CN112819245A - User complaint behavior prediction method, device, equipment and storage medium - Google Patents

User complaint behavior prediction method, device, equipment and storage medium Download PDF

Info

Publication number
CN112819245A
CN112819245A CN202110208102.2A CN202110208102A CN112819245A CN 112819245 A CN112819245 A CN 112819245A CN 202110208102 A CN202110208102 A CN 202110208102A CN 112819245 A CN112819245 A CN 112819245A
Authority
CN
China
Prior art keywords
call
user
data
historical
pieces
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.)
Pending
Application number
CN202110208102.2A
Other languages
Chinese (zh)
Inventor
刘斌彬
纪诚
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sunshine Insurance Group Co Ltd
Original Assignee
Sunshine Insurance Group Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Sunshine Insurance Group Co Ltd filed Critical Sunshine Insurance Group Co Ltd
Priority to CN202110208102.2A priority Critical patent/CN112819245A/en
Publication of CN112819245A publication Critical patent/CN112819245A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/332—Query formulation
    • G06F16/3329—Natural language query formulation or dialogue systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/3331—Query processing
    • G06F16/334—Query execution
    • G06F16/3343—Query execution using phonetics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33—Querying
    • G06F16/3331—Query processing
    • G06F16/334—Query execution
    • G06F16/3344—Query execution using natural language analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/35—Clustering; Classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/01—Customer relationship services
    • G06Q30/015—Providing customer assistance, e.g. assisting a customer within a business location or via helpdesk
    • G06Q30/016—After-sales
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/08—Insurance

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Computational Linguistics (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Development Economics (AREA)
  • Marketing (AREA)
  • Artificial Intelligence (AREA)
  • Human Resources & Organizations (AREA)
  • Mathematical Physics (AREA)
  • Operations Research (AREA)
  • Game Theory and Decision Science (AREA)
  • Human Computer Interaction (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Technology Law (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Health & Medical Sciences (AREA)
  • Acoustics & Sound (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • General Health & Medical Sciences (AREA)
  • Telephonic Communication Services (AREA)

Abstract

The application provides a method, a device, equipment and a storage medium for predicting user complaint behaviors, and relates to the technical field of artificial intelligence. The method comprises the following steps: acquiring a plurality of pieces of historical call data of a user; each piece of historical call data respectively comprises: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call; according to the incoming call time, sequencing a plurality of pieces of historical call data to obtain user sequencing data; and predicting the intention of the user for the next call based on the serialized data of the user and by adopting a complaint behavior prediction model obtained by pre-training, wherein the intention of the user for the next call is used for indicating whether the complaint behavior can occur to the user. In the scheme, the plurality of pieces of historical call data of the user are characterized into the sequence data, so that the content characteristics and the time sequence characteristics of each call of the user are reserved, the information loss in the serialization process is reduced, and the accuracy of the prediction of the complaint behaviors of the user is improved.

Description

User complaint behavior prediction method, device, equipment and storage medium
Technical Field
The application relates to the technical field of artificial intelligence, in particular to a method, a device, equipment and a storage medium for predicting user complaint behaviors.
Background
With the intelligent development of customer service systems, by using intelligent technologies such as Automatic Speech Recognition (ASR for short), Natural Language Processing (NLP for short) and the like, when a customer makes a call, the customer carries out semantic understanding on the call Speech of the customer, knows the intention of the call, finds the risk of customer complaints in time, and then accurately navigates the customer to the corresponding customer service staff or self-service process to pertinently process the customer complaints, thereby achieving the effect of improving the service quality.
At present, a customer complaint behavior prediction model is mainly constructed by adopting a logistic regression or random forest model. Firstly, in a logistic regression model, data of different dimensions need to be standardized in the same data space, and complaint data of customers have the characteristic of sparseness, so that a sparse characteristic matrix is easily formed, and the problem of under-fitting of the constructed model is caused; secondly, through a model constructed by random forests, the influence of discrete variables and sparse characteristics on results can be effectively reduced, but certain information loss exists in a time sequence.
Therefore, the methods in the prior art have the problem that the accuracy of the prediction of the customer complaint behaviors is low.
Disclosure of Invention
The objective of the present application is to provide a method, an apparatus, a device, and a storage medium for predicting a user complaint behavior, so as to reduce information loss of the complaint data of the user in the serialization process, and achieve the purpose of improving the accuracy of the prediction of the user complaint behavior.
In order to achieve the above purpose, the technical solutions adopted in the embodiments of the present application are as follows:
in a first aspect, an embodiment of the present application provides a method for predicting a complaint behavior of a user, including:
acquiring a plurality of pieces of historical call data of a user; wherein each piece of the historical call data respectively comprises: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call;
according to the incoming call time, carrying out serialization processing on the plurality of pieces of historical call data to obtain serialization data of the user;
and predicting to obtain the next call intention of the user by adopting a complaint behavior prediction model obtained by pre-training based on the serialized data of the user, wherein the next call intention of the user is used for indicating whether the user can have complaint behaviors.
Optionally, the serializing the multiple pieces of historical call data according to the incoming call time to obtain the serialized data of the user includes:
respectively connecting all the historical call data in series into a serialization subsequence;
and according to the incoming call time, all the serialization subsequences are connected in series to form the serialization data of the user.
Optionally, the concatenating each piece of historical call data into one serialized subsequence includes:
and sequentially connecting the call intention, the call scene, the call duration and the time interval of the next adjacent call in each piece of historical call data in series to obtain the serialization subsequence.
Optionally, before the concatenating the historical call data into a serialized sub-sequence, the method further includes:
and according to the value distribution, carrying out box separation processing on the call duration in the plurality of pieces of historical call data and the time interval between the call duration and the next adjacent call.
Optionally, the predicting the next call intention of the user by using a complaint behavior prediction model obtained by pre-training based on the serialized data of the user includes:
constructing a feature vector corresponding to the serialized data;
and inputting the feature vector into the complaint behavior prediction model, and predicting to obtain the next call intention of the user.
Optionally, the acquiring multiple pieces of historical call data of the user includes:
acquiring a plurality of historical call voices of the user;
respectively converting each historical call voice into each text data;
and carrying out natural language processing on the text data to obtain a plurality of pieces of historical call data of the user.
Optionally, before the obtaining of the plurality of pieces of historical call data of the user; the method comprises the following steps:
acquiring a plurality of pieces of historical call data of a user, wherein each piece of historical call data respectively comprises: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call;
according to the incoming call time, carrying out serialization processing on the plurality of pieces of historical call data to obtain serialization data of the user;
and training an obtained complaint behavior prediction model by taking the serialized data of the user as a training sample, wherein the complaint behavior prediction model is used for predicting whether the user can have complaint behaviors.
In a second aspect, an embodiment of the present application further provides a device for predicting a complaint behavior of a user, where the device includes: the device comprises an acquisition module, a processing module and a prediction module;
the acquisition module is used for acquiring a plurality of pieces of historical call data of the user; wherein each piece of the historical call data respectively comprises: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call;
the processing module is used for carrying out serialization processing on the plurality of pieces of historical call data according to the incoming call time to obtain serialization data of the user;
the prediction module is used for predicting and obtaining the next conversation intention of the user by adopting a complaint behavior prediction model obtained by pre-training based on the serialized data of the user, wherein the next conversation intention of the user is used for indicating whether the user can have complaint behaviors.
Optionally, the processing module is further configured to:
respectively connecting all the historical call data in series into a serialization subsequence;
and according to the incoming call time, all the serialization subsequences are connected in series to form the serialization data of the user.
Optionally, the processing module is further configured to:
and sequentially connecting the call intention, the call scene, the call duration and the time interval of the next adjacent call in each piece of historical call data in series to obtain the serialization subsequence.
Optionally, the processing module is further configured to:
and according to the value distribution, carrying out box separation processing on the call duration in the plurality of pieces of historical call data and the time interval between the call duration and the next adjacent call.
Optionally, the prediction module is further configured to:
constructing a feature vector corresponding to the serialized data;
and inputting the feature vector into the complaint behavior prediction model, and predicting to obtain the next call intention of the user.
Optionally, the obtaining module is further configured to:
acquiring a plurality of historical call voices of the user;
respectively converting each historical call voice into each text data;
and carrying out natural language processing on the text data to obtain a plurality of pieces of historical call data of the user.
Optionally, the apparatus further comprises a training module; the training module is configured to:
acquiring a plurality of pieces of historical call data of a user, wherein each piece of historical call data respectively comprises: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call;
according to the incoming call time, carrying out serialization processing on the plurality of pieces of historical call data to obtain serialization data of the user;
and training an obtained complaint behavior prediction model by taking the serialized data of the user as a training sample, wherein the complaint behavior prediction model is used for predicting whether the user can have complaint behaviors.
In a third aspect, an embodiment of the present application further provides an electronic device, including: a processor, a storage medium and a bus, the storage medium storing machine-readable instructions executable by the processor, the processor and the storage medium communicating via the bus when the electronic device is operating, the processor executing the machine-readable instructions to perform the steps of the method as provided by the first aspect.
In a fourth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor to perform the steps of the method as provided in the first aspect.
The beneficial effect of this application is:
the application provides a method, a device, equipment and a storage medium for predicting user complaint behaviors, wherein the method comprises the following steps: acquiring a plurality of pieces of historical call data of a user; each piece of historical call data respectively comprises: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call; according to the incoming call time, sequencing a plurality of pieces of historical call data to obtain user sequencing data; and predicting the intention of the user for the next call based on the serialized data of the user and by adopting a complaint behavior prediction model obtained by pre-training, wherein the intention of the user for the next call is used for indicating whether the complaint behavior can occur to the user. In the scheme, the plurality of pieces of historical call data of the user are characterized into the sequence data, so that the content characteristics and the time sequence characteristics of each call of the user are reserved, the information loss in the serialization process is reduced, and the accuracy of the prediction of the complaint behaviors of the user is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Fig. 2 is a schematic flowchart of a method for predicting a complaint behavior of a user according to an embodiment of the present disclosure;
fig. 3 is a schematic flowchart of another method for predicting a complaint behavior of a user according to an embodiment of the present disclosure;
fig. 4 is a schematic flowchart of another method for predicting a complaint behavior of a user according to an embodiment of the present application;
fig. 5 is a schematic flowchart of another method for predicting a complaint behavior of a user according to an embodiment of the present application;
fig. 6 is a schematic flowchart of another method for predicting a complaint behavior of a user according to an embodiment of the present application;
fig. 7 is a schematic structural diagram of a device for predicting user complaint behavior according to an embodiment of the present application.
Detailed Description
In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it should be understood that the drawings in the present application are for illustrative and descriptive purposes only and are not used to limit the scope of protection of the present application. Additionally, it should be understood that the schematic drawings are not necessarily drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of the present application. It should be understood that the operations of the flow diagrams may be performed out of order, and steps without logical context may be performed in reverse order or simultaneously. One skilled in the art, under the guidance of this application, may add one or more other operations to, or remove one or more operations from, the flowchart.
In addition, the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present application without making any creative effort, shall fall within the protection scope of the present application.
In order to enable a person skilled in the art to make use of the present disclosure, the following embodiments are given in conjunction with a specific application scenario. It will be apparent to those skilled in the art that the general principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the application.
It should be noted that in the embodiments of the present application, the term "comprising" is used to indicate the presence of the features stated hereinafter, but does not exclude the addition of further features.
Before explaining the embodiments of the present application in detail, an application scenario of the present application will be described first. The application scene can be specifically applied to a plurality of fields such as insurance industry, logistics industry, internet industry and the like, and the application is not limited to the fields.
The method and the device are mainly based on historical call data of the user and a customer service system in the past, whether complaint behaviors can occur when the user calls next time is predicted, complaint risks of the user can be predicted in advance before the user initiates complaints, so that the user can accurately navigate the user to corresponding customer service personnel when the user calls next time, and then the complaints of the user are specifically processed, the service quality of enterprises is improved, and meanwhile the experience feeling of the user is also improved.
Fig. 1 is a schematic structural diagram of an electronic device according to an embodiment of the present disclosure; as shown in fig. 1, the electronic device 100 may be an electronic device with a processing function, such as a computer, a server, or the like, for implementing the method for predicting the complaint behavior of the user provided by the present application.
As shown in fig. 1, the electronic device 100 includes: memory 101, processor 102.
The memory 101 and the processor 102 are electrically connected directly or indirectly to realize data transmission or interaction. For example, electrical connections may be made through one or more communication buses or signal lines.
The Memory 101 may be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Read-Only Memory (EPROM), an electrically Erasable Read-Only Memory (EEPROM), and the like.
The processor 102 may be an integrated circuit chip having signal processing capability. The Processor 102 may be a general-purpose Processor including a Central Processing Unit (CPU), a Network Processor (NP), and the like. The various methods, steps and logic blocks disclosed in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory 101 is used for storing programs, and the processor 102 calls the programs stored in the memory 101 to execute the method for predicting the complaint behavior of the user provided in the following embodiment.
It will be appreciated that the configuration shown in FIG. 1 is merely illustrative and that electronic device 100 may include more or fewer modules than shown in FIG. 1 or have a different configuration than shown in FIG. 1. The modules shown in fig. 1 may be implemented in hardware, software, or a combination thereof.
The following describes the implementation principle and the corresponding beneficial effects of the method for predicting the complaint behavior of the user according to the present application with a plurality of specific embodiments.
Fig. 2 is a schematic flowchart of a method for predicting a complaint behavior of a user according to an embodiment of the present disclosure; the execution subject of the method may be the electronic device, as shown in fig. 2, the method includes:
s201, acquiring a plurality of pieces of historical call data of the user.
Wherein, each historical conversation data respectively includes: incoming call time, call intention, call scenario, call duration, time interval to next adjacent call.
The call-in time refers to the time when the user a calls into the customer service system each time.
For example, the time of the first call to the customer service system is 2020/7/158: 15:50, the call intention is roadside assistance, the call scene is manual, the call duration is 15s, and the time interval from the next adjacent call is 180 s.
In the embodiment, whether complaints will occur when the user makes a next call is predicted mainly based on past historical call data of the user and the customer service system.
For example, if it is required to predict whether complaints will occur when the user a calls the customer service system at the 4 th time, 3 pieces of historical call data of the previous 3 times calls of the user a may be acquired, for example, the acquired 3 pieces of historical call data respectively include: the number of the user A, the call-in time, the call intention, the call scene, the call duration and the time interval between the next adjacent call. The client number of the user A is used as a unique identifier for identifying different users.
The specific formula can be shown in the following table 1:
TABLE 1 multiple historical call data for incoming call by user A
Figure BDA0002949998180000081
S202, according to the incoming call time, sequencing the multiple pieces of historical call data to obtain the user sequencing data.
For example, 3 pieces of history call data of the user a may be serialized according to the time sequence of each incoming call of the user a in table 1, for example, the serialization data of the user: the call data 1- > the call data 2- > the call data 3, and the history call data of each item respectively includes: intention- > scene- > call duration- > time interval of next adjacent call to obtain serialized data of 3 pieces of history call data of previous calls of the user a.
After the serialization processing, a plurality of pieces of historical call data of the user A are characterized into the sequence data with the indefinite length, so that the content characteristics of each call of the user A can be reserved, the time sequence among different calls can be embodied, the information loss in the serialization process is reduced, and the accuracy of the prediction of the complaint behaviors of the user can be improved.
And S203, predicting to obtain the next call intention of the user by adopting a complaint behavior prediction model obtained by pre-training based on the serialized data of the user.
Wherein, the next call intention of the user is used for indicating whether the complaint behavior of the user can occur.
The complaint behavior prediction model is a Neural Network model obtained by training based on past multiple pieces of historical call data of multiple users, and can be a Recurrent Neural Network (RNN) model, for example.
For example, on the basis of the above embodiment, serialized data of 3 pieces of historical call data incoming before the user a is input to a complaint behavior prediction model obtained through pre-training, a classification label predicting the call intention of the user a when the user a calls the customer service system for the 4 th time is obtained, that is, a complaint behavior occurs or does not occur, and if the complaint behavior is predicted to occur when the user a calls the customer service system for the 4 th time, the complaint can be early warned in advance and relevant managers and customer service staff can be reminded, so as to improve the service quality.
To sum up, the embodiment of the present application provides a method for predicting a complaint behavior of a user, including: acquiring a plurality of pieces of historical call data of a user; wherein, each historical conversation data respectively includes: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call; according to the incoming call time, sequencing a plurality of pieces of historical call data to obtain user sequencing data; and predicting the intention of the user for the next call based on the serialized data of the user and by adopting a complaint behavior prediction model obtained by pre-training, wherein the intention of the user for the next call is used for indicating whether the complaint behavior can occur to the user. In the scheme, the plurality of pieces of historical call data of the user A are characterized into the sequence data with the indefinite length, so that the content characteristics and the time sequence characteristics of each call of the user can be reserved, the information loss in the serialization process is reduced, and the accuracy of the prediction of the complaint behaviors of the user is effectively improved.
Fig. 3 is a schematic flowchart of another method for predicting a complaint behavior of a user according to an embodiment of the present disclosure; as shown in fig. 3, the above step S202: according to the incoming call time, sequencing a plurality of pieces of historical call data to obtain the user sequencing data, and the method comprises the following steps:
s301, all the historical call data are respectively connected in series to form a serialization subsequence.
Optionally, the call intention, the call scene, the call duration, and the time interval between the next call and the next call in each piece of historical call data may be concatenated in sequence to obtain a serialized subsequence.
For example, the 3 pieces of historical call data shown in table 1 may be respectively concatenated into a plurality of serialized subsequences, for example, the 1 st subsequence of historical call data is: rescue-artificial-15 s-200s, as for the subsequence of the 2 nd historical call data: the 3 rd sub-sequence of the historical call data is as follows: claims query-self-service-20 s-180 s. Similarly, if the 4 th historical call data of the user a is obtained, a subsequence of the 4 th historical call data can be obtained, so that a plurality of pieces of historical call data of the user a can be serialized into sequence data of indefinite length, and information loss in the serialization process is effectively avoided.
And S302, according to the incoming call time, all the serialization subsequences are connected in series to form the serialization data of the user.
For example, according to the incoming call time in the 3 pieces of historical call data shown in table 1, the obtained 3 subsequences are concatenated into the serialized data of the user, which is specifically as follows:
the method comprises the following steps of road rescue, manual operation, 15s-200s, vehicle insurance declaration, manual operation, 20s-100s, claim settlement query, self-service, 20s-180 s.
In this embodiment, according to the variation characteristic of the incoming call time in the plurality of pieces of historical call data, the plurality of subsequences of the plurality of pieces of historical call data are subjected to linear serialization processing, and the processing characteristic of the complaint behavior prediction model on the linear time sequence is utilized to better fit the time sequence data, so that the accuracy of the prediction on the call intention of the user at the next time is improved.
Alternatively, in the step S301: before each piece of historical call data is respectively connected in series into a serialization subsequence, the method further comprises the following steps: and according to the value distribution, carrying out box separation processing on the call duration in the plurality of pieces of historical call data and the time interval between the call duration and the next adjacent call.
Wherein, the box separation treatment comprises: equidistant binning and isocratic binning. In the present embodiment, the equal-distance binning process is adopted, which may also be called equal-width binning, and if K spaces are provided, the pitch I of each space is (Max-Min)/K.
For example, the distance between the call durations of the user a shown in table 1 may be set to 10, and the split space may be: the classification of 10s or less is 1, the classification of 11-20 is 2, and the classification of 21-30 is 3.
Setting the interval between the time intervals of the next adjacent calls as 100, the split space obtained by the division may be: the classification is 1 in 100s or less, 2 in 101 to 200, and 3 in 201 to 300.
The interval setting of the divided box-dividing spaces can be redefined according to the actual situation, and is not particularly limited herein.
Thus, the "call duration" and the "time interval between the call and the next adjacent call" in the 3 pieces of historical call data shown in table 1 can be subjected to binning processing, and the obtained 3 subsequences can be respectively: road rescue-manual-2-2, vehicle insurance declaration-manual-2-1 and claim settlement inquiry-self-service-2-2.
In addition, the obtained 3 subsequences are concatenated into the serialized data of the user a, which is specifically as follows: the method comprises the following steps of road rescue, manual operation, 2-2-vehicle insurance declaration, manual operation, 2-1-claim settlement inquiry, self-service and 2-2.
In this embodiment, the call duration and the time interval of the next adjacent call are subjected to binning discretization, so that the efficiency of predicting the call intention of the user at the next time can be improved.
Fig. 4 is a schematic flowchart of another method for predicting a complaint behavior of a user according to an embodiment of the present application; as shown in fig. 4, the step S203: based on the serialized data of the user, predicting and obtaining the next call intention of the user by adopting a complaint behavior prediction model obtained by pre-training, wherein the method comprises the following steps:
s401, constructing a feature vector corresponding to the serialized data.
The feature vector may be included in a plurality, and the plurality of feature vectors may form a multi-dimensional feature vector matrix.
For example, the serialized data for user a may be: "road rescue-manual work-15 s-200 s-vehicle insurance application-manual work-20 s-100 s-claim settlement query-self-help-20 s-180 s", generate a 3-row and 4-column multidimensional eigenvector matrix, which is specifically as follows:
Figure BDA0002949998180000111
and S402, inputting the feature vector into the complaint behavior prediction model, and predicting to obtain the next call intention of the user.
In this embodiment, for example, the multidimensional feature vector matrix generated according to the serialized data of the user a may be used as an input feature of the complaint behavior prediction model, and the prediction result of the call intention of the user a on the 4 th incoming client system is output through the processing of the complaint behavior prediction model, for example, it is predicted that the complaint behavior of the user a on the 4 th incoming client system occurs.
Fig. 5 is a schematic flowchart of another method for predicting a complaint behavior of a user according to an embodiment of the present application; as shown in fig. 5, the step S201 of acquiring a plurality of pieces of historical call data of the user includes:
s501, obtaining a plurality of historical call voices of the user.
For example, when the user a calls in the customer service system of an insurance company or a logistics company, the user a can start to acquire the call voice data between the user a and the customer service system.
And S502, respectively converting each historical call voice into each text data.
Illustratively, when a user a needs a case information query service, after calling in a customer service system, and according to a language prompt message in the customer service system, entering a manual service platform, and connecting with a customer service 1, when the user a says "query a current progress situation of a case" to the customer service 1, and after the customer service 1 reports the progress situation of the case to the user a, the user a says "the progress is too slow".
In the whole process of the call between the user a and the customer service 1, the call voice data of the user a "query the current progress of a certain case" and "the progress is too slow" can be translated into corresponding text data by using an Automatic Speech Recognition technology (ASR for short).
S503, natural language processing is carried out on each text data to obtain a plurality of pieces of historical call data of the user.
For example, in addition to the above embodiments, the text data obtained by translation may be classified and identified by using a Natural Language Processing algorithm (NLP) model, and information such as a call intention of the user a calling into the customer service system and a scene of handling a service may be accurately identified, so that information such as a call intention and a call scene in each piece of history call data of the user a may be obtained.
Fig. 6 is a schematic flowchart of another method for predicting a complaint behavior of a user according to an embodiment of the present application; as shown in fig. 6, before acquiring a plurality of pieces of historical call data of the user in step S201, the method includes:
s601, acquiring a plurality of pieces of historical call data of the user.
Wherein, each historical conversation data respectively includes: incoming call time, call intention, call scenario, call duration, time interval to next adjacent call.
For example, 3 pieces of historical call data of the previous 3 calls of the user a may be acquired, where acquiring the 3 pieces of historical call data of the user a includes: incoming call time, call intention, call scenario, call duration, time interval to next adjacent call. Specific examples are shown in table 1 above.
S602, according to the incoming call time, sequencing a plurality of pieces of historical call data to obtain the user sequencing data.
In this embodiment, 3 pieces of historical call data shown in table 1 may be serialized according to the call-in time of each time the user a calls into the customer service system in table 1, so as to obtain the serialized data of the user a, specifically: the call data 1- > the call data 2- > the call data 3, and the history call data of each item respectively includes: intention- > scene- > call duration- > time interval of next adjacent call, and the like, so that 3 pieces of historical call data of the user A are connected into a data chain by using a serialization processing method, call content and time sequence information in the 3 pieces of historical call data of the user A can be fully mined, content characteristics and time sequence characteristics of each call are kept, the purpose of reducing information loss in the process of carrying out characteristic quantization processing on the call data is achieved, and the prediction accuracy of a complaint behavior prediction model obtained through training can be improved.
And S603, training to obtain a complaint behavior prediction model by taking the serialized data of the user as a training sample.
The complaint behavior prediction model is used for predicting whether complaint behaviors can occur to the user.
On the basis of the above embodiment, in order to obtain the complaint behavior prediction model based on the serialized data of the user a, if the RNN model is used as the initial complaint behavior prediction model, the serialized data of the user a is used as the training sample, and is input into the initial complaint behavior prediction model, and after the initial complaint behavior prediction model is trained for multiple times, the two classification models for identifying complaints/non-complaints are obtained, so that the trained complaint behavior prediction model is obtained.
Therefore, the probability of the complaint behavior of the user when the user calls the customer service system next time can be predicted by the obtained historical call data of the user and the trained complaint behavior prediction model, and the user opinion can be effectively processed.
The following describes a device, a storage medium, and the like corresponding to the method for predicting the user complaint behavior provided by the present application, and specific implementation processes and technical effects thereof are referred to above, and will not be described again below.
Fig. 7 is a schematic structural diagram of a device for predicting a complaint behavior of a user according to an embodiment of the present application; as shown in fig. 7, the apparatus includes: an obtaining module 701, a processing module 702, and a predicting module 703;
an obtaining module 701, configured to obtain multiple pieces of historical call data of a user; wherein, each historical conversation data respectively includes: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call;
the processing module 702 is configured to perform serialization processing on multiple pieces of historical call data according to the incoming call time to obtain serialization data of the user;
the prediction module 703 is configured to predict, based on the serialized data of the user, a next call intention of the user by using a complaint behavior prediction model obtained through pre-training, where the next call intention of the user is used to indicate whether the user will have a complaint behavior.
Optionally, the processing module 702 is further configured to:
respectively connecting all the historical call data in series into a serialization subsequence;
and according to the incoming call time, all the serialization subsequences are concatenated into the serialization data of the user.
Optionally, the processing module 702 is further configured to:
and sequentially connecting the call intention, the call scene, the call duration and the time interval of the next adjacent call in each piece of historical call data in series to obtain a serialized subsequence.
Optionally, the processing module 702 is further configured to:
and according to the value distribution, carrying out box separation processing on the call duration in the plurality of pieces of historical call data and the time interval between the call duration and the next adjacent call.
Optionally, the prediction module 703 is further configured to:
constructing a feature vector corresponding to the serialized data;
and inputting the feature vector into a complaint behavior prediction model, and predicting to obtain the next call intention of the user.
Optionally, the obtaining module 701 is further configured to:
acquiring a plurality of historical call voices of a user;
respectively converting each historical call voice into each text data;
and performing natural language processing on each text data to obtain a plurality of pieces of historical call data of the user.
Optionally, the apparatus further comprises: a training module; the training module is used for:
acquiring a plurality of pieces of historical call data of a user, wherein each piece of historical call data respectively comprises: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call;
according to the incoming call time, sequencing a plurality of pieces of historical call data to obtain user sequencing data;
and training the obtained complaint behavior prediction model by taking the serialized data of the user as a training sample, wherein the complaint behavior prediction model is used for predicting whether the user can have complaint behaviors.
The above-mentioned apparatus is used for executing the method provided by the foregoing embodiment, and the implementation principle and technical effect are similar, which are not described herein again.
These above modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or one or more microprocessors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs), among others. For another example, when one of the above modules is implemented in the form of a Processing element scheduler code, the Processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).
Optionally, the invention also provides a program product, for example a computer-readable storage medium, comprising a program which, when being executed by a processor, is adapted to carry out the above-mentioned method embodiments.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, a division of a unit is merely a logical division, and an actual implementation may have another division, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.
The integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute some steps of the methods according to the embodiments of the present invention. And the aforementioned storage medium includes: a U disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

Claims (10)

1. A method for predicting complaint behaviors of a user is characterized by comprising the following steps:
acquiring a plurality of pieces of historical call data of a user; wherein each piece of the historical call data respectively comprises: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call;
according to the incoming call time, carrying out serialization processing on the plurality of pieces of historical call data to obtain serialization data of the user;
and predicting to obtain the next call intention of the user by adopting a complaint behavior prediction model obtained by pre-training based on the serialized data of the user, wherein the next call intention of the user is used for indicating whether the user can have complaint behaviors.
2. The method of claim 1, wherein the serializing the plurality of pieces of historical call data according to the incoming call time to obtain the serialized data of the user comprises:
respectively connecting all the historical call data in series into a serialization subsequence;
and according to the incoming call time, all the serialization subsequences are connected in series to form the serialization data of the user.
3. The method of claim 2, wherein the concatenating the historical call data into a serialized subsequence comprises:
and sequentially connecting the call intention, the call scene, the call duration and the time interval of the next adjacent call in each piece of historical call data in series to obtain the serialization subsequence.
4. The method of claim 2, wherein before the step of concatenating the historical call data into a serialized sub-sequence, the step of:
and according to the value distribution, carrying out box separation processing on the call duration in the plurality of pieces of historical call data and the time interval between the call duration and the next adjacent call.
5. The method according to any one of claims 1 to 4, wherein the predicting the next call intention of the user by using a pre-trained complaint behavior prediction model based on the serialized data of the user comprises:
constructing a feature vector corresponding to the serialized data;
and inputting the feature vector into the complaint behavior prediction model, and predicting to obtain the next call intention of the user.
6. The method according to any one of claims 1-4, wherein the obtaining a plurality of pieces of historical call data of the user comprises:
acquiring a plurality of historical call voices of the user;
respectively converting each historical call voice into each text data;
and carrying out natural language processing on the text data to obtain a plurality of pieces of historical call data of the user.
7. The method according to any one of claims 1-4, wherein before said obtaining a plurality of pieces of historical call data of a user; the method comprises the following steps:
acquiring a plurality of pieces of historical call data of a user, wherein each piece of historical call data respectively comprises: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call;
according to the incoming call time, carrying out serialization processing on the plurality of pieces of historical call data to obtain serialization data of the user;
and training an obtained complaint behavior prediction model by taking the serialized data of the user as a training sample, wherein the complaint behavior prediction model is used for predicting whether the user can have complaint behaviors.
8. A user complaint behavior prediction apparatus, characterized in that the apparatus comprises: the device comprises an acquisition module, a processing module and a prediction module;
the acquisition module is used for acquiring a plurality of pieces of historical call data of the user; wherein each piece of the historical call data respectively comprises: incoming call time, call intention, call scene, call duration, and time interval to the next adjacent call;
the processing module is used for carrying out serialization processing on the plurality of pieces of historical call data according to the incoming call time to obtain serialization data of the user;
the prediction module is used for predicting and obtaining the next conversation intention of the user by adopting a complaint behavior prediction model obtained by pre-training based on the serialized data of the user, wherein the next conversation intention of the user is used for indicating whether the user can have complaint behaviors.
9. An electronic device, comprising: a processor, a storage medium and a bus, the storage medium storing machine-readable instructions executable by the processor, the processor and the storage medium communicating via the bus when the electronic device is operating, the processor executing the machine-readable instructions to perform the steps of the method according to any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that the storage medium has stored thereon a computer program which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.
CN202110208102.2A 2021-02-24 2021-02-24 User complaint behavior prediction method, device, equipment and storage medium Pending CN112819245A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110208102.2A CN112819245A (en) 2021-02-24 2021-02-24 User complaint behavior prediction method, device, equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110208102.2A CN112819245A (en) 2021-02-24 2021-02-24 User complaint behavior prediction method, device, equipment and storage medium

Publications (1)

Publication Number Publication Date
CN112819245A true CN112819245A (en) 2021-05-18

Family

ID=75865463

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110208102.2A Pending CN112819245A (en) 2021-02-24 2021-02-24 User complaint behavior prediction method, device, equipment and storage medium

Country Status (1)

Country Link
CN (1) CN112819245A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113313316A (en) * 2021-06-11 2021-08-27 北京明略昭辉科技有限公司 Method and device for outputting prediction data, storage medium and electronic equipment
CN113569021A (en) * 2021-06-29 2021-10-29 杭州摸象大数据科技有限公司 Method for user classification, computer device and readable storage medium
CN113610399A (en) * 2021-08-09 2021-11-05 广州品唯软件有限公司 Risk monitoring method, system and device for customer service background
CN113673905A (en) * 2021-08-31 2021-11-19 广东省信息网络有限公司 Complaint service early warning monitoring system based on big data

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109559172A (en) * 2018-11-29 2019-04-02 北京车联天下信息技术有限公司 Data processing method, device, electronic equipment and computer readable storage medium
CN109858702A (en) * 2019-02-14 2019-06-07 中国联合网络通信集团有限公司 Client upgrades prediction technique, device, equipment and the readable storage medium storing program for executing complained
CN110889526A (en) * 2018-09-07 2020-03-17 中国移动通信集团有限公司 Method and system for predicting user upgrade complaint behavior
CN111797318A (en) * 2020-07-01 2020-10-20 喜大（上海）网络科技有限公司 Information recommendation method, device, equipment and storage medium
CN112257458A (en) * 2020-10-21 2021-01-22 阳光保险集团股份有限公司 Intention recognition model training method, intention recognition method, device and equipment

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110889526A (en) * 2018-09-07 2020-03-17 中国移动通信集团有限公司 Method and system for predicting user upgrade complaint behavior
CN109559172A (en) * 2018-11-29 2019-04-02 北京车联天下信息技术有限公司 Data processing method, device, electronic equipment and computer readable storage medium
CN109858702A (en) * 2019-02-14 2019-06-07 中国联合网络通信集团有限公司 Client upgrades prediction technique, device, equipment and the readable storage medium storing program for executing complained
CN111797318A (en) * 2020-07-01 2020-10-20 喜大（上海）网络科技有限公司 Information recommendation method, device, equipment and storage medium
CN112257458A (en) * 2020-10-21 2021-01-22 阳光保险集团股份有限公司 Intention recognition model training method, intention recognition method, device and equipment

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113313316A (en) * 2021-06-11 2021-08-27 北京明略昭辉科技有限公司 Method and device for outputting prediction data, storage medium and electronic equipment
CN113569021A (en) * 2021-06-29 2021-10-29 杭州摸象大数据科技有限公司 Method for user classification, computer device and readable storage medium
CN113569021B (en) * 2021-06-29 2023-08-04 杭州摸象大数据科技有限公司 Method for classifying users, computer device and readable storage medium
CN113610399A (en) * 2021-08-09 2021-11-05 广州品唯软件有限公司 Risk monitoring method, system and device for customer service background
CN113673905A (en) * 2021-08-31 2021-11-19 广东省信息网络有限公司 Complaint service early warning monitoring system based on big data

Similar Documents

Publication Publication Date Title
CN112819245A (en) User complaint behavior prediction method, device, equipment and storage medium
US10777207B2 (en) Method and apparatus for verifying information
US11392775B2 (en) Semantic recognition method, electronic device, and computer-readable storage medium
EP3851999A1 (en) Method and apparatus for processing information
CN113590776B (en) Knowledge graph-based text processing method and device, electronic equipment and medium
CN113268610B (en) Intent jump method, device, equipment and storage medium based on knowledge graph
CN110825827B (en) Entity relationship recognition model training method and device and entity relationship recognition method and device
CN110555451A (en) information identification method and device
CN113657100A (en) Entity identification method and device, electronic equipment and storage medium
CN112307337A (en) Association recommendation method and device based on label knowledge graph and computer equipment
CN110807311A (en) Method and apparatus for generating information
CN111414732A (en) Text style conversion method and device, electronic equipment and storage medium
CN111159378B (en) Method and device for classifying problem description information
CN113609865A (en) Text emotion recognition method and device, electronic equipment and readable storage medium
CN115146653B (en) Dialogue scenario construction method, device, equipment and storage medium
CN113420140B (en) User emotion prediction method and device, electronic equipment and readable storage medium
CN113268575B (en) Entity relationship identification method and device and readable medium
CN115221892A (en) Work order data processing method and device, storage medium and electronic equipment
CN115510219A (en) Method and device for recommending dialogs, electronic equipment and storage medium
CN110795424B (en) Characteristic engineering variable data request processing method and device and electronic equipment
CN114792086A (en) Information extraction method, device, equipment and medium supporting text cross coverage
CN113344405B (en) Method, device, equipment, medium and product for generating information based on knowledge graph
CN112784034B (en) Digest generation method and device and computer equipment
CN112989797B (en) Model training and text expansion methods, devices, equipment and storage medium
CN113705206B (en) Emotion prediction model training method, device, equipment and storage medium

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20210518

RJ01 Rejection of invention patent application after publication