parameters of session content; the gating loop unit network performs convolution calculation on the input data, and an output result of the gating loop unit network is usually a feature matrix or feature vector obtained after the convolution calculation is performed on the input data. Because the input data of each layer of gating cycle unit network comprises the output result of the previous layer of gating cycle unit network, the input data of the last layer of gating cycle unit network comprises the output results corresponding to all previous layers, and the output result of the last layer of gating cycle unit network can represent the corresponding characteristics of all messages in the conversation content.
And step 06, taking the output result of the last layer of gating circulating unit network as a characteristic value corresponding to the current training sample.
The output result of the last layer of gating cyclic unit network usually contains a multi-dimensional feature matrix, so that the feature value corresponding to the current training sample also contains the multi-dimensional feature matrix, such as 300-dimensional or even more dimensions.
Step S516, inputting the characteristic value into a loss function of the initial model, and outputting a prediction result of the session closing reason of the current training sample;
the loss function may be a cross entropy function, a softmax function, a sigmoid function, etc.; the loss function can calculate the probability of the characteristic value of each dimension and output the appointed characteristic value; in this embodiment, the feature value of each dimension generally represents different meanings, but the feature value is generally included to represent the reason for closing the session, for example, the feature value a represents normal closing, and the feature value B represents timeout hanging-up; the loss function can specify the probability of outputting the characteristic value A and the probability of outputting the characteristic value B, wherein the probability of outputting the characteristic value A and the probability of outputting the characteristic value B are prediction results of the session closing reasons.
For further explanation, the implementation of the above step S516 is described below by taking the example that the loss function includes a softmax function, and specifically includes the following steps 12-16:
Step 12, calculating an exponential function value of each characteristic value through a softmax function;
specifically, the softmax function is formulated as follows:
wherein x is i Representing an ith eigenvalue; x is x j Represents the j-th feature value; n represents the total number of the characteristic values.
The exponential function value of the eigenvalue may expand the difference between the individual eigenvalues with respect to the eigenvalue itself, e.g. the vector of eigenvalues is [3,1, -3], and after calculating the exponential function value of each eigenvalue, the corresponding vector of exponential function values of the eigenvalue is [20,2.7,0.05]. The probability of each characteristic value is calculated by adopting the exponential function value of the characteristic value, so that the probability difference between the characteristic values can be increased, the probability of a correct recognition result is higher, and the accuracy of the recognition result is facilitated.
Step 14, determining the probability of each characteristic value according to the index function value of each characteristic value and the sum of the index function values of each characteristic value of the current training sample;
specifically, the probability of the feature value is obtained by dividing the exponential function value of each feature value by the sum of the exponential function values of each feature value.
And step 16, outputting the probability of the appointed characteristic value in each characteristic value, and determining the probability of the appointed characteristic value and the session closing reason corresponding to the appointed characteristic value as the prediction result of the session closing reason of the current training sample.
Since the feature value output by the recursive network generally contains a plurality of dimensions, and only the probability of the feature value related to the reason for the closing of the session is concerned in the present embodiment, the feature value related to the reason for the closing of the session may be determined as the specified feature value in advance, and the loss function may output only the probability of the specified feature value.
For example, the specified characteristic value is a characteristic value A and a characteristic value B, wherein the characteristic value A represents normal closing, and the characteristic value B represents overtime hanging-up; the probability corresponding to the characteristic value A is 0.7, the probability corresponding to the characteristic value B is 0.2, and the probability of normal closing of the session content is 0.7, the probability of overtime hanging-up is 0.2, and the probability of normal closing is higher; if the reason for closing the session of the session content is normal closing, the predicted result of the session content is consistent with the actual reason for closing the session.
Step S518, updating the prediction accuracy of the initial model according to the session closing reason of the current training sample and the prediction result of the session closing reason;
in the model training process, whether the model predicts correctly or incorrectly for each training sample can be recorded, so that the prediction accuracy of the initial model is counted; the prediction accuracy may be updated based on the current training sample whenever a new current training sample is predicted.
Step S520, judging whether to terminate training based on preset conditions; if not, go to step S522; if yes, go to step S524; wherein, the preset condition includes: the prediction accuracy is greater than or equal to a preset accuracy threshold, and/or the number of current training samples input to the convolutional network is equal to a preset iteration number threshold;
judging whether the preset condition is met or not after completing the prediction of one current training sample; the preset condition may include that the prediction accuracy is greater than or equal to a preset accuracy threshold, or that the number of current training samples input to the convolutional network is equal to one of the preset iteration count thresholds, or may also include that the prediction accuracy is greater than or equal to the preset accuracy threshold, and that the number of current training samples input to the convolutional network is equal to the preset iteration count threshold, that is, when the prediction accuracy of the model is greater than or equal to the preset accuracy threshold, and that the number of current training samples input to the convolutional network is equal to the preset iteration count threshold, training may be terminated.
The accuracy rate threshold and the iteration number threshold can be set according to actual requirements, for example, the accuracy rate threshold can be set to 80%, 90% and the like; the iteration number threshold may be located 1000 times, 5000 times, etc.
Step S522, determining the next current training sample from the plurality of training samples, and continuing to perform step S512 described above, that is, continuing to train the convolutional network, the recursive network, and the loss function. And (5) ending.
Step S524, taking the initial model trained by the current training sample as a probability prediction model.
The probability prediction model obtained through training in the mode is reliable and stable in performance and high in accuracy, the continuous probability of the customer service session can be predicted according to the session information of the customer service session, and the online state of the session is controlled based on the continuous probability, so that the session is disconnected without solving the user problem, the time waste of the customer service after the user problem is solved and the response of the user can be avoided, and the working efficiency of the customer service is improved while the user experience is ensured.
Further, the embodiment of the invention also provides another control method of the online customer service session, which mainly describes a specific implementation mode for controlling the online state of the customer service session based on the continuous probability; as shown in fig. 6, the method includes the steps of:
step S602, monitoring a customer service session on line; wherein the customer service session is a session between a user and a service provider;
Step S604, when the service provider transmits the current message, starting a timer; the timing duration of the timer is preset duration;
step S606, monitoring whether a message of a user is received in the timing duration of the timer; if yes, ending; if not, go to step S608;
step S608, obtaining session information corresponding to the customer service session.
The session information comprises session content and session statistical parameters; wherein, the session content includes: based on the current message of the service provider, the current message and at least one part of messages which are continuous before the current message; for example, the session content may be preset to include a specified number of messages, and from the current message of the service provider, a continuous specified number of messages are acquired; the specified number may be five, or other numbers, for example. If there are more messages in the session, the specified number of messages is only a portion of the messages of the session; if the number of the messages in the session is a specified number in total, the specified number of messages is all the messages of the session; if the number of messages in the session is less than the specified number, the number of messages in the entire session is obtained.
The session statistics parameters include: the interval duration between two adjacent messages in the customer service session, the duration of the current message from the current time, the number of messages of the user, and the number of messages of the service provider. In these session statistics parameters, the interval duration between two adjacent messages in the customer service session and the duration of the current message from the current time will usually represent the intention of the user more obviously, and the probability prediction is greatly affected, so in actual implementation, the session statistics parameters may only include the interval duration between two adjacent messages in the customer service session and the duration of the current message from the current time. Because the system usually automatically records the sending time of the message when the user or the service provider sends the message, the messages in the session can be ordered according to the sending time, and then the sending time of the adjacent session is subjected to difference operation, so that the interval duration between the two adjacent messages in the customer service session can be obtained, and the duration of the current message from the current time can be obtained by difference operation between the current time and the sending time of the current message.
Step S610, inputting session information into a pre-trained probability prediction model to obtain the continuous probability of customer service session;
step S612, judging whether the continuation probability is smaller than a preset probability threshold value; if not, go to step S614; if yes, go to step S616;
the probability threshold may be set according to actual requirements, for example, the probability threshold may be 50%, 70%, etc. If the continuation probability is smaller than a preset probability threshold, the probability of replying to the message by the user is not high, and the system can disconnect the customer service session directly; in addition, in order to more reliably control the customer service session, the system may not disconnect the customer service session directly, and send a prompt message to the service provider of the customer service session, so that the service provider is informed that the possibility of replying to the message is very low, and after receiving the prompt message, the service provider may further determine whether the customer service session can be disconnected according to the session content.
Step S614, monitoring whether a message of the user is received within a preset timeout threshold; if yes, ending; if not, go to step S616;
if the continuation threshold is higher, the customer service session is not disconnected temporarily or the service provider is instructed to disconnect the customer service session, at this time, a new timer may be started, the time duration of the timer is the timeout threshold, and if the user message is still not received during the time duration of the timer, the customer service session is disconnected or the service provider is instructed to disconnect the customer service session.
Step S616, disconnect the customer service session or instruct the service provider to disconnect the customer service session.
In the above manner, if the continuation probability output by the probability prediction model is smaller than the preset probability threshold, the customer service session is disconnected or the service provider is instructed to disconnect the customer service session; if the continuing probability is greater than or equal to a preset probability threshold, disconnecting the customer service session in a timeout disconnection mode or indicating a service provider to disconnect the customer service session; the method can avoid the problem of users from disconnecting the session without solving the problem of users, and can also avoid the time waste of customer service after the problem of users is solved and waiting for the response of users, thereby improving the working efficiency of the customer service while guaranteeing the experience of the users.
In another way, if the continuation probability is equal to or higher than a preset probability threshold, it may be monitored whether the message of the user is received within a preset timeout time threshold; if not, predicting new continuing probability of the customer service session through the probability prediction model continuously; and when the new continuation probability is lower than a preset probability threshold value, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
Specifically, if the message of the user is still not received within the timeout time threshold, session information corresponding to the customer service session may be continuously acquired; the session content in the session information is generally the same as the session content input to the probability prediction model last time; the session statistical parameters in the session information are usually different from the session statistical parameters input to the probability prediction model last time; for example, the duration of the current message sent by the service provider from the current time may be increased due to the change of the current time, and thus the continuation probability of the probability prediction model output may be changed. It will be appreciated that the longer the current message is from the current time, the lower the probability of continuation. Therefore, even if the probability prediction model predicts that the output continuation probability is lower for the first time, if the user does not send a message all the time, in the subsequent prediction process, the new continuation probability becomes lower and lower until the new continuation probability is lower than the preset probability threshold, the customer service session is disconnected or the service provider is instructed to disconnect the customer service session.
It should be noted that, if the above-mentioned multiple prediction method is adopted, the training sample needs to include the duration of the last message sent by the service provider from the session disconnection in the training process of the probability prediction model, and in the first prediction process, the session statistics parameter input into the probability prediction model includes the duration of the current message sent by the service provider from the current time.
In the above manner, if the continuous probability output by the probability prediction model is greater than or equal to the preset probability threshold, session information corresponding to the customer service session is acquired again, and is input into the probability prediction model for prediction, and a new probability threshold is output; the method can also avoid the problem of users from disconnecting the session without solving the problem of users, and can also avoid the time waste of customer service after the problem of users is solved and waiting for the response of users, thereby improving the working efficiency of the customer service while ensuring the experience of the users.
An embodiment of a method for controlling an online customer service session according to the foregoing disclosure is shown in fig. 7, which is a schematic structural diagram of a device for controlling an online customer service session; the functions implemented by the device correspond to the steps executed by the method. The apparatus may be understood as the above server, or a processor of the server, or may be understood as a component, which is independent from the above server or processor and is controlled by the server, to implement the functions of the present application, as shown in fig. 7, and includes:
A session monitoring module 70, configured to monitor a currently online customer service session; the customer service session is a session between a user and a service provider;
an information obtaining module 72, configured to obtain session information corresponding to a customer service session if the service provider does not receive the message of the user within a preset duration after sending the current message; the session information comprises session content and/or session statistical parameters in the customer service session;
the probability prediction module 74 is configured to input session information into a pre-trained probability prediction model, so as to obtain a continuous probability of the customer service session;
a session control module 76 for controlling the online status of the customer service session based on the probability of continuation.
In the control device for the online customer service session provided by the embodiment of the invention, if the fact that the service provider does not receive the message of the user within the preset time after sending the current message in the current online customer service session is monitored, session information corresponding to the customer service session is obtained; and then inputting the session information into a probability prediction model to obtain the continuous probability of the customer service session, and controlling the online state of the customer service session based on the continuous probability. In the mode, the probability prediction model can predict the probability of user response according to the session information of the customer service session, namely the continuing probability of the customer service session, and the online state of the session is controlled based on the continuing probability; compared with the overtime disconnection mode in the related art, the method can avoid the session from being disconnected without solving the user problem, can also avoid the time waste of customer service after the user problem is solved and wait for the user to respond, and improves the working efficiency of the customer service while guaranteeing the user experience.
The modules in the control device of the online customer service session described above may be connected or communicate with each other via a wired or wireless connection. The wired connection may include a metal cable, optical cable, hybrid cable, or the like, or any combination thereof. The wireless connection may include a connection through a LAN, WAN, bluetooth, zigBee, or NFC, or any combination thereof. Two or more modules may be combined into a single module, and any one module may be divided into two or more units.
In some embodiments, the information obtaining module is configured to: starting a timer when the service provider sends a current message; the timing duration of the timer is preset duration; monitoring whether a message of a user is received within the timing duration of the timer; if not, obtaining the session information corresponding to the customer service session.
In some embodiments, the session content includes: based on the current message of the service provider, the current message and at least one part of messages which are continuous before the current message; the session statistics parameters include: the interval duration between two adjacent messages in the customer service session, the duration of the current message from the current time, the number of messages of the user, and the number of messages of the service provider.
In some embodiments, the apparatus further comprises a model training module for: extracting session data from a preset session database; determining a plurality of training samples based on the session data; each training sample comprises a session identifier, and session content, a session closing reason and session statistical parameters corresponding to the session identifier; training a preset initial model based on a training sample to obtain a probability prediction model; wherein the initial model comprises: convolutional networks, recursive networks, and loss functions.
In some embodiments, the model training module is configured to: dividing the session data into a plurality of pieces according to the session identification in the extracted session data; each session data corresponds to a session identifier; and extracting a session identifier contained in each session data, wherein the session identifier corresponds to session content, a session closing reason and session statistical parameters, and obtaining a training sample corresponding to each session data.
In some embodiments, the apparatus further comprises: the message judging module is used for judging whether the training samples contain the message of the user and the message of the session service provider for each training sample; and the sample removing module is used for removing the training sample if the training sample does not contain the user message and the session service provider message.
In some embodiments, the model training module is configured to: determining a current training sample from a plurality of training samples, inputting session content of the current training sample into a convolution network of a preset initial model, and outputting a feature matrix corresponding to the session content; inputting the feature matrix and the session statistical parameters into a recursive network of the initial model, and outputting feature values corresponding to the current training samples; inputting the characteristic value into a loss function of the initial model, and outputting a prediction result of a session closing reason of the current training sample; updating the prediction accuracy of the initial model according to the session closing reason of the current training sample and the prediction result of the session closing reason; judging whether to terminate training based on preset conditions, wherein the preset conditions comprise: the prediction accuracy is greater than or equal to a preset accuracy threshold, and/or the number of current training samples input to the convolutional network is equal to a preset iteration number threshold; if not, determining the next current training sample from the plurality of training samples, and continuing to train the convolutional network, the recursive network and the loss function; and if so, taking the initial model trained by the current training sample as a probability prediction model.
In some embodiments, the convolutional network comprises a 3D convolutional neural network; the model training module is used for: and mapping the conversation content of the current training sample through the 3D convolutional neural network to obtain a feature matrix corresponding to the conversation content.
In some embodiments, the recursive network includes a gating loop unit network with a preset number of layers; the gating circulation unit network with the preset layer number is sequentially connected; the number of layers of the gating circulation unit network corresponds to the number of messages in the session content; the model training module is used for: inputting the feature matrix of the message corresponding to the first layer gating circulation unit network in the feature matrix and session statistics parameters of the session content into the first layer gating circulation unit network to obtain an output result of the first layer gating circulation unit network; for each layer of gating and circulating unit network except the first layer of gating and circulating unit network, inputting the output result of the previous layer of gating and circulating unit network of the layer of gating and circulating unit network, the feature matrix of the message corresponding to the layer of gating and circulating unit network and the session statistical parameter of the session content into the layer of gating and circulating unit network until the last layer of gating and circulating unit network; and taking the output result of the last layer of gating circulating unit network as a characteristic value corresponding to the current training sample.
In some embodiments, the loss function comprises a softmax function; the number of the characteristic values corresponding to the current training samples is a plurality of; the model training module is used for: calculating an exponential function value of each characteristic value through a softmax function; determining the probability of each characteristic value according to the index function value of each characteristic value and the sum of the index function values of each characteristic value of the current training sample; and outputting the probability of the appointed characteristic value in each characteristic value, and determining the probability of the appointed characteristic value and the session closing reason corresponding to the appointed characteristic value as the prediction result of the session closing reason of the current training sample.
In some embodiments, the session control module is configured to: if the continuation probability is smaller than a preset probability threshold, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
In some embodiments, the session control module is configured to: if the continuation probability is equal to or higher than a preset probability threshold, monitoring whether a message of the user is received within a preset timeout time threshold; if not, the customer service session is disconnected or the service provider is instructed to disconnect the customer service session.
In some embodiments, the session control module is configured to: if the continuation probability is equal to or higher than a preset probability threshold, monitoring whether a message of the user is received within a preset timeout time threshold; if not, predicting new continuing probability of the customer service session through the probability prediction model continuously; and when the new continuation probability is lower than a preset probability threshold value, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
The embodiment also provides an electronic device, including: the system comprises a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, when the electronic device is running, the processor and the storage medium are communicated through the bus, and the processor executes the machine-readable instructions to execute the steps of the control method of the online customer service session.
The present embodiment also provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the control method of an online customer service session as described above.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described system and apparatus may refer to corresponding procedures in the method embodiments, which are not described in detail in this application. In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. The above-described apparatus embodiments are merely illustrative, and the division of the modules is merely a logical function division, and there may be additional divisions when actually implemented, and for example, multiple modules or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some communication interface, indirect coupling or communication connection of devices or modules, electrical, mechanical, or other form.
The modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical units, may be located in one place, or may be distributed over multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a non-volatile computer readable storage medium executable by a processor. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk, etc.
The foregoing is merely a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think about changes or substitutions within the technical scope of the present application, and the changes or substitutions are covered in the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims (28)

1. A method for controlling an online customer service session, the method comprising:
monitoring a customer service session on line; wherein, the customer service session is a session between a user and a service provider;
if the service provider does not receive the message of the user within the preset time after sending the current message, session information corresponding to the customer service session is obtained; the session information comprises session content and/or session statistical parameters in the customer service session; the session statistical parameters comprise at least two of interval duration between two adjacent messages in the customer service session, duration of the current message from the current time, message quantity of the user and message quantity of the service provider;
inputting the session information into a pre-trained probability prediction model to obtain the continuous probability of the customer service session; wherein the continuation probability is determined by the probability prediction model based on feature values in a feature matrix of the session information; the continuation probability is used for representing the probability of the user replying to the message after the service provider sends the current message;
Controlling the online state of the customer service session based on the continuation probability;
the probability prediction model is obtained through training in the following mode:
determining a current training sample from a plurality of training samples, inputting session content of the current training sample into a convolution network of a preset initial model, and outputting a feature matrix corresponding to the session content;
inputting the feature matrix and session statistical parameters of the current training sample into a recursive network of the initial model, and outputting feature values corresponding to the current training sample;
inputting the characteristic value into a loss function of the initial model, and outputting a prediction result of a session closing reason of the current training sample;
updating the prediction accuracy of the initial model according to the session closing reason of the current training sample and the prediction result of the session closing reason;
judging whether to terminate training or not based on preset conditions;
if not, determining a next current training sample from a plurality of training samples, and continuing to train the convolution network, the recursion network and the loss function;
and if so, taking the initial model trained by the current training sample as a probability prediction model.
2. The method of claim 1, wherein the step of obtaining session information corresponding to the customer service session if the message of the user is not received within a preset time period after the service provider sends the current message comprises:
starting a timer when the service provider sends a current message; the timing duration of the timer is preset duration;
monitoring whether the message of the user is received within the timing duration of the timer;
if not, obtaining the session information corresponding to the customer service session.
3. The method according to claim 1 or 2, wherein the session content comprises: the current message and at least a portion of messages that precede the current message are consecutive with respect to the current message of the service provider.
4. The method as recited in claim 1, further comprising:
extracting session data from a preset session database;
determining a plurality of training samples based on the session data; each training sample comprises a session identifier, corresponding session content, a session closing reason and session statistical parameters.
5. The method of claim 4, wherein the step of determining a plurality of training samples based on the session data comprises:
Dividing the session data into a plurality of pieces according to the session identification in the extracted session data; each session data corresponds to a session identifier;
and extracting a session identifier contained in each session data, wherein session content, a session closing reason and session statistical parameters corresponding to the session identifier are extracted, and a training sample corresponding to each session data is obtained.
6. The method of claim 5, wherein after obtaining the training samples corresponding to each session data, the method further comprises:
for each training sample, judging whether the training sample contains the information of the user and the information of the session service provider;
if not, the training sample is rejected.
7. The method of claim 1, wherein the preset conditions include: the prediction accuracy is greater than or equal to a preset accuracy threshold, and/or the number of current training samples input to the convolutional network is equal to a preset iteration number threshold.
8. The method of claim 1, wherein the convolutional network comprises a 3D convolutional neural network;
the step of inputting the session content of the current training sample into the convolution network of the preset initial model and outputting the feature matrix corresponding to the session content comprises the following steps:
And mapping the session content of the current training sample through the 3D convolutional neural network to obtain a feature matrix corresponding to the session content.
9. The method of claim 1, wherein the recursive network comprises a network of gated loop units of a preset number of layers; the gating circulation unit network with the preset layer number is sequentially connected; the layer number of the gating cycle unit network corresponds to the number of messages in the session content of the current training sample;
inputting the feature matrix and the session statistical parameters of the current training sample into a recursive network of the initial model, and outputting feature values corresponding to the current training sample, wherein the step comprises the following steps:
inputting a feature matrix of a message corresponding to a first layer of gating circulating unit network in the feature matrix and session statistical parameters of the current training sample into the first layer of gating circulating unit network to obtain an output result of the first layer of gating circulating unit network;
for each layer of gating and circulating unit network except the first layer of gating and circulating unit network, inputting an output result of a previous layer of gating and circulating unit network of the layer of gating and circulating unit network, a feature matrix of a message corresponding to the layer of gating and circulating unit network and session statistical parameters of the current training sample into the layer of gating and circulating unit network until the last layer of gating and circulating unit network;
And taking the output result of the last layer of gating circulating unit network as a characteristic value corresponding to the current training sample.
10. The method of claim 1, wherein the loss function comprises a softmax function; the number of the characteristic values corresponding to the current training samples is a plurality of;
inputting the characteristic value into a loss function of the initial model, and outputting a prediction result of a session closure reason of the current training sample, wherein the step comprises the following steps:
calculating an exponential function value for each of the eigenvalues by the softmax function;
determining the probability of each characteristic value according to the exponential function value of each characteristic value and the sum of the exponential function values of each characteristic value of the current training sample;
and outputting the probability of the appointed characteristic value in each characteristic value, and determining the probability of the appointed characteristic value and the session closing reason corresponding to the appointed characteristic value as the prediction result of the session closing reason of the current training sample.
11. The method of claim 1, wherein the step of controlling the online status of the customer service session based on the continuation probability comprises:
And if the continuation probability is smaller than a preset probability threshold, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
12. The method of claim 1, wherein the step of controlling the online status of the customer service session based on the continuation probability comprises:
if the continuation probability is equal to or higher than a preset probability threshold, monitoring whether the message of the user is received within a preset timeout time threshold;
if not, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
13. The method of claim 1, wherein the step of controlling the online status of the customer service session based on the continuation probability comprises:
if the continuation probability is equal to or higher than a preset probability threshold, monitoring whether the message of the user is received within a preset timeout time threshold;
if not, continuing to predict the new continuing probability of the customer service session through the probability prediction model;
and when the new continuation probability is lower than a preset probability threshold, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
14. A control device for an online customer service session, the device comprising:
the session monitoring module is used for monitoring the currently online customer service session; wherein, the customer service session is a session between a user and a service provider;
the information acquisition module is used for acquiring session information corresponding to the customer service session if the service provider does not receive the message of the user within a preset time period after sending the current message; the session information comprises session content and/or session statistical parameters in the customer service session; the session statistical parameters comprise at least two of interval duration between two adjacent messages in the customer service session, duration of the current message from the current time, message quantity of the user and message quantity of the service provider;
the probability prediction module is used for inputting the session information into a pre-trained probability prediction model to obtain the continuous probability of the customer service session; wherein the continuation probability is determined by the probability prediction model based on feature values in a feature matrix of the session information; the continuation probability is used for representing the probability of the user replying to the message after the service provider sends the current message;
The session control module is used for controlling the online state of the customer service session based on the continuation probability;
the model training module is used for determining a current training sample from a plurality of training samples, inputting the session content of the current training sample into a convolution network of a preset initial model, and outputting a feature matrix corresponding to the session content; inputting the feature matrix and session statistical parameters of the current training sample into a recursive network of the initial model, and outputting feature values corresponding to the current training sample; inputting the characteristic value into a loss function of the initial model, and outputting a prediction result of a session closing reason of the current training sample; updating the prediction accuracy of the initial model according to the session closing reason of the current training sample and the prediction result of the session closing reason; judging whether to terminate training or not based on preset conditions; if not, determining a next current training sample from a plurality of training samples, and continuing to train the convolution network, the recursion network and the loss function; and if so, taking the initial model trained by the current training sample as a probability prediction model.
15. The apparatus of claim 14, wherein the information acquisition module is configured to:
starting a timer when the service provider sends a current message; the timing duration of the timer is preset duration;
monitoring whether the message of the user is received within the timing duration of the timer;
if not, obtaining the session information corresponding to the customer service session.
16. The apparatus according to claim 14 or 15, wherein the session content comprises: the current message and at least a portion of messages that precede the current message are consecutive with respect to the current message of the service provider.
17. The apparatus of claim 14, wherein the model training module is configured to:
extracting session data from a preset session database;
determining a plurality of training samples based on the session data; each training sample comprises a session identifier, corresponding session content, a session closing reason and session statistical parameters.
18. The apparatus of claim 17, wherein the model training module is configured to:
dividing the session data into a plurality of pieces according to the session identification in the extracted session data; each session data corresponds to a session identifier;
And extracting a session identifier contained in each session data, wherein session content, a session closing reason and session statistical parameters corresponding to the session identifier are extracted, and a training sample corresponding to each session data is obtained.
19. The apparatus of claim 18, wherein the apparatus further comprises:
the message judging module is used for judging whether the training samples contain the message of the user and the message of the session service provider for each training sample;
and the sample removing module is used for removing the training sample if the training sample does not contain the user message and the session service provider message.
20. The apparatus of claim 14, wherein the preset condition comprises: the prediction accuracy is greater than or equal to a preset accuracy threshold, and/or the number of current training samples input to the convolutional network is equal to a preset iteration number threshold.
21. The apparatus of claim 14, wherein the convolutional network comprises a 3D convolutional neural network;
the model training module is used for: and mapping the session content of the current training sample through the 3D convolutional neural network to obtain a feature matrix corresponding to the session content.
22. The apparatus of claim 14, wherein the recursive network comprises a network of gated loop units of a preset number of layers; the gating circulation unit network with the preset layer number is sequentially connected; the layer number of the gating circulation unit network corresponds to the number of the messages in the session content;
the model training module is used for: inputting a feature matrix of a message corresponding to a first layer of gating circulating unit network in the feature matrix and session statistical parameters of the current training sample into the first layer of gating circulating unit network to obtain an output result of the first layer of gating circulating unit network;
for each layer of gating and circulating unit network except the first layer of gating and circulating unit network, inputting an output result of a previous layer of gating and circulating unit network of the layer of gating and circulating unit network, a feature matrix of a message corresponding to the layer of gating and circulating unit network and session statistical parameters of the current training sample into the layer of gating and circulating unit network until the last layer of gating and circulating unit network;
and taking the output result of the last layer of gating circulating unit network as a characteristic value corresponding to the current training sample.
23. The apparatus of claim 14, wherein the loss function comprises a softmax function; the number of the characteristic values corresponding to the current training samples is a plurality of;
the model training module is used for: calculating an exponential function value for each of the eigenvalues by the softmax function;
determining the probability of each characteristic value according to the exponential function value of each characteristic value and the sum of the exponential function values of each characteristic value of the current training sample;
and outputting the probability of the appointed characteristic value in each characteristic value, and determining the probability of the appointed characteristic value and the session closing reason corresponding to the appointed characteristic value as the prediction result of the session closing reason of the current training sample.
24. The apparatus of claim 14, wherein the session control module is configured to:
and if the continuation probability is smaller than a preset probability threshold, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
25. The apparatus of claim 14, wherein the session control module is configured to:
if the continuation probability is equal to or higher than a preset probability threshold, monitoring whether the message of the user is received within a preset timeout time threshold;
If not, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
26. The apparatus of claim 14, wherein the session control module is configured to:
if the continuation probability is equal to or higher than a preset probability threshold, monitoring whether the message of the user is received within a preset timeout time threshold;
if not, continuing to predict the new continuing probability of the customer service session through the probability prediction model;
and when the new continuation probability is lower than a preset probability threshold, disconnecting the customer service session or instructing the service provider to disconnect the customer service session.
27. An electronic device, comprising: a processor, a storage medium and a bus, the storage medium storing machine-readable instructions executable by the processor, the processor and the storage medium communicating over the bus when the electronic device is running, the processor executing the machine-readable instructions to perform the steps of the method of controlling an online customer service session according to any one of claims 1 to 13 when executed.
28. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, performs the steps of the method of controlling an online customer service session according to any of claims 1 to 13.
CN201910138876.5A 2019-02-25 2019-02-25 Control method and device for online customer service session and electronic equipment Active CN111611351B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910138876.5A CN111611351B (en) 2019-02-25 2019-02-25 Control method and device for online customer service session and electronic equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910138876.5A CN111611351B (en) 2019-02-25 2019-02-25 Control method and device for online customer service session and electronic equipment

Publications (2)

Publication Number Publication Date
CN111611351A CN111611351A (en) 2020-09-01
CN111611351B true CN111611351B (en) 2023-08-01

Family

ID=72199945

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910138876.5A Active CN111611351B (en) 2019-02-25 2019-02-25 Control method and device for online customer service session and electronic equipment

Country Status (1)

Country Link
CN (1) CN111611351B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112565527A (en) * 2020-12-09 2021-03-26 北京达佳互联信息技术有限公司 Session state control method and device, electronic equipment and storage medium
CN114884794A (en) * 2021-02-05 2022-08-09 华为技术有限公司 Communication method and device
CN113315876B (en) * 2021-05-27 2023-01-20 中国银行股份有限公司 Telephone bank service control method, device, server and storage medium
CN114281967A (en) * 2021-12-17 2022-04-05 深圳市欧瑞博科技股份有限公司 Intelligent processing method and device for man-machine conversation, electronic equipment and storage medium
CN114625846A (en) * 2022-02-08 2022-06-14 阿里巴巴（中国）有限公司 Online session disconnection method, device, system, medium and computer program product
CN115118762B (en) * 2022-05-19 2024-09-20 北京京东乾石科技有限公司 Session processing method, client and system
CN116522958A (en) * 2023-07-04 2023-08-01 京东科技信息技术有限公司 Session sample generation method, model training method, emotion recognition method and device
CN117809657B (en) * 2024-02-29 2024-05-24 国网山东省电力公司东营供电公司 Self-answering intelligent auxiliary telephone robot

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009005004A1 (en) * 2007-06-29 2009-01-08 Nec Corporation Session control system, session control method, and session control program
CN101764740A (en) * 2008-12-25 2010-06-30 华为技术有限公司 Method, device and system for selecting neighbor node
CN105553833A (en) * 2015-12-30 2016-05-04 上海智臻智能网络科技股份有限公司 Customer service system and service method and robot customer service thereof
CN108289053A (en) * 2017-01-10 2018-07-17 阿里巴巴集团控股有限公司 Control method, the device and system of instant communication session
CN108399201A (en) * 2018-01-30 2018-08-14 武汉大学 A kind of Web user access path prediction technique based on Recognition with Recurrent Neural Network

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9641681B2 (en) * 2015-04-27 2017-05-02 TalkIQ, Inc. Methods and systems for determining conversation quality
CN108628908B (en) * 2017-03-24 2021-02-26 北京京东尚科信息技术有限公司 Method, device and electronic equipment for classifying user question-answer boundaries
CN107995377B (en) * 2017-11-03 2020-06-16 平安科技（深圳）有限公司 Customer service management method, electronic device and computer readable storage medium

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009005004A1 (en) * 2007-06-29 2009-01-08 Nec Corporation Session control system, session control method, and session control program
CN101764740A (en) * 2008-12-25 2010-06-30 华为技术有限公司 Method, device and system for selecting neighbor node
CN105553833A (en) * 2015-12-30 2016-05-04 上海智臻智能网络科技股份有限公司 Customer service system and service method and robot customer service thereof
CN108289053A (en) * 2017-01-10 2018-07-17 阿里巴巴集团控股有限公司 Control method, the device and system of instant communication session
CN108399201A (en) * 2018-01-30 2018-08-14 武汉大学 A kind of Web user access path prediction technique based on Recognition with Recurrent Neural Network

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
融合潜在主题信息和卷积语义特征的文本主题分类;陈培新;郭武;;信号处理(第08期);68-74 *

Also Published As

Publication number Publication date
CN111611351A (en) 2020-09-01

Similar Documents

Publication Publication Date Title
CN111611351B (en) Control method and device for online customer service session and electronic equipment
JP7404467B2 (en) Dynamic communication routing based on consistency weighting and routing rules
CN110399550B (en) Information recommendation method and device
CN109990790B (en) Unmanned aerial vehicle path planning method and device
CN111339433B (en) Information recommendation method and device based on artificial intelligence and electronic equipment
US11010697B1 (en) On-demand resource scheduling
CN104750745A (en) Agents and information processing method thereof
CN112346936A (en) Application fault root cause positioning method and system
CN111935140B (en) Abnormal message identification method and device
CN105740331A (en) Information push method and device
CN113947260A (en) User satisfaction prediction method and device and electronic equipment
CN112235164B (en) Neural network flow prediction device based on controller
CN116756576B (en) Data processing method, model training method, electronic device and storage medium
CN115883392B (en) Data perception method and device of computing power network, electronic equipment and storage medium
CN117009070A (en) Method, device and equipment for constructing power-calculation scheduling knowledge graph and readable storage medium
CN116506305A (en) Network traffic prediction method, model training method and device
CN116800886A (en) Abnormal number identification method and device, storage medium and electronic equipment
CN114205320B (en) Message display method and device, electronic equipment and storage medium
CN113539275B (en) Method, device and storage medium for determining speech technology
CN115221892A (en) Work order data processing method and device, storage medium and electronic equipment
CN110633361B (en) Input control method and device and intelligent session server
CN110087230B (en) Data processing method, data processing device, storage medium and electronic equipment
Gohar et al. An online cost minimization of the slice broker based on deep reinforcement learning
CN116151601A (en) Stream service modeling method, device, platform, electronic equipment and storage medium
CN113850390A (en) Method, device, equipment and medium for sharing data in federal learning system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant