CN111917878B - Message processing method, device, equipment and storage medium - Google Patents
Message processing method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN111917878B CN111917878B CN202010766697.9A CN202010766697A CN111917878B CN 111917878 B CN111917878 B CN 111917878B CN 202010766697 A CN202010766697 A CN 202010766697A CN 111917878 B CN111917878 B CN 111917878B
- Authority
- CN
- China
- Prior art keywords
- customer service
- message
- user account
- intention information
- intention
- 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
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 21
- 238000000034 method Methods 0.000 claims abstract description 47
- 238000012545 processing Methods 0.000 claims abstract description 29
- 230000003993 interaction Effects 0.000 claims description 38
- 230000008451 emotion Effects 0.000 claims description 27
- 238000012549 training Methods 0.000 claims description 27
- 238000012937 correction Methods 0.000 claims description 25
- 238000000605 extraction Methods 0.000 claims description 7
- 230000011218 segmentation Effects 0.000 claims description 5
- 239000000284 extract Substances 0.000 claims description 4
- 238000009394 selective breeding Methods 0.000 claims 1
- 238000013473 artificial intelligence Methods 0.000 abstract description 10
- 238000003058 natural language processing Methods 0.000 abstract description 8
- 230000006870 function Effects 0.000 description 17
- 238000012544 monitoring process Methods 0.000 description 16
- 230000008569 process Effects 0.000 description 15
- 238000010586 diagram Methods 0.000 description 13
- 238000005516 engineering process Methods 0.000 description 13
- 238000004891 communication Methods 0.000 description 10
- 238000004590 computer program Methods 0.000 description 10
- 239000002699 waste material Substances 0.000 description 6
- 230000002159 abnormal effect Effects 0.000 description 4
- 238000013507 mapping Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 239000013589 supplement Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/14—Session management
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
- G10L15/18—Speech classification or search using natural language modelling
- G10L15/1822—Parsing for meaning understanding
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/28—Constructional details of speech recognition systems
- G10L15/30—Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L51/00—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
- H04L51/04—Real-time or near real-time messaging, e.g. instant messaging [IM]
- H04L51/046—Interoperability with other network applications or services
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L51/00—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
- H04L51/07—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail characterised by the inclusion of specific contents
- H04L51/10—Multimedia information
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Artificial Intelligence (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The application discloses a message processing method, a device, equipment and a storage medium, and belongs to the technical field of artificial intelligence natural language processing. The method comprises the following steps: acquiring a dialogue message between a user account and an intelligent customer service; if the conversation message between the user account and the intelligent customer service meets the condition, sending the conversation message between the user account and the intelligent customer service to a manual auxiliary platform; receiving first intention information sent by a manual assistance platform; a first reply message associated with the first intent information is determined. The scheme that this application provided reduces manpower resources consumption, improves the ability of intelligent customer service on-line service on a large scale, reduces intelligent customer service backstage because the wasting of resources that idle time overlength caused, improves the utilization ratio of resources at intelligent customer service backstage, keeps the original intention of intelligence processing simple problem, artifical solution complex problem, avoids the wasting of manpower, and the help realizes artifical working value.
Description
Technical Field
The present application relates to the field of artificial intelligence natural language processing technologies, and in particular, to a method, an apparatus, a device, and a storage medium for processing a message.
Background
The intelligent customer service is an on-line question-answering system which provides various question-answering services for users through computer equipment.
Currently, in the related art, during the interaction process between the user and the intelligent customer service, the intelligent customer service performs language analysis on the problem proposed by the user during the interaction process through the algorithm logic of the computer, and automatically replies to solve the problem. Meanwhile, the interaction process between the intelligent customer service and the user is monitored by the staff, if the problem that the intelligent customer service cannot answer occurs in the interaction process, the automatic reply is converted into the manual reply under the control of the staff, the manual customer service replaces the intelligent customer service to continue interacting with the user, and the problem answering service is provided for the user.
However, in the above related art, the interaction process between the smart customer service and the user is monitored by the staff, and the human resource consumption is large.
Disclosure of Invention
The embodiment of the application provides a message processing method, a message processing device, message processing equipment and a storage medium, so that the consumption of human resources can be reduced, and the capacity of large-scale intelligent customer service online service is improved. The technical scheme is as follows:
according to an aspect of an embodiment of the present application, there is provided a message processing method, including:
acquiring a dialogue message between a user account and an intelligent customer service;
if the conversation message between the user account and the intelligent customer service meets the condition, sending the conversation message between the user account and the intelligent customer service to a manual assistance platform;
receiving first intention information sent by the manual assistance platform, wherein the first intention information is used for indicating a conversation intention corresponding to a conversation message input by the user account;
and acquiring a first reply message associated with the first intention information, wherein the first reply message is used for replying by the intelligent customer service.
According to an aspect of an embodiment of the present application, there is provided a message processing method, including:
receiving a dialogue message between a user account and an intelligent customer service, which is sent by an intelligent customer service background;
determining candidate intention information corresponding to the dialogue message input by the user account according to the dialogue message between the user account and the intelligent customer service;
sending the dialogue information between the user account and the intelligent customer service to a working terminal, wherein the working terminal is used for displaying the dialogue information between the user account and the intelligent customer service, and the dialogue information between the user account and the intelligent customer service comprises the dialogue information between the user account and the intelligent customer service and the candidate intention information;
receiving first intention information which is sent by the working terminal and selected according to the candidate intention information, wherein the first intention information is used for indicating a conversation intention corresponding to a conversation message input by the user account;
and sending the first intention information to the smart customer service background, wherein a first reply message associated with the first intention information is used for the smart customer service to reply.
According to an aspect of an embodiment of the present application, there is provided a message processing apparatus, including:
the message acquisition module is used for acquiring a dialogue message between a user account and the intelligent customer service;
the message sending module is used for sending the conversation message between the user account and the intelligent customer service to a manual auxiliary platform if the conversation message between the user account and the intelligent customer service meets the condition;
the intention receiving module is used for receiving first intention information sent by the manual assistance platform, and the first intention information is used for indicating a conversation intention corresponding to a conversation message input by the user account;
and the reply acquisition module is used for acquiring a first reply message associated with the first intention information, and the first reply message is used for the smart customer service to reply.
According to an aspect of an embodiment of the present application, there is provided a message processing apparatus, including:
the message receiving module is used for receiving a conversation message between the user account and the intelligent customer service, which is sent by the intelligent customer service background;
the information determining module is used for determining candidate intention information corresponding to the dialogue information input by the user account according to the dialogue information between the user account and the intelligent customer service;
the information sending module is used for sending the dialogue information between the user account and the intelligent customer service to a working terminal, the working terminal is used for displaying the dialogue information between the user account and the intelligent customer service, and the dialogue information between the user account and the intelligent customer service comprises dialogue information between the user account and the intelligent customer service and the candidate intention information;
the intention receiving module is used for receiving first intention information which is sent by the working terminal and selected according to the candidate intention information, and the first intention information is used for indicating a conversation intention corresponding to a conversation message input by the user account;
and the intention sending module is used for sending the first intention information to the intelligent customer service background, wherein a first reply message associated with the first intention information is used for the intelligent customer service to reply.
According to an aspect of embodiments of the present application, there is provided a computer device including a processor and a memory, the memory having stored therein at least one instruction, at least one program, a set of codes, or a set of instructions, which is loaded and executed by the processor to implement the above-mentioned message processing method.
According to an aspect of embodiments of the present application, there is provided a computer-readable storage medium having at least one instruction, at least one program, a set of codes, or a set of instructions stored therein, which is loaded and executed by a processor to implement the above-mentioned message processing method.
According to an aspect of embodiments herein, there is provided a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the message processing method.
The technical scheme provided by the embodiment of the application can bring the following beneficial effects:
through the manual assistance platform, under the condition that the conversation information between the user account and the intelligent customer service meets the condition, the staff assists the intelligent customer service background to determine the intention information, so that the accuracy of the intention information determination is ensured, the staff does not need to monitor the conversation information between the user account and the intelligent customer service in real time, and only needs to provide assistance when the intelligent customer service works abnormally, so that the consumption of human resources is reduced, the monitoring errors caused by human factors, such as fatigue work, are avoided, and the capacity of large-scale intelligent customer service online service is improved; and after the intention information is determined, the intelligent customer service background determines the reply message according to the intention, workers do not need to arrange manual reply in real time, the timeliness of reply message feedback is ensured, the consumption of human resources is reduced, meanwhile, the resource waste of the intelligent customer service background due to overlong idle time is reduced, the resource utilization rate of the intelligent customer service background is improved, the interaction between the intelligent customer service and the user account is continuously realized by the intelligent customer service background after the first intention information is determined manually, the intelligent handling simple problem is kept, the original purpose of manually solving the complex problem is achieved, the manpower waste is avoided, and the manual work value is realized.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a schematic diagram of an intelligent customer service control system provided by one embodiment of the present application;
FIG. 2 is a flow chart of a message processing method provided by an embodiment of the present application;
FIG. 3 is a diagram illustrating a display of a reply message in a dialog interface;
fig. 4 is a flowchart of a message processing method according to another embodiment of the present application;
FIG. 5 is a diagram illustrating a display manner of dialog information in a user interface;
FIG. 6 illustrates a flow diagram of the interaction of the smart customer service back office with the human assistance platform;
FIG. 7 is a block diagram of a message processing apparatus provided in one embodiment of the present application;
fig. 8 is a block diagram of a message processing apparatus according to another embodiment of the present application;
FIG. 9 is a block diagram of a message processing apparatus according to another embodiment of the present application;
fig. 10 is a block diagram of a message processing apparatus according to another embodiment of the present application;
fig. 11 is a block diagram of a computer device according to an embodiment of the present application.
Detailed Description
To make the objects, technical solutions and advantages of the present application more clear, embodiments of the present application will be described in further detail below with reference to the accompanying drawings.
Artificial Intelligence (AI) is a theory, method, technique and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and expand human Intelligence, perceive the environment, acquire knowledge and use the knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technique of computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can react in a manner similar to human intelligence. Artificial intelligence is the research of the design principle and the realization method of various intelligent machines, so that the machines have the functions of perception, reasoning and decision making.
The artificial intelligence technology is a comprehensive subject and relates to the field of extensive technology, namely the technology of a hardware level and the technology of a software level. The artificial intelligence infrastructure generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and the like.
Natural Language Processing (NLP) is an important direction in the fields of computer science and artificial intelligence. It studies various theories and methods that enable efficient communication between a person and a computer using natural language. Natural language processing is a science integrating linguistics, computer science and mathematics. Therefore, the research in this field will involve natural language, i.e. the language that people use everyday, so it is closely related to the research of linguistics. Natural language processing techniques typically include text processing, semantic understanding, machine translation, robotic question and answer, knowledge mapping, and the like.
In the embodiment of the application, the computer device performs Text preprocessing (Text preprocessing) and Semantic understanding (Semantic understanding) on the dialog messages input by the user account based on a natural language processing technology, determines the intention information corresponding to the dialog messages input by the user, acquires the reply messages corresponding to the intention information from a reply message library according to the intention information, and further displays the reply messages to the user by the intelligent customer service through the user terminal. The intention information is used for indicating a conversation purpose corresponding to the conversation message input by the user, and the reply message is used for providing an auxiliary effect for the user to realize the conversation purpose. For example, the user intends to solve problem a and the reply message is how to solve problem a.
By way of example, with reference to fig. 1, a schematic diagram of an intelligent customer service control system provided by an embodiment of the present application is shown. The intelligent customer service control system can comprise a user terminal 10, an intelligent customer service background 20, a manual assistance platform 30, a work terminal 40 and an intelligent training background 50.
The user terminal 10 is used for acquiring a dialog message between a user account and an intelligent customer service, and the user terminal 10 refers to a terminal used by any user. Alternatively, the user terminal 10 may be an electronic device such as a mobile phone, a tablet Computer, a game console, an e-book reader, a multimedia playing device, a wearable device, a PC (Personal Computer), and the like. A client of the application may be installed in the user terminal 10. The application program refers to any computer program capable of providing an interaction platform between a user account and the intelligent customer service. Optionally, the intelligent customer service may be a customer service for the application program, or a customer service for another application program, which is not limited in this embodiment of the present application. Taking the purchase application as an example, the intelligent customer service may be a customer service of the purchase application, a merchant customer service using the purchase application, or a customer service of a payment application associated with the purchase application and displayed in the purchase application. The application program may be an application program that needs to be downloaded and installed, or may be an application program that is ready to use, and this is not limited in this embodiment of the present application.
In one possible implementation, the intelligent customer service is displayed in the application in the form of a customer service account number. Optionally, the user account number is conversed with the intelligent customer service through the customer service account number in the interaction platform provided by the application program. For example, the intelligent customer service of the game application is displayed in the social application in the form of a customer service account, and the user account can be in conversation with the intelligent customer service of the corresponding game application through the customer service account in the social application. Certainly, in order to meet the requirements of different users, the intelligent customer service may correspond to multiple customer service accounts, and then different user accounts may have a conversation with the intelligent customer service in the interaction platforms provided by different application programs through different customer service accounts. For example, the smart customer service of the game application has a first customer account and a second customer account, the user account may have a smart customer service session with the game application through the first customer account in the first social application, and similarly, other user accounts may have a smart customer service session with the game application through the second customer account in the second social application.
In another possible embodiment, the intelligent customer service is displayed in the application in the form of a customer service icon. Optionally, the user account is conversed with the intelligent customer service in an interaction platform provided by the application program through a customer service icon. For example, by clicking on a client in the application program, a jump is made to a dialog interface, and the user account is then in dialog with the smart customer service in the dialog interface.
In this embodiment, the user terminal 10 may obtain a session message between the user account and the smart customer service during a session between the user account and the smart customer service, and send the session message to the smart customer service backend. The dialog message may include a dialog message input by the user account, or may include a dialog message input by the user account and a dialog message replied by the smart customer service at the same time.
The intelligent customer service background 20 is used for controlling the interaction between the intelligent customer service and the user account. The intelligent customer service is the mapping of the intelligent background 20 on the application program. Optionally, an auto-reply program is set in the smart customer service backend 20, and the auto-reply program refers to a computer program having an auto-reply function for a dialog message input by a user account. Optionally, the intelligent customer service background 20 analyzes and processes the dialog message input by the user account through an automatic reply program, determines intention information corresponding to the dialog message input by the user, and obtains a corresponding reply message from a reply message library according to the intention information. Further, the smart customer service background 20 sends the reply message to the user terminal 10, and correspondingly, the smart customer service displays the reply message to the user account in the user terminal 10. It should be noted that, in this embodiment of the application, in the case that the intelligent customer service back-stage 20 determines that the session message between the user account and the intelligent customer service meets the condition, the intelligent customer service back-stage may send the session message between the user account and the intelligent customer service to the artificial assistance platform 30, and obtain the corresponding reply message from the reply message library according to the intention information manually selected by the artificial assistance platform 30, and the intelligent customer service displays the reply message in the user terminal 10.
The manual assistance platform 30 is used for providing an assistance function for the intelligent customer service background 20, and optionally, the manual assistance platform 30 obtains a manual selection through the work terminal 40 to provide an assistance function for the intelligent customer service background. For example, the human assistance platform 30 may be some backend server of an application that establishes a connection with the work terminal 40. Optionally, the smart customer service background 20 sends a conversation message between the user account and the smart customer service to the manual assistance platform 30 to seek assistance from the manual assistance platform 30, in case it is determined that the conversation message between the user account and the smart customer service satisfies the condition. Correspondingly, the manual assistance platform 30 analyzes the dialog message input by the user account, determines candidate intention information of the dialog message input by the user account, packages the dialog message input by the user account and the candidate intention information of the dialog message to generate dialog information, and sends the dialog information to the work terminal 40. Correspondingly, the human assistance platform 30 determines the manually selected intention information corresponding to the dialog message input by the user through the working terminal 40, and sends the manually selected intention information to the intelligent customer service background 20. Optionally, in this embodiment of the present application, after acquiring the manually selected intention information, the manual assistance platform 30 packages the manually selected user intention information with the dialog message input by the user account to generate correction information, and sends the correction information to the intelligent training platform 50.
The work terminal 40 refers to a terminal used by a worker. Alternatively, the work terminal 40 may be an electronic device such as a mobile phone, a tablet Computer, a game console, an electronic book reader, a multimedia player, a wearable device, a PC (Personal Computer), and the like. In the embodiment of the present application, after receiving the dialog information, the work terminal 40 displays the dialog message input by the user account and the candidate intention information in the user interface, and further, the worker selects corresponding intention information from the candidate intention information according to the dialog message input by the user, and sends the manual selection result to the manual assistance platform 30. Wherein the manual selection result is used for indicating manually selected intention information corresponding to the dialog message input by the user.
The intelligent training platform 50 is used to train the intelligent customer service back office 20. Optionally, after acquiring the correction information, the intelligent training platform 50 trains the intelligent customer service background 20 according to the correction information, so as to improve the ability of the intelligent customer service background 20 to determine the intention information. In a possible implementation manner, the smart customer service backend 20 is provided with a knowledge graph, the knowledge graph includes an association relationship between the intention information, and the smart training platform 50 may also supplement the knowledge graph during the training of the smart customer service backend 20, for example, increase the association relationship between the intention information, add a new user intention, and the like. Of course, during the training process, the smart customer service backend 20 may also supplement or modify the reply messages in the reply message library to optimize the corresponding reply messages.
It should be noted that the intelligent customer service background 20, the manual assistance platform 30, and the intelligent training background 50 may be disposed in the same computer device, or may be disposed in different computer devices, which is not limited in this embodiment of the present invention.
Optionally, the user terminal 10, the intelligent customer service background 20, the manual assistance platform 30, the work terminal 40 and the intelligent training background 50 communicate with each other through a network.
The technical solution of the present application will be described in detail with reference to several embodiments.
Referring to fig. 2, a flowchart of a message processing method according to an embodiment of the present application is shown. The execution subject of each step may be a computer device, such as the intelligent customer service background 20 in fig. 1. The method can comprise the following steps (201-204):
The user account refers to an account used by any user. Optionally, the user uses the user account to talk to the intelligent customer service. In a possible implementation manner, the user account is an account corresponding to an application program, and is used for logging in the application program by using the user account and performing a conversation with the intelligent customer service in an interaction platform provided by the application program. Alternatively, the application may be any computer program that provides an interactive platform between the user's account and the intelligent customer service. The application program may be an application program that needs to be downloaded and installed, or may be an application program that is to be used on demand, which is not limited in this embodiment of the application.
The intelligent customer service refers to the mapping of the intelligent customer service background in the application program. That is, the smart service is a representation of the smart service backend at the user terminal. Optionally, the user can realize a conversation with the intelligent customer service background through the intelligent customer service. In one possible implementation, the intelligent customer service is displayed in the application in the form of a customer service account number. Optionally, the user account number is conversed with the intelligent customer service through the customer service account number in the interaction platform provided by the application program. In another possible embodiment, the intelligent customer service is displayed in the application in the form of a customer service icon. Optionally, the user account is conversed with the intelligent customer service in an interaction platform provided by the application program through a customer service icon.
In the embodiment of the application, the intelligent customer service background can acquire the conversation message between the user account and the intelligent customer service in the conversation process between the user account and the intelligent customer service. Optionally, the dialog message between the user account and the intelligent customer service includes a dialog message input by the user account and a dialog message replied by the intelligent customer service. In a possible implementation manner, the intelligent customer service background acquires the dialogue message input by the user account through the user terminal, and acquires the dialogue message replied by the intelligent customer service through the intelligent customer service background. Optionally, after detecting the dialog message input by the user account, the user terminal sends the dialog message input by the user account to the smart customer service background in real time if the dialog message input by the user is directed to the smart customer service. Correspondingly, the intelligent customer service background receives the dialogue message input by the user account, determines a corresponding reply message according to the dialogue message input by the user account, and records the reply message as the dialogue message replied by the intelligent customer service.
Optionally, the dialog message refers to communication information generated in the dialog process, and the communication information may include a text message, a voice message, a picture message, or the like, which is not limited in this embodiment of the application. The text message refers to a dialog message displayed in a text message, the voice message refers to a dialog message displayed in a sound form, and the picture message refers to a dialog message displayed in an image form.
The condition is a judgment condition for judging whether the manual assistance platform is required to provide help. And the manual auxiliary platform is used for providing an auxiliary function for the intelligent customer service background. Optionally, after acquiring the session message between the user account and the smart customer service, the smart customer service background determines the session message between the user account and the smart customer service based on the determination condition. And if the conversation message between the user account and the intelligent customer service meets the condition, determining that the manual assistance platform is required to provide help, and sending the conversation message between the user account and the intelligent customer service to the manual assistance platform.
In one possible embodiment, the condition includes a first condition. Wherein, the first condition refers to a judgment condition of the dialog message input aiming at the user account. In this embodiment of the application, the intelligent customer service background may determine, based on the first condition, a dialog message input by the user account, and determine that the assistance of the manual assistance platform is required when the dialog message input by the user account satisfies the first condition. Optionally, the step 202 includes the following steps:
1. and monitoring the dialogue messages input by the user account.
In this embodiment of the application, after acquiring the session message input by the user account sent by the user terminal, the smart customer service background may monitor the session message input by the user account to determine whether the session message input by the user account satisfies the first condition.
In a possible implementation manner, the intelligent customer service background monitors keywords corresponding to the dialog messages input by the user account. Optionally, after obtaining the dialog message input by the user account, the intelligent customer service background extracts a keyword from the dialog message input by the user account, obtains a corresponding keyword, and determines that the dialog message input by the user account satisfies the first condition when the keyword includes a negative emotion keyword.
The negative emotion keywords refer to keywords indicating discontent emotion of the user. Optionally, more than one negative emotion keyword may be stored in the smart customer service background in advance, after the keywords are acquired, the keywords are traversed, and if the keywords include the prestored negative emotion keywords, it is determined that the dialog message input by the user account satisfies the first condition. It should be noted that, in order to ensure the accuracy of the determination result, the intelligent customer service background may update the negative emotion keywords at certain time intervals. The updating method of the negative emotion keyword includes, but is not limited to: retrieving, acquiring and storing new negative emotion keywords, and removing the infrequent negative emotion keywords. Of course, in another possible implementation, after obtaining the above keyword, the smart client may also perform semantic analysis on the keyword to determine whether the keyword includes a negative emotion keyword.
In another possible implementation, the intelligent customer service background monitors the accuracy probability of the intention information corresponding to the dialog message input by the user account. Optionally, after acquiring the dialog message input by the user account, the smart customer service background acquires intention information corresponding to the dialog message input by the user account and an accurate probability of the intention information. The accuracy probability is used to indicate the matching degree between the intention information and the dialog message input by the user account, that is, the accuracy probability may be used to characterize the accuracy of the intention information. And under the condition that the accuracy probability is smaller than a first threshold value, the intelligent customer service background determines that the dialogue message input by the user account meets a first condition. The intention information refers to a dialog intention corresponding to a dialog message input by the user account, and the dialog intention can be used for representing a problem to be solved by the user. For example, the above-mentioned conversation message is "yesterday has a customer who purchased 243 dollars of merchandise with 24.3 dollars, i have no time to verify that he is not already connected and what is done at present", and the intention information corresponding to the conversation message is "pay less" and money recollection ".
Of course, in other possible implementations, the intelligent customer service background may monitor the keyword and the accurate probability at the same time, which is not limited in this application.
2. And if the conversation message input by the user account meets the first condition, sending the conversation message between the user account and the intelligent customer service to the manual assistance platform.
In the embodiment of the application, after determining that the dialog message input by the user account meets the first condition, the intelligent customer service background sends the dialog message between the user account and the intelligent customer service to the manual assistance platform. Optionally, the intelligent customer service background sends the dialog message between the user account and the intelligent customer service to the manual assistance platform, and at the same time, sends the reason why the dialog message input by the user account meets the first condition. For example, the reason may be that the user expresses an discontent emotion, that the intention information determination is ambiguous, or the like.
In another possible embodiment, the condition includes a second condition. The second condition refers to a judgment condition for the dialog message replied by the intelligent customer service. In this embodiment of the application, the intelligent customer service background may determine, based on the second condition, a dialog message replied by the intelligent customer service, and determine that the assistance of the manual assistance platform is needed when the dialog message replied by the intelligent customer service satisfies the second condition. Optionally, the step 202 includes the following steps:
1. and monitoring the dialogue messages replied by the intelligent customer service.
In this embodiment, after recording the dialog message replied by the smart customer service, the smart customer service background may monitor the dialog message replied by the smart customer service to determine whether the dialog message replied by the smart customer service satisfies the second condition.
In one possible implementation, the smart customer service background monitors the number of replies to the smart customer service. And the reply times are the times of acquiring the reply message by the intelligent customer service background. Optionally, after recording the dialog message replied by the intelligent customer service, the intelligent customer service background acquires the number of replies to the user account by the intelligent customer service in the current interaction, and determines that the dialog message replied by the intelligent customer service satisfies the second condition when the number of replies is greater than a threshold value.
In another possible implementation manner, the intelligent customer service background monitors the amount of intention information corresponding to the dialog message replied by the intelligent customer service. Optionally, after recording the dialog message replied by the smart customer service, the smart customer service background acquires the number of intention information included in the dialog message replied by the smart customer service, and if the number of intention information is not unique, acquires the historical dialog message replied by the smart customer service, and determines the historical reply frequency of the dialog message including the number of intention information that is not unique from the historical dialog message replied by the smart customer service, and determines that the dialog message replied by the smart customer service satisfies the second condition when the historical reply frequency is greater than the target value. The historical dialogue message refers to all dialogue messages replied by the intelligent customer service in the interaction.
In another possible implementation, the intelligent customer service background monitors the number of times of displaying the dialog message replied by the intelligent customer service. The display times refer to the times of the intelligent customer service replying the same dialogue message in the interaction. Optionally, after recording the dialog message replied by the smart customer service, the smart customer service background obtains the number of times of displaying the dialog message replied by the smart customer service, and determines that the dialog message replied by the smart customer service satisfies the second condition when the number of times of displaying is greater than a threshold value. Optionally, the smart customer service background may traverse through words in the historical dialog messages of the smart customer service to determine the number of presentations. In another possible implementation, the smart customer service may compare the identifications of the historical dialog messages of the smart customer service to determine the number of presentations. Wherein the same historical dialog messages have the same identity.
Certainly, in other possible embodiments, the smart customer service background may simultaneously monitor the number of replies, the number of intention messages, and the number of presentations, which is not limited in this embodiment of the present application.
2. And if the conversation message replied by the intelligent customer service meets a second condition, sending the conversation message between the user account and the intelligent customer service to the manual assistance platform.
In the embodiment of the application, after determining that the dialog message replied by the intelligent customer service meets the second condition, the intelligent customer service background sends the dialog message between the user account and the intelligent customer service to the manual assistance platform. Optionally, the intelligent customer service background sends the dialog message between the user account and the intelligent customer service to the manual assistance platform, and at the same time, sends the reason why the dialog message replied by the intelligent customer service meets the second condition. For example, the reason may be that the number of replies is too large, the number of times the intention information cannot be determined is too large, or the same message is replied a plurality of times, or the like.
Of course, in other possible embodiments, the intelligent customer service background may monitor the dialog message input by the user account and the dialog message replied by the intelligent customer service at the same time, which is not limited in this application.
The first intention information is used for indicating a dialog intention corresponding to the dialog message input by the user account, and the dialog intention can be used for characterizing the problem to be solved by the user. In the embodiment of the application, after the intelligent customer service background sends the dialogue message between the user account and the intelligent customer service background to the manual auxiliary platform, the manual auxiliary platform obtains a manual selection result based on the working terminal, determines first intention information according to the manual selection result, and sends the first intention information to the intelligent customer service background. Correspondingly, the intelligent client background receives the first intention information sent by the artificial assistance platform.
The working terminal is a terminal used by a worker, and the worker is a worker responsible for carrying out a manual monitoring part on the conversation message received by the manual auxiliary platform; the manual selection result is used for indicating the conversation intention determined by the staff according to the conversation message input by the user account.
At step 204, a first reply message associated with the first intention information is obtained.
The first reply message is used for the intelligent customer service to reply. The first reply message may be a text message, a voice message, a picture message, or the like, which is not limited in this embodiment of the application. Optionally, after receiving the first intention message, the smart customer service background acquires a first reply message associated with the first intention information, and sends the first reply message to the user terminal. Correspondingly, the user terminal receives the first reply message, and the intelligent customer service displays the first reply message to the user in the conversation interface. In the embodiment of the application, the reply messages associated with different intention information are different.
In one possible embodiment, the number of reply messages associated with the first intention message is unique. Optionally, after receiving the first intention message, the smart customer service background uses a reply message uniquely associated with the first intention message as the first reply message.
In another possible embodiment, the number of reply messages associated with the first intention message is not unique. Optionally, after receiving the first intention information, the smart customer service background randomly selects any reply message associated with the first intention information as a first reply message. Or, the smart customer service background acquires the first reply message associated with the first intention information according to the user information corresponding to the user account, optionally, step 204 includes the following steps:
1. and acquiring the reply message associated with the first intention information from the reply message library as a candidate reply message.
The reply message library refers to a database for storing reply messages. In the embodiment of the application, more than one intention information is included in the reply message library, and the reply messages are associated with the intention information. In one possible embodiment, the reply message library stores different reply messages in a classified manner. Optionally, the reply message library stores reply messages corresponding to different intention information in different file directories, so that the intelligent customer service background can obtain the associated reply message after determining the intention information. Of course, the reply message library may store different types of reply messages in a classified manner for the same intention information. For example, according to the content type of the reply message, storing the text message, the voice message and the picture message in a file directory of the intention information in a classified manner; for another example, the reply message is sorted and stored according to the association degree of the reply message and the user information. In another possible embodiment, the reply message library classifies and identifies different reply messages. Optionally, the reply message library employs the feature identifier to mark the reply message associated with the same intention information, so that the intelligent customer service background can retrieve the associated reply message according to the feature identifier after determining the intention information.
In the embodiment of the application, after receiving the first intention information, the smart customer service background acquires a reply message associated with the first intention from a reply message library as a candidate reply message.
2. And acquiring a reply message matched with the user information corresponding to the user account from the candidate reply messages as a first reply message.
The user information refers to information which is associated with the user account and used for representing the characteristics of the user. The user information includes, but is not limited to, at least one of: the current geographic position corresponding to the user account, the historical geographic position corresponding to the user account, the user conversation habit corresponding to the user account, the user age corresponding to the user account, and the user gender corresponding to the user account. The current geographic position is used for indicating the real-time geographic position of the user; the historical geographic position is used for indicating the geographic position of the user with the stay time exceeding a target numerical value; the conversation habit is used for indicating the communication habit of the user, such as voice communication, character communication or picture communication.
In this embodiment of the application, after acquiring the candidate reply message, the smart customer service background acquires the user information corresponding to the user account, and acquires a matched reply message from the candidate reply message as a first reply message according to the user information. Optionally, the smart customer service background may analyze the candidate message to determine a first reply message matching the user information; the first reply message matching the user information may also be determined based on the identity of the candidate message, e.g., the identity of the candidate reply message is a numeric identity, the second digit in the identity indicates current geographical location information, the third through fifth digits indicate historical geographical location information, etc.
It should be noted that, in this embodiment of the application, after the smart customer service displays the first reply message to the user, if the user account continues to input the dialog message for the first intention information, there may be a case that the first reply message cannot provide accurate help without the user, that is, the first reply message for the first intention information cannot be understood by the user, and at this time, the smart customer service may explain the first reply message in more detail. Optionally, after the first reply message is acquired, in a case that the smart customer service background acquires a second dialogue message input by the user account, performing word segmentation processing and keyword extraction on the second dialogue message, and determining third intention information corresponding to the second dialogue message. If the third intention information is the same as the first intention information, a second reply message associated with the first reply message is obtained, the second reply message is sent to the user terminal, correspondingly, the intelligent customer service displays the second reply message to the user in the conversation interface, the first reply message is explained in detail, and the intelligent customer service is guaranteed to accurately provide help for the user. The second reply message refers to an explanation message for the first reply message. For example, the first reply message is "the unpaid money can be recovered by leaving a message from the user", and the second reply message is "the message can be left to the user in the a or B mode, and a message reminder is initiated to the user to recover the unpaid money".
In summary, in the technical scheme provided in the embodiment of the present application, through the manual assistance platform, when the session message between the user account and the intelligent customer service satisfies the condition, the worker assists the intelligent customer service background to determine the intention information, so as to ensure the accuracy of the intention information determination, and the worker does not need to monitor the session message between the user account and the intelligent customer service in real time, and only needs to provide an assistance function when the intelligent customer service is abnormal, so as to reduce the consumption of human resources, avoid monitoring errors caused by human factors, such as fatigue work, and improve the capacity of the large-scale intelligent customer service online service; and after the intention information is determined, the intelligent customer service background determines the reply message according to the intention, workers do not need to arrange manual reply in real time, the timeliness of reply message feedback is ensured, the consumption of human resources is reduced, meanwhile, the resource waste of the intelligent customer service background due to overlong idle time is reduced, the resource utilization rate of the intelligent customer service background is improved, the interaction between the intelligent customer service and the user account is continuously realized by the intelligent customer service background after the first intention information is determined manually, the intelligent handling simple problem is kept, the original purpose of manually solving the complex problem is achieved, the manpower waste is avoided, and the manual work value is realized.
In addition, the conversation information input by the user account is monitored by taking the first condition as a reference, and help can be requested from the manual assistance platform in time under the condition that the conversation information input by the user account has problems; monitoring the conversation message replied by the intelligent customer service by taking a second condition as a reference, and timely requesting help from the artificial auxiliary platform when the conversation message replied by the intelligent customer service has a problem; and the interaction process is monitored from two aspects of the user account and the intelligent customer service, so that the timeliness of requesting help from the manual auxiliary platform can be ensured while the accuracy of the monitoring result is ensured.
In addition, the first reply message is matched according to the user information corresponding to the user account, so that the matching degree between the first reply message and the user is ensured, and the user can accurately understand the meaning of the first reply message.
In addition, for the condition that the user carries out the question hunting under the same intention information, the explanation message of the first reply message is displayed to the user, and the probability of solving the problem by the user is improved.
It should be noted that, in the embodiment of the present application, after receiving the first intention information, the smart customer service background may lock an intention for the user account according to the first intention information in order to facilitate smooth subsequent conversation between the smart customer service and the user account. Optionally, the step 203 includes the following sub-steps:
1. according to the first intention information, second intention information associated with the first intention information is determined.
In one possible embodiment, the first intention information is lower intention information corresponding to the second intention information, and the second intention information is upper intention information corresponding to the first intention information. Optionally, the content indicated by the first intention information is overlaid in the content indicated by the second intention information. For example, the second intention information is "payment related", and the first intention information is "XX platform payment". In another possible implementation, the second intention information is intention information which is stored in the smart customer service background in advance and is associated with the first intention information. The related information of the first intention information and the second intention information can be stored in the intelligent customer service background by a worker according to experience in advance. For example, the second intention information is "payment-related", and the first intention information is "inquiry of other person information".
In the embodiment of the application, after receiving the first intention information, the smart customer service background determines second intention information associated with the first intention information according to the first intention information.
2. And establishing a locking relation between the user account and the second intention information.
The locking relation is used for indicating the acquisition range of the intention information corresponding to the conversation message input by the user account subsequently. In the embodiment of the application, after the smart customer service background determines the second intention information, a locking relationship is established between the user account and the second intention information. It should be noted that the locking relationship only takes effect in the process of the interaction.
3. And in the case of receiving the first dialogue message input by the user account, determining intention information corresponding to the first dialogue message from the intention range indicated by the second intention information.
The intention range refers to an intention determination range for a dialog message input for a user account. The first dialogue message is the dialogue message input by the user account after the locking relation is established. Optionally, the smart customer service background determines intention information corresponding to the first dialogue message from the intention range indicated by the second intention information when receiving the first dialogue message input by the user account.
In the embodiment of the application, after receiving the first dialogue message, the smart customer service background extracts keywords from the first dialogue message to obtain the keywords in the first dialogue message. If the keyword comprises a keyword used for indicating other intention information, the locking relation between the user account and the second intention information is released; and if the keywords do not comprise the keywords used for indicating other intention information, determining the intention information corresponding to the first dialogue message from the intention range indicated by the second intention information, and acquiring the reply message associated with the intention information corresponding to the first dialogue message. The other intention information means intention information outside the above-mentioned intention range.
Optionally, the smart customer service background may store an indication relationship between each intention information and a keyword, and after acquiring the keyword of the first dialogue message, the keyword may be traversed to determine whether the keyword includes a keyword indicating other intention information. Of course, in order to ensure the accuracy of judgment of the indication relationship, the indication relationship is used only to represent a strong correlation between the keyword and the intention information.
The above describes a reply mode of the intelligent customer service background with the help of the manual assistance platform. In the following, a simple introduction is made to the reply rule when the intelligent customer service background does not obtain the help of the manual assistance platform (e.g., the manual assistance platform has no feedback for a long time). Optionally, the reply rule of the smart customer service background is as follows:
1. the method comprises the steps of performing word segmentation processing and keyword extraction on conversation messages input by a user account, and determining intention information corresponding to the conversation messages input by the user account and the accurate probability corresponding to the intention information.
The accuracy probability is used to indicate the degree of match between the intent information and the dialog message entered by the user account. In the embodiment of the application, after acquiring the dialog message input by the user account, the intelligent customer service background performs word segmentation and keyword extraction on the dialog message input by the user account, and determines the intention information corresponding to the dialog message input by the user account and the accuracy probability corresponding to the intention information according to the keyword.
Optionally, a conscious-graph determination model may be set in the smart customer service background. After the keywords are acquired, the keywords are input into the intention determining model, and then the intention determining module outputs intention information corresponding to the dialogue information input by the user account and the accurate probability corresponding to the intention information. The training samples of the intention determining model are various dialogue messages and intention information corresponding to the dialogue messages.
2. And if the accuracy probability is smaller than a first threshold value, determining that the dialog message replied by the intelligent customer service includes first prompt information.
The first prompt message is used to prompt for a new dialog message to be re-entered. In this embodiment of the application, if the accuracy probability is smaller than a first threshold, the intelligent customer service determines that the intention information corresponding to the dialog message input by the user account cannot be acquired, and further determines that the dialog message replied by the intelligent customer service includes the first prompt information to prompt the user to re-input a new dialog message, and determines the corresponding intention information through the new dialog message to ensure smooth interaction between the user account and the intelligent customer service.
3. And if the accuracy probability is greater than the first threshold and less than the second threshold, determining that the dialog message replied by the intelligent customer service includes second prompt information.
The second prompt message is used for prompting the intention information corresponding to the dialogue message input by the selected user account. Wherein the first threshold is less than the second threshold. In this embodiment of the application, if the accuracy probability is greater than a first threshold and less than a second threshold, the intelligent customer service determines that the intention information corresponding to the dialog message input by the user account cannot be accurately obtained, and further determines that the dialog message replied by the intelligent customer service includes second prompt information, where the second prompt information includes more than one intention information, and the user may select the intention information corresponding to the dialog message input by the user account by clicking the corresponding intention information, and correspondingly, the intelligent customer service background determines the corresponding intention information according to the user selection, so as to ensure smooth interaction between the user account and the intelligent customer service.
4. And if the accuracy probability is greater than a second threshold value, selecting a reply message corresponding to the intention information from a reply message library.
In this embodiment of the application, if the accuracy probability is greater than the second threshold, the intelligent customer service determines that the intention information is intention information corresponding to the dialog information input by the user account, selects a reply message corresponding to the intention information from a reply message library, and displays the reply message to the user by the intelligent customer service to solve the problem expected to be solved by the user.
Illustratively, referring to fig. 3 in combination, the reply rule when the smart customer service background does not obtain the help of the human-assisted platform is visually displayed.
The user account 31 is included in the dialog interface 30 to input a first message 32, and accordingly, the smart customer service background acquires the first message 32 and determines corresponding intention information. And under the condition that the accuracy probability corresponding to the intention information acquired by the intelligent customer service background is smaller than a first threshold, the intelligent customer service 33 displays a first reply 35 in the conversation interface 34, wherein the first reply 35 comprises first prompt information for prompting the user to re-input a new conversation message. Under the condition that the accuracy probability corresponding to the intention information acquired by the intelligent customer service background is greater than the first threshold and smaller than the second threshold, the intelligent customer service 33 displays a second reply 37 in the conversation interface 36, wherein the second reply 37 comprises a plurality of intention information to prompt the user to select appropriate intention information according to the own requirements. And under the condition that the accuracy probability corresponding to the intention information acquired by the intelligent customer service background is greater than a second threshold, the intelligent customer service 33 displays a third reply 39 in the conversation interface 38, wherein the third reply 39 is a reply message acquired by the intelligent customer service background from the reply message library according to the intention information.
Referring to fig. 4, a flowchart of a message processing method according to another embodiment of the present application is shown. The execution subject of each step may be a computer device, such as the human assistance platform 30 in fig. 1. The method may comprise the following steps (401 to 405):
The user account refers to an account used by any user. Optionally, the user uses the user account to talk to the intelligent customer service. The intelligent customer service refers to the mapping of the intelligent customer service background in the application program. The dialogue message refers to communication information generated in the process of communication between the user account and the intelligent customer service.
In the embodiment of the application, the intelligent customer service background sends the dialogue message between the user account and the intelligent customer service to the manual assistance platform under the condition that the dialogue message between the user account and the intelligent customer service meets the conditions.
The candidate intention information is used to indicate possible situations of the first intention information. Optionally, the first intention information is determined by selecting from candidate intention information. The first intention information is used for indicating a dialogue intention corresponding to the dialogue message input by the user account, and the dialogue intention can be used for representing the problem to be solved by the user. In the embodiment of the application, after the manual assistance platform acquires the dialog message between the user account and the intelligent customer service, the corresponding candidate intention information can be determined according to the dialog message between the user account and the intelligent customer service.
In one possible implementation, the human-assisted platform invokes the intent recognition model to obtain the candidate intent information. Optionally, the dialog message between the user account and the intelligent customer service includes a dialog message input by the user account, the artificial assistant platform calls the intention recognition model after acquiring the dialog message input by the user account, and determines the first candidate intention information corresponding to the dialog message input by the user account according to the dialog message input by the user account. The intention recognition model is trained on the historical dialogue messages and the intention information corresponding to the historical dialogue messages, and the selection results refer to manual selection results obtained through the work terminal. Further, the first candidate intention information and the second candidate intention information are matched, and intention information belonging to both the first candidate intention information and the second candidate intention information is selected as candidate intention information.
The second candidate intention information is intention information corresponding to a candidate dialogue intention of a dialogue message input by a user and determined by the intelligent customer service background, and the accuracy probability of the second candidate intention information is greater than the first threshold and smaller than the second threshold. Optionally, the manual assistance platform acquires the second candidate intention information sent by the smart customer service background while acquiring the dialog message between the user account and the smart customer service. Of course, in the case that the intelligent customer service background cannot determine the second candidate intention, the human assistance platform may directly take the first candidate intention as the candidate intention.
It should be noted that, in this embodiment of the application, the intelligent customer service background sends, to the manual assistance platform, the reason why the session message between the user account and the intelligent customer service background meets the above condition while sending the session message between the user account and the intelligent customer service background, and correspondingly, after the manual assistance platform obtains the session message input by the user account, the manual assistance platform may process the session message input by the user account based on the reason, so as to obtain more accurate candidate intention information. For example, different intention recognition models can be trained in the manual assistance platform according to different reasons, so that the intention recognition models can obtain accurate candidate intention information for different reasons.
And step 403, sending dialogue information between the user account and the intelligent customer service to the working terminal.
The working terminal refers to a terminal used by a worker. In the embodiment of the application, the working terminal is used for displaying the dialogue information between the user account and the intelligent customer service. The dialogue information between the user account and the intelligent customer service comprises dialogue information between the user account and the intelligent customer service and the candidate intention information. In one possible implementation mode, in order to ensure the tidiness of the user interface, the work terminal only displays the dialogue messages input by the user and the candidate intention information corresponding to the dialogue messages input by the user. As shown in fig. 5, a first message 51 and a second message 52 entered by a user are presented in a user interface. When the worker clicks on the first message 51, the first candidate intent 53, the second candidate intent 54, and the third candidate intent 55 corresponding to the first message 51 are presented in the user interface. The presentation positions of the first candidate intention 53, the second candidate intention 54 and the third candidate intention 55 can be sorted from large to small according to the accuracy probability, and the first candidate intention 53 with the highest accuracy probability is presented on the top. Similarly, when the worker clicks on the second message 52, the fourth candidate intent 56 and the fifth candidate intent 57 corresponding to the second message 52 are presented in the user interface. In another possible implementation mode, in order to ensure the accuracy of the selection of the staff member, the working terminal simultaneously displays the dialogue messages input by the user account and the dialogue messages replied by the intelligent customer service. Of course, in practical applications, in a case that the user terminal does not display the dialog message replied by the intelligent customer service, the dialog message between the user account and the intelligent customer service may include the dialog message input by the user account and the candidate intention information.
Optionally, in this embodiment of the application, after obtaining the candidate intention information, the manual assistance platform may generate a dialog message between the user account and the intelligent customer service according to the candidate intention and the dialog message between the user account and the intelligent customer service, and send the dialog message between the user account and the intelligent customer service to the work terminal. Correspondingly, the working terminal receives and displays the dialogue information between the user account and the intelligent customer service.
And step 404, receiving first intention information which is sent by the working terminal and selected according to the candidate intention information.
The first intention information is used for indicating the dialogue intention corresponding to the dialogue information input by the user account. In the embodiment of the application, after the working terminal displays the dialogue information between the user account and the intelligent customer service, the working terminal acquires first intention information selected by a worker according to the candidate intention information, and sends the first intention information to the manual assistance platform, and correspondingly, the manual assistance platform receives the first intention information.
In one possible embodiment, the first intention information is selected from candidate intention information. In another possible implementation, the first intention information is associated intention information corresponding to the candidate intention information. Illustratively, referring to fig. 5 in combination, in the case that the worker clicks the fourth candidate intention 56, the candidate sub-intentions 58 corresponding to the fourth candidate intention 56 may be presented in the user interface, and further, the worker selects the corresponding candidate sub-intention 58 from the candidate sub-intentions 58 as the first intention information.
And 405, sending the first intention information to the intelligent customer service background.
In the embodiment of the application, after the artificial assistance platform obtains the first intention information, the artificial assistance platform sends the first intention information to the intelligent customer service background. Correspondingly, the smart customer service background receives the first intention information and determines a first reply message associated with the first intention information, wherein the first reply message is used for the smart customer service to reply.
Optionally, in this embodiment of the application, after the artificial intelligence obtains the first intention information, it may further record a session message between the user account and the intelligent customer service and the first intention information, establish an association relationship between the session message between the user account and the intelligent customer service and the first intention information, generate correction information according to the association relationship, and send the correction information to the intelligent training background. The correction information is used for prompting correction of the intelligent customer service background based on the incidence relation, and the intelligent training background is used for training and correcting the intelligent customer service background through the incidence relation. For example, the intention determination model in the intelligent customer service background is trained according to a dialogue message input by a user and first intention information; or adding a new reply message in the reply message library; or adding new intention information in the knowledge graph; and so on.
It should be noted that the training and correction of the intelligent customer service by the intelligent training platform may be performed in real time after the correction information is acquired, or may be performed after a plurality of correction information are acquired at time intervals, which is not limited in the embodiment of the present application. Of course, in practical applications, the intelligent training platform may select the training time interval according to practical situations. Optionally, under the condition that a certain keyword or certain intention information is included in the plurality of correction information, the intelligent training platform determines that the certain keyword or the certain intention information is a topic with higher discussion degree in a special period; or the certain keyword or the certain intention information is a weak link of the intelligent customer service background. At the moment, when the intelligent training platform receives the correction information with the certain keyword and the certain intention information next time, the intelligent customer service background is trained in real time based on the correction information.
Of course, in the embodiment of the present application, the manual assistance platform may also send the dialog message between the user account and the intelligent customer service to the manual service platform when the dialog message input by the user explicitly indicates that manual service is needed; or, the manual assistance platform may also directly send the dialog message to the manual service platform after acquiring the dialog message in the process of a certain interaction for multiple times.
In summary, in the technical solution provided in the embodiment of the present application, through the manual assistance platform, when a session message between a user account and an intelligent customer service meets a condition, a worker assists the intelligent customer service background to determine intent information, so as to ensure accuracy of the intent information determination, and the worker does not need to monitor the session message between the user account and the intelligent customer service in real time, and only needs to provide an assistance function when the intelligent customer service is abnormal, so as to reduce human resource consumption.
In addition, the artificial auxiliary platform screens the second candidate intention information determined by the intelligent customer service background to obtain the candidate intention information, so that the accuracy of the candidate intention information is ensured.
In addition, the intelligent training platform trains the intelligent customer service background according to the correction information, so that the working capacity of the intelligent customer service background is provided, the intelligent customer service background can accurately identify intention information, the interaction between the intelligent customer service background and the manual auxiliary platform is reduced, the information transmission overhead is reduced, the intelligent online customer service user experience is integrally promoted, and meanwhile, the manual correction records are accumulated to help the intelligent training platform to more accurately optimize the intelligent online customer service.
The present application is now fully described with reference to fig. 6.
The intelligent customer service background monitors the conversation messages between the user account and the intelligent customer service and judges whether the conversation messages between the user account and the intelligent customer service meet the conditions or not. If the conversation message between the user account and the intelligent customer service meets the condition, sending the conversation message between the user account and the intelligent customer service to the manual assistance platform; and if the conversation message between the user account and the intelligent customer service does not meet the condition, continuing to monitor the conversation message between the user account and the intelligent customer service. And then, the manual assistance platform judges whether the conversation message between the user account and the intelligent customer service needs to be transferred to the manual service. If the conversation message between the user account and the intelligent customer service needs to be transferred to the manual service, the conversation message between the user account and the intelligent customer service is sent to the manual service platform; and if the conversation message between the user account and the intelligent customer service does not need to be transferred to the manual service, acquiring first intention information corresponding to the conversation message between the user account and the intelligent customer service through the working terminal, and sending the first intention information to the intelligent customer service background. Correspondingly, after the intelligent customer service background acquires the first intention information, acquiring reply information associated with the first intention information, wherein the reply information is information for the intelligent customer service to display to the user.
The following are embodiments of the apparatus of the present application that may be used to perform embodiments of the method of the present application. For details which are not disclosed in the embodiments of the apparatus of the present application, reference is made to the embodiments of the method of the present application.
Referring to fig. 7, a block diagram of a message processing apparatus according to an embodiment of the present application is shown. The device has the function of realizing the message processing method, and the function can be realized by hardware or by hardware executing corresponding software. The device can be a computer device, and can also be arranged in the computer device. The apparatus 700 may include: a message acquisition module 701, a message sending module 702, an intention receiving module 703, and a reply acquisition module 704.
A message acquiring module 701, configured to acquire a session message between a user account and an intelligent customer service;
a message sending module 702, configured to send a session message between the user account and the intelligent customer service to a manual assistance platform if the session message between the user account and the intelligent customer service meets a condition;
an intention receiving module 703, configured to receive first intention information sent by the human assistance platform, where the first intention information is used to indicate a dialog intention corresponding to a dialog message input by the user account;
a reply obtaining module 704, configured to obtain a first reply message associated with the first intention information, where the first reply message is used for the smart customer service to reply.
In an exemplary embodiment, the condition includes a first condition and a second condition; the first condition refers to a judgment condition for the dialog message input by the user account, and the second condition refers to a judgment condition for the dialog message replied by the intelligent customer service; the message sending module 702 includes: the monitoring device comprises a first monitoring unit and a second monitoring unit.
The first monitoring unit is used for monitoring the dialogue information input by the user account; and if the conversation message input by the user account meets the first condition, sending the conversation message between the user account and the intelligent customer service to the manual assistance platform.
The second monitoring unit is used for monitoring the conversation message replied by the intelligent customer service; and if the conversation message replied by the intelligent customer service meets the second condition, sending the conversation message between the user account and the intelligent customer service to the manual assistance platform.
In an exemplary embodiment, the first monitoring unit is configured to perform keyword extraction on a dialog message input by the user account to obtain a corresponding keyword; if the keywords comprise negative emotion keywords, determining that the input conversation message of the user account meets the first condition, wherein the negative emotion keywords are keywords representing discontent emotions; or acquiring intention information corresponding to the dialogue message input by the user account and the accurate probability of the intention information corresponding to the dialogue message input by the user account, wherein the accurate probability is used for indicating the matching degree between the intention information and the dialogue message input by the user account; and if the accuracy probability is smaller than a first threshold value, determining that the dialogue message input by the user account meets the first condition.
In an exemplary embodiment, the second monitoring unit is configured to acquire the number of times that the intelligent customer service replies to the user account in the current interaction; if the reply times are larger than a threshold value, determining that the dialog message replied by the intelligent customer service meets the second condition; or acquiring the number of intention information contained in the dialog message replied by the intelligent customer service; if the number of the intention information is not unique, determining the historical reply times of the dialogue message with the non-unique number of the intention information from the historical dialogue message replied by the intelligent customer service; if the historical reply times are larger than a target value, determining that the dialog message replied by the intelligent customer service meets the second condition; or acquiring the display times of the dialog messages replied by the intelligent customer service, wherein the display times refer to the times of replying the same dialog messages in the interaction by the intelligent customer service; and if the display times are larger than an upper limit value, determining that the dialog message replied by the intelligent customer service meets the second condition.
In an exemplary embodiment, as shown in fig. 8, the apparatus 700 further comprises: a first determination module 705, a relationship establishment module 706, and a second determination module 707.
A first determining module 705, configured to determine second intention information associated with the first intention information according to the first intention information.
A relationship establishing module 706, configured to establish a locking relationship between the user account and the second intention information.
A second determining module 707, configured to, in a case that the first dialog message input by the user account is received, determine intention information corresponding to the first dialog message from within an intention range indicated by the second intention information.
The reply obtaining module 704 is further configured to obtain a reply message associated with the intention information corresponding to the first dialogue message.
In an exemplary embodiment, as shown in fig. 8, the apparatus 700 further comprises: a key acquisition module 708 and a relationship release module 709.
The key obtaining module 708 is configured to, in a case that the first dialog message input by the user account is received, perform keyword extraction on the first dialog message to obtain a keyword in the first dialog message.
A relation releasing module 709, configured to release a locking relation between the user account and the second intention information if the keyword includes a keyword indicating other intention information, where the other intention information is intention information outside the intention range.
The second determining module 707 is further configured to, if a keyword used for indicating other intention information is not included in the keywords, perform the step of determining the intention information corresponding to the first dialogue message from the range of intention indicated by the second intention information.
In an exemplary embodiment, the reply obtaining module 704 is configured to obtain, as candidate reply messages, reply messages associated with the first intention information from a reply message library, where the reply message library includes more than one intention information and reply messages associated with the intention information; acquiring a reply message matched with the user information corresponding to the user account from the candidate reply messages as a first reply message; wherein the user information comprises at least one of: the current geographic position corresponding to the user account, the historical geographic position corresponding to the user account, the user conversation habit corresponding to the user account, the user age corresponding to the user account, and the user gender corresponding to the user account.
In an exemplary embodiment, as shown in fig. 8, the apparatus 700 further comprises: an intent determination module 710.
The intention determining module 710 is configured to perform word segmentation processing and keyword extraction on the second dialog message input by the user account, and determine third intention information corresponding to the second dialog message.
The reply obtaining module 704 is further configured to obtain a second reply message associated with the first reply message if the third intention information is the same as the first intention information, where the second reply message is an interpreted message for the first reply message.
In summary, in the technical scheme provided in the embodiment of the present application, through the manual assistance platform, when the session message between the user account and the intelligent customer service satisfies the condition, the worker assists the intelligent customer service background to determine the intention information, so as to ensure the accuracy of the intention information determination, and the worker does not need to monitor the session message between the user account and the intelligent customer service in real time, and only needs to provide an assistance function when the intelligent customer service is abnormal, so as to reduce the consumption of human resources, avoid monitoring errors caused by human factors, such as fatigue work, and improve the capacity of the large-scale intelligent customer service online service; and after the intention information is determined, the intelligent customer service background determines the reply message according to the intention, workers do not need to lay out manual reply in real time, the timeliness of reply message feedback is ensured, the consumption of human resources is reduced, meanwhile, the resource waste of the intelligent customer service background caused by overlong idle time is reduced, the resource utilization rate of the intelligent customer service background is improved, the interaction between the intelligent customer service and the user account is continuously realized by the intelligent customer service background after the first intention information is determined manually, the purpose of intelligently processing simple problems and manually solving complex problems is kept, the waste of manpower is avoided, and the manual work value is helped to be realized.
Referring to fig. 9, a block diagram of a message processing apparatus according to another embodiment of the present application is shown. The device has the function of realizing the method for acquiring the intention information, and the function can be realized by hardware or hardware executing corresponding software. The device can be computer equipment, and can also be arranged in the computer equipment. The apparatus 900 may include: a message receiving module 901, an information determining module 902, an information transmitting module 903, an intention receiving module 904, and an intention transmitting module 905.
The message receiving module 901 is configured to receive a session message between a user account and an intelligent customer service, where the session message is sent by the intelligent customer service background.
And an information determining module 902, configured to determine candidate intention information corresponding to the dialog message input by the user account according to the dialog message between the user account and the intelligent customer service.
The information sending module 903 is configured to send session information between the user account and the intelligent customer service to a work terminal, where the work terminal is configured to display the session information between the user account and the intelligent customer service, and the session information between the user account and the intelligent customer service includes session information between sessions between the user account and the intelligent customer service and the candidate intention information.
An intention receiving module 904, configured to receive first intention information selected according to the candidate intention information and sent by the work terminal, where the first intention information is used to indicate a conversation intention corresponding to a conversation message input by the user account.
An intention sending module 905, configured to send the first intention information to the smart customer service background, where a first reply message associated with the first intention information is used for the smart customer service to reply.
In an exemplary embodiment, the dialog messages between the user account and the smart customer service include dialog messages entered by the user account; the information determining module 902 is further configured to invoke an intention recognition model, and determine, according to the dialog message input by the user account, first candidate intention information corresponding to the dialog message input by the user account, where the intention recognition model is a model trained based on historical dialog messages and selection results of intention information corresponding to the historical dialog messages; and matching the first candidate intention information with second candidate intention information, and selecting the intention information belonging to the first candidate intention information and the second candidate intention information as candidate intention information, wherein the second candidate intention information refers to intention information corresponding to the candidate dialogue intention of the dialogue information input by the user and determined by the intelligent customer service background.
In an exemplary embodiment, as shown in fig. 10, the apparatus 900 further comprises: a message recording module 906, an association establishing module 907, a revision generating module 908 and a revision sending module 909.
A message recording module 906, configured to record a dialog message between the user account and the smart customer service, and the first intention information.
An association establishing module 907, configured to establish an association relationship between the session message between the user account and the smart customer service and the first intention information.
And a correction generation module 908, configured to generate correction information according to the association relationship, where the correction information is used to prompt that the intelligent customer service background is corrected based on the association relationship.
A correction sending module 909, configured to send the correction information to an intelligent training background, where the intelligent training background is configured to perform training correction on the intelligent customer service background through the association relationship.
In summary, in the technical solution provided in the embodiment of the present application, through the manual assistance platform, when a session message between a user account and an intelligent customer service meets a condition, a worker assists the intelligent customer service background to determine intent information, so as to ensure accuracy of the intent information determination, and the worker does not need to monitor the session message between the user account and the intelligent customer service in real time, and only needs to provide an assistance function when the intelligent customer service is abnormal, so as to reduce human resource consumption.
It should be noted that, when the apparatus provided in the foregoing embodiment implements the functions thereof, only the division of the functional modules is illustrated, and in practical applications, the functions may be distributed by different functional modules according to needs, that is, the internal structure of the apparatus may be divided into different functional modules to implement all or part of the functions described above. In addition, the apparatus and method embodiments provided by the above embodiments belong to the same concept, and specific implementation processes thereof are described in the method embodiments for details, which are not described herein again.
Referring to fig. 11, a block diagram of a computer device according to an embodiment of the present application is shown. The computer equipment can be used for realizing the function of the reply message acquisition method. Specifically, the method comprises the following steps:
the computer apparatus 1100 includes a Central Processing Unit (CPU) 1101, a system Memory 1104 including a Random Access Memory (RAM) 1102 and a Read Only Memory (ROM) 1103, and a system bus 1105 connecting the system Memory 1104 and the Central Processing Unit 1101. The computer device 1100 also includes a basic Input/Output (I/O) system 1106, which facilitates transfer of information between devices within the computer, and a mass storage device 1107 for storing an operating system 1113, application programs 1114, and other program modules 1115.
The basic input/output system 1106 includes a display 1108 for displaying information and an input device 1109 such as a mouse, keyboard, etc. for user input of information. Wherein the display 1108 and the input device 1109 are connected to the central processing unit 1101 through an input output controller 1110 connected to the system bus 1105. The basic input/output system 1106 may also include an input/output controller 1110 for receiving and processing input from a number of other devices, such as a keyboard, mouse, or electronic stylus. Similarly, input-output controller 1110 also provides output to a display screen, a printer, or other type of output device.
The mass storage device 1107 is connected to the central processing unit 1101 through a mass storage controller (not shown) that is connected to the system bus 1105. The mass storage device 1107 and its associated computer-readable media provide non-volatile storage for the computer device 1100. That is, the mass storage device 807 may include a computer readable medium (not shown) such as a hard disk or CD-ROM (Compact disk Read-Only Memory) drive.
Without loss of generality, computer readable media may comprise computer storage media and communication media. Computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes RAM, ROM, EPROM (Erasable Programmable Read Only Memory), EEPROM (Electrically Erasable Programmable Read Only Memory), flash Memory or other solid state Memory technology, CD-ROM, DVD (Digital Video Disc), or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices. Of course, those skilled in the art will appreciate that computer storage media is not limited to the foregoing. The system memory 1104 and mass storage device 1107 described above may be collectively referred to as memory.
According to various embodiments of the present application, the computer device 1100 may also operate as a remote computer connected to a network via a network, such as the Internet. That is, the computer device 1100 may connect to the network 1112 through the network interface unit 1111 that is connected to the system bus 1105, or may connect to other types of networks or remote computer systems (not shown) using the network interface unit 1111.
The memory also includes a computer program stored in the memory and configured to be executed by the one or more processors to implement the message processing method described above.
In an exemplary embodiment, there is also provided a computer readable storage medium having stored therein at least one instruction, at least one program, set of codes, or set of instructions which, when executed by a processor, implement the above-described message processing method.
Optionally, the computer-readable storage medium may include: ROM (Read Only Memory), RAM (Random Access Memory), SSD (Solid State drive), or optical disc. The Random Access Memory may include a ReRAM (resistive Random Access Memory) and a DRAM (Dynamic Random Access Memory).
In an exemplary embodiment, a computer program product or computer program is also provided, the computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the message processing method.
It should be understood that reference to "a plurality" herein means two or more. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship. In addition, the step numbers described herein only exemplarily show one possible execution sequence among the steps, and in some other embodiments, the steps may also be executed out of the numbering sequence, for example, two steps with different numbers are executed simultaneously, or two steps with different numbers are executed in a reverse order to the order shown in the figure, which is not limited by the embodiment of the present application.
The above description is only exemplary of the application and should not be taken as limiting the application, and any modifications, equivalents, improvements and the like that are made within the spirit and principle of the application should be included in the protection scope of the application.
Claims (13)
1. A method of message processing, the method comprising:
acquiring a dialogue message between a user account and an intelligent customer service;
if the conversation message between the user account and the intelligent customer service meets the condition, sending the conversation message between the user account and the intelligent customer service to a manual auxiliary platform; wherein the conditions include at least one of: the dialog messages input by the user account comprise negative emotion keywords, the accuracy probability of intention information corresponding to the dialog messages input by the user account is smaller than a first threshold, the reply times of the intelligent customer service in the current interaction for the user account is larger than a threshold, the historical reply times of the dialog messages with the non-unique intention information quantity contained in the historical dialog messages replied by the intelligent customer service are larger than a target value, the display times are larger than an upper limit value, the negative emotion keywords are keywords representing unsatisfied emotions, the accuracy probability is used for indicating the matching degree between the intention information and the dialog messages input by the user account, and the display times are times of replying the same dialog messages in the current interaction by the intelligent customer service;
receiving first intention information sent by the manual assistance platform, wherein the first intention information is used for indicating a conversation intention corresponding to a conversation message input by the user account, the conversation intention is used for representing a problem needing to be solved by the user account, and the first intention information is determined based on a manual selection result;
determining a first reply message associated with the first intention information, the first reply message for the smart customer service to reply to;
determining second intention information associated with the first intention information according to the first intention information, wherein the second intention information is upper intention information corresponding to the first intention information;
establishing a locking relationship between the user account and the second intention information;
after receiving a first dialogue message input by the user account, extracting keywords of the first dialogue message to obtain keywords in the first dialogue message; the first dialogue message refers to the dialogue message input by the user account after the locking relation is established;
if the keywords in the first dialogue message do not comprise keywords used for indicating other intention information, determining intention information corresponding to the first dialogue message from the range of intention indicated by the second intention information; wherein the other intention information refers to intention information outside the intention range;
determining a reply message associated with intent information corresponding to the first conversation message.
2. The method of claim 1, further comprising:
extracting keywords from the dialogue information input by the user account to obtain corresponding keywords; if the corresponding keywords comprise the negative emotion keywords, determining that the conversation messages between the user account and the intelligent customer service meet the conditions;
alternatively, the first and second electrodes may be,
acquiring intention information corresponding to the dialogue message input by the user account and the accurate probability of the intention information corresponding to the dialogue message input by the user account; and if the accuracy probability is smaller than the first threshold value, determining that the conversation message between the user account and the intelligent customer service meets the condition.
3. The method of claim 1, further comprising:
acquiring the number of times of replying to the user account by the intelligent customer service in the interaction; if the reply times are larger than the threshold value, determining that the dialogue message between the user account and the intelligent customer service meets the condition;
alternatively, the first and second electrodes may be,
acquiring the number of intention information contained in the dialog message replied by the intelligent customer service; if the number of the intention information is not unique, determining the historical reply times of the dialogue message with the non-unique number of the intention information from the historical dialogue message replied by the intelligent customer service; if the historical reply times are larger than the target value, determining that the dialogue information between the user account and the intelligent customer service meets the condition;
alternatively, the first and second electrodes may be,
acquiring the display times of the dialog messages replied by the intelligent customer service; and if the display times are larger than the upper limit value, determining that the conversation message between the user account and the intelligent customer service meets the condition.
4. The method of claim 1, further comprising:
and if the keywords in the first dialogue message comprise the keywords for indicating the other intention information, releasing the locking relationship between the user account and the second intention information.
5. The method of claim 1, wherein the determining the first reply message associated with the first intention information comprises:
acquiring reply messages associated with the first intention information from a reply message library as candidate reply messages, wherein the reply message library comprises more than one intention information and the reply messages associated with the intention information;
acquiring a reply message matched with the user information corresponding to the user account from the candidate reply messages as a first reply message;
wherein the user information comprises at least one of: the current geographic position corresponding to the user account, the historical geographic position corresponding to the user account, the user conversation habit corresponding to the user account, the user age corresponding to the user account, and the user gender corresponding to the user account.
6. The method of any of claims 1 to 5, wherein after determining the first reply message associated with the first intention information, further comprising:
performing word segmentation processing and keyword extraction on a second dialogue message input by the user account, and determining third intention information corresponding to the second dialogue message;
if the third intention information is the same as the first intention information, acquiring a second reply message associated with the first reply message, wherein the second reply message refers to an explanation message aiming at the first reply message.
7. A method of message processing, the method comprising:
receiving a dialogue message between a user account and an intelligent customer service, which is sent by an intelligent customer service background; wherein the dialog message between the user account and the smart customer service is sent by the smart customer service background under the condition that the dialog message satisfies a condition, and the condition includes at least one of the following: the dialog messages input by the user account comprise negative emotion keywords, the accuracy probability of the intention information corresponding to the dialog messages input by the user account is smaller than a first threshold value, the reply times of the intelligent customer service in the current interaction to the user account is larger than a threshold value, the historical reply times of the dialog messages with the non-unique intention information number contained in the historical dialog messages replied by the intelligent customer service are larger than a target value, the display times are larger than an upper limit value, the emotion keywords are keywords representing unsatisfied emotions, the accuracy probability is used for indicating the matching degree between the intention information corresponding to the dialog messages input by the user account and the dialog messages input by the user account, and the display times are the times of replying the same dialog messages in the current interaction by the intelligent customer service;
determining candidate intention information corresponding to the dialogue information input by the user account according to the dialogue information between the user account and the intelligent customer service;
sending the dialogue information between the user account and the intelligent customer service to a working terminal, wherein the working terminal is used for displaying the dialogue information between the user account and the intelligent customer service; wherein the dialogue information between the user account and the intelligent customer service comprises dialogue information between the dialogue between the user account and the intelligent customer service and the candidate intention information;
receiving first intention information which is sent by the working terminal and selected according to the candidate intention information, wherein the first intention information is used for indicating a conversation intention corresponding to a conversation message input by the user account, the conversation intention is used for representing a problem needing to be solved by the user account, and the first intention information is determined based on a manual selection result;
sending the first intention information to the smart customer service background, wherein a first reply message associated with the first intention information is used for the smart customer service to reply; after receiving the first intention information, the intelligent customer service background determines second intention information associated with the first intention information according to the first intention information, wherein the second intention information is upper intention information corresponding to the first intention information; the intelligent customer service background establishes a locking relation between the user account and the second intention information; after receiving a first dialogue message input by the user account, the intelligent customer service background extracts keywords from the first dialogue message to obtain keywords in the first dialogue message; the first dialogue message refers to the dialogue message input by the user account after the locking relation is established; if the keywords in the first dialogue message do not comprise keywords used for indicating other intention information, the intelligent customer service background determines intention information corresponding to the first dialogue message from the intention range indicated by the second intention information; wherein the other intention information refers to intention information outside the intention range; the smart customer service backend determines a reply message associated with intent information corresponding to the first conversation message.
8. The method of claim 7, wherein the determining candidate intention information corresponding to the dialog message input by the user account according to the dialog message between the user account and the smart customer service comprises:
calling an intention recognition model, and determining first candidate intention information corresponding to the dialogue information input by the user account according to the dialogue information input by the user account, wherein the intention recognition model is trained on the basis of the historical dialogue information and an artificial selection result of the intention information corresponding to the historical dialogue information;
and matching the first candidate intention information with second candidate intention information, and selecting the intention information belonging to the first candidate intention information and the second candidate intention information as candidate intention information, wherein the second candidate intention information is intention information corresponding to candidate dialogue intentions of dialogue messages input by the user account determined by the intelligent customer service background.
9. The method according to claim 7 or 8, characterized in that the method further comprises:
recording a dialogue message between the user account and the intelligent customer service and the first intention information;
establishing an association relation between a dialogue message between the user account and the intelligent customer service and the first intention information;
generating correction information according to the incidence relation, wherein the correction information is used for prompting that the intelligent customer service background is corrected based on the incidence relation;
and sending the correction information to an intelligent training background, wherein the intelligent training background is used for training and correcting the intelligent customer service background through the incidence relation.
10. A message processing apparatus, characterized in that the apparatus comprises:
the message acquisition module is used for acquiring a dialogue message between a user account and the intelligent customer service;
the message sending module is used for sending the conversation message between the user account and the intelligent customer service to a manual auxiliary platform if the conversation message between the user account and the intelligent customer service meets the condition; wherein the conditions include at least one of: the dialog messages input by the user account comprise negative emotion keywords, the accuracy probability of intention information corresponding to the dialog messages input by the user account is smaller than a first threshold, the reply times of the intelligent customer service in the current interaction for the user account is larger than a threshold, the historical reply times of the dialog messages with the non-unique intention information quantity contained in the historical dialog messages replied by the intelligent customer service are larger than a target value, the display times are larger than an upper limit value, the negative emotion keywords are keywords representing unsatisfied emotions, the accuracy probability is used for indicating the matching degree between the intention information and the dialog messages input by the user account, and the display times are times of replying the same dialog messages in the current interaction by the intelligent customer service;
the intention receiving module is used for receiving first intention information sent by the manual assistance platform, the first intention information is used for indicating a conversation intention corresponding to a conversation message input by the user account, the conversation intention is used for representing a problem needing to be solved by the user account, and the first intention information is determined based on a manual selection result;
a reply determination module for determining a first reply message associated with the first intention information, the first reply message for the smart customer service to reply to;
a first determination module, configured to determine, according to the first intention information, second intention information associated with the first intention information, where the second intention information is upper intention information corresponding to the first intention information;
the relation establishing module is used for establishing a locking relation between the user account and the second intention information;
the key acquisition module is used for extracting keywords of the first dialogue message after receiving the first dialogue message input by the user account to obtain the keywords in the first dialogue message; the first dialogue message refers to the dialogue message input by the user account after the locking relation is established;
a second determining module, configured to determine intention information corresponding to the first dialogue message from an intention range indicated by the second intention information if a keyword in the first dialogue message does not include a keyword used for indicating other intention information; wherein the other intention information refers to intention information outside the intention range;
the reply acquisition module is further used for determining a reply message associated with the intention information corresponding to the first dialogue message.
11. A message processing apparatus, characterized in that the apparatus comprises:
the message receiving module is used for receiving a conversation message between the user account and the intelligent customer service, which is sent by the intelligent customer service background; wherein the dialog message is sent by the smart customer service background under the condition that the dialog message satisfies a condition, and the condition comprises at least one of the following: the dialog messages input by the user account comprise negative emotion keywords, the accuracy probability of intention information corresponding to the dialog messages input by the user account is smaller than a first threshold, the reply times of the intelligent customer service in the current interaction to the user account is larger than a threshold, the historical reply times of the dialog messages with the non-unique intention information quantity contained in the historical dialog messages replied by the intelligent customer service are larger than a target value, the display times are larger than an upper limit value, the negative emotion keywords are keywords representing unsatisfied emotions, the accuracy probability is used for indicating the matching degree between the intention information corresponding to the dialog messages input by the user account and the dialog messages input by the user account, and the display times are times of replying the same dialog messages by the intelligent customer service in the current interaction;
the information determining module is used for determining candidate intention information corresponding to the dialogue information input by the user account according to the dialogue information between the user account and the intelligent customer service;
the information sending module is used for sending the dialogue information between the user account and the intelligent customer service to a working terminal, the working terminal is used for displaying the dialogue information between the user account and the intelligent customer service, and the dialogue information between the user account and the intelligent customer service comprises dialogue information between the user account and the intelligent customer service and the candidate intention information;
an intention receiving module, configured to receive first intention information selected according to the candidate intention information and sent by the work terminal, where the first intention information is used to indicate a dialog intention corresponding to a dialog message input by the user account, the dialog intention is used to characterize a problem that needs to be solved by the user account, and the first intention information is determined based on a manual selection result;
the intention sending module is used for sending the first intention information to the intelligent customer service background, wherein a first reply message associated with the first intention information is used for the intelligent customer service to reply; after receiving the first intention information, the intelligent customer service background determines second intention information associated with the first intention information according to the first intention information, wherein the second intention information is upper intention information corresponding to the first intention information; the intelligent customer service background establishes a locking relation between the user account and the second intention information; after receiving a first dialogue message input by the user account, the intelligent customer service background extracts keywords from the first dialogue message to obtain keywords in the first dialogue message; the first dialogue message refers to the dialogue message input by the user account after the locking relation is established; if the keywords in the first dialogue message do not comprise the keywords used for indicating other intention information, the intelligent customer service background determines the intention information corresponding to the first dialogue message from the intention range indicated by the second intention information; wherein the other intention information refers to intention information outside the intention range; the smart customer service backend determines a reply message associated with intent information corresponding to the first conversation message.
12. A computer device comprising a processor and a memory, the memory having stored therein at least one program which is loaded and executed by the processor to implement the message processing method of any of claims 1 to 6 or to implement the message processing method of any of claims 7 to 9.
13. A computer-readable storage medium, in which at least one program is stored, which is loaded and executed by a processor to implement the message processing method according to any one of claims 1 to 6, or to implement the message processing method according to any one of claims 7 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010766697.9A CN111917878B (en) | 2020-08-03 | 2020-08-03 | Message processing method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010766697.9A CN111917878B (en) | 2020-08-03 | 2020-08-03 | Message processing method, device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111917878A CN111917878A (en) | 2020-11-10 |
| CN111917878B true CN111917878B (en) | 2023-01-10 |
Family
ID=73286952
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010766697.9A Active CN111917878B (en) | 2020-08-03 | 2020-08-03 | Message processing method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111917878B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112364149B (en) * | 2021-01-12 | 2021-04-23 | 广州云趣信息科技有限公司 | User question obtaining method and device and electronic equipment |
| CN113065362B (en) * | 2021-03-19 | 2024-06-21 | 联想(北京)有限公司 | Intelligent reply method and device |
| CN113590785A (en) * | 2021-07-29 | 2021-11-02 | 百度在线网络技术（北京）有限公司 | Message reply method and device, electronic equipment and storage medium |
| CN114117157B (en) * | 2021-11-19 | 2024-04-09 | 招联消费金融股份有限公司 | Session processing method, apparatus, computer device and storage medium |
| CN114860912B (en) * | 2022-05-20 | 2023-08-29 | 马上消费金融股份有限公司 | Data processing method, device, electronic equipment and storage medium |
| CN115016707A (en) * | 2022-05-31 | 2022-09-06 | 北京达佳互联信息技术有限公司 | Information processing method, information processing device, electronic equipment and storage medium |
| CN115208846B (en) * | 2022-07-08 | 2023-06-09 | 武汉联影医疗科技有限公司 | Session interaction method and system |
| CN116029740A (en) * | 2023-02-09 | 2023-04-28 | 湖南快乐阳光互动娱乐传媒有限公司 | Information processing method and device |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106354835A (en) * | 2016-08-31 | 2017-01-25 | 上海交通大学 | Artificial dialogue auxiliary system based on context semantic understanding |
| CN109104406B (en) * | 2018-06-28 | 2020-09-11 | 腾讯科技（深圳）有限公司 | Online customer service management method and device, server and storage medium |
| CN111477231B (en) * | 2019-01-24 | 2023-12-01 | 科沃斯商用机器人有限公司 | Man-machine interaction method, device and storage medium |
| CN110020873B (en) * | 2019-03-12 | 2024-06-21 | 平安科技（深圳）有限公司 | Manual customer service switching method, related device, equipment and computer readable medium |
| CN111444323A (en) * | 2020-02-25 | 2020-07-24 | 平安国际智慧城市科技股份有限公司 | Accident information rapid acquisition method and device, computer equipment and storage medium |
| CN113079263B (en) * | 2021-03-16 | 2023-05-02 | 京东科技控股股份有限公司 | Method, device, system and medium for intelligent customer service switching |
-
2020
- 2020-08-03 CN CN202010766697.9A patent/CN111917878B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN111917878A (en) | 2020-11-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111917878B (en) | Message processing method, device, equipment and storage medium | |
| CN108073600B (en) | Intelligent question-answer interaction method and device and electronic equipment | |
| AU2020201883B2 (en) | Call center system having reduced communication latency | |
| CN109514586B (en) | Method and system for realizing intelligent customer service robot | |
| CN111932144B (en) | Customer service agent distribution method and device, server and storage medium | |
| CN108829682B (en) | Computer readable storage medium, intelligent question answering method and intelligent question answering device | |
| CN112153426B (en) | Content account management method and device, computer equipment and storage medium | |
| CN112035638B (en) | Information processing method, device, storage medium and equipment | |
| CN110321273A (en) | A kind of business statistical method and device | |
| CN110543550B (en) | Method and device for automatically generating test questions | |
| JP6442807B1 (en) | Dialog server, dialog method and dialog program | |
| JP2023158593A (en) | Automatic question answering system and question-answer pair data generation method | |
| CN115470318A (en) | Customer service problem processing method and device | |
| CN107832342B (en) | Robot chatting method and system | |
| CN115221892A (en) | Work order data processing method and device, storage medium and electronic equipment | |
| CN111414609B (en) | Object verification method and device | |
| CN113868396A (en) | Task intelligent dialogue construction method and system based on knowledge graph | |
| CN111008373A (en) | Intelligent question and answer processing method and device, computer readable medium and electronic equipment | |
| CN110929005A (en) | Emotion analysis-based task follow-up method, device, equipment and storage medium | |
| CN113449503A (en) | Session type information management method, device, equipment and storage medium thereof | |
| CN117874211B (en) | Intelligent question-answering method, system, medium and electronic equipment based on SAAS software | |
| CN112966077B (en) | Method, device and equipment for determining conversation state and storage medium | |
| CN112291329B (en) | Information display method, device and equipment | |
| CN116011599B (en) | Intelligent reservation method and system for biological laboratory | |
| EP4089592A1 (en) | Method for determining annotation capability information, related apparatus and computer program product |
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 |