CN112131369A - Service class determination method and device - Google Patents

Service class determination method and device Download PDF

Info

Publication number
CN112131369A
CN112131369A CN202011049699.2A CN202011049699A CN112131369A CN 112131369 A CN112131369 A CN 112131369A CN 202011049699 A CN202011049699 A CN 202011049699A CN 112131369 A CN112131369 A CN 112131369A
Authority
CN
China
Prior art keywords
teller
service
probability
business
category
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.)
Granted
Application number
CN202011049699.2A
Other languages
Chinese (zh)
Other versions
CN112131369B (en
Inventor
黄文强
黄雅楠
浮晨琪
李蚌蚌
徐晨敏
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.)
Bank of China Ltd
Original Assignee
Bank of China 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 Bank of China Ltd filed Critical Bank of China Ltd
Priority to CN202011049699.2A priority Critical patent/CN112131369B/en
Publication of CN112131369A publication Critical patent/CN112131369A/en
Application granted granted Critical
Publication of CN112131369B publication Critical patent/CN112131369B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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
    • G06F16/35—Clustering; Classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/12—Computing arrangements based on biological models using genetic models
    • G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N7/00—Computing arrangements based on specific mathematical models
    • G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
    • 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
    • G06Q40/02—Banking, e.g. interest calculation or account maintenance
    • 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
    • G10L15/26—Speech to text systems
    • 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
    • G10L17/00—Speaker identification or verification techniques
    • G10L17/02—Preprocessing operations, e.g. segment selection; Pattern representation or modelling, e.g. based on linear discriminant analysis [LDA] or principal components; Feature selection or extraction
    • 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
    • G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
    • G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
    • G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Physics (AREA)
  • Acoustics & Sound (AREA)
  • Artificial Intelligence (AREA)
  • Business, Economics & Management (AREA)
  • Multimedia (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Biophysics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computing Systems (AREA)
  • Finance (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Software Systems (AREA)
  • Evolutionary Biology (AREA)
  • Databases & Information Systems (AREA)
  • Accounting & Taxation (AREA)
  • Physiology (AREA)
  • Signal Processing (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Mathematical Analysis (AREA)
  • Algebra (AREA)
  • Probability & Statistics with Applications (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Biomedical Technology (AREA)
  • Genetics & Genomics (AREA)
  • Computational Mathematics (AREA)
  • Development Economics (AREA)

Abstract

The embodiment of the application provides a method and a device for determining service classes, wherein the method comprises the following steps: the method comprises the steps of obtaining target voice information corresponding to each service in at least one service of a teller, obtaining first target voice characteristic information of the teller according to the target voice information of the teller, determining the probability of the first target voice characteristic information corresponding to at least one service category, converting the target voice information into text information, matching the text information with preset answer text information to obtain a matching degree, inputting the probability and the matching degree of the first target voice characteristic information corresponding to at least one service category into a first Bayesian model to obtain the service category probability of the teller, and determining the service category of the teller according to the service category probability of the teller. The method and the device for determining the business category enable a bank to accurately determine the business handling category of a teller, and enable the teller to better know the business category.

Description

Service class determination method and device
Technical Field
The present application relates to the field of computer technologies, and in particular, to a method and an apparatus for determining a service class.
Background
At present, with the rapid development of computer network technology, the business handling of banks is more and more intelligent. The teller, as a primary service person for the bank to handle the business for the customer, needs to have a qualified business handling capability, that is, the bank needs to make standardized approval on the teller's business level. At present, the business class of the teller is judged artificially and subjectively, and the classification result of the business class of the teller is not fair and accurate.
Therefore, how to accurately determine the business category of the teller is a specific problem faced by the bank.
Disclosure of Invention
The embodiment of the application provides a method and a device for determining business categories, so that a bank can accurately determine the business handling categories of tellers, and the business categories of the tellers can be better known.
The embodiment of the application provides a service class determination method, which comprises the following steps:
acquiring target voice information corresponding to each service in at least one service of a teller;
respectively executing the following processing aiming at the target voice information corresponding to each service to obtain the service category of the teller corresponding to each service:
obtaining first target sound characteristic information of the teller according to the target voice information of the teller, wherein the first target sound characteristic information embodies the sound characteristic of the teller;
determining a probability that the first target sound characteristic information corresponds to at least one traffic class;
converting the target voice information into text information, wherein the text information comprises an answer sentence of the teller for replying to a customer question;
matching the text information with preset answer text information to obtain a matching degree, wherein the preset answer text information comprises preset answer sentences, and the matching degree reflects the matching degree of the answer sentences of the teller and the preset answer sentences;
inputting the probability of the first target sound characteristic information corresponding to at least one service class and the matching degree into a first Bayesian model to obtain the service class probability of the teller;
and determining the business category of the teller according to the business category probability of the teller.
Optionally, the method further includes:
determining the performance corresponding to each business according to the business category of each business of the teller and the corresponding relation between the business category and the performance of the teller;
and determining the total performance of the teller according to the performance corresponding to each service.
Optionally, the method further includes:
acquiring second target sound characteristic information of a client corresponding to each service in at least one service, wherein the second target sound characteristic information embodies the sound characteristics of the client;
classifying the second target sound characteristic information according to a preset interval, and determining an interval probability corresponding to the second target sound characteristic information; the interval probability is the probability of the interval in which the second target sound characteristic information is located;
respectively inputting the interval probabilities corresponding to the second target sound characteristic information into a second Bayesian model to obtain emotion category probabilities corresponding to the second target sound characteristic information; the emotion category probability is the probability that the second target sound characteristic information corresponds to an emotion category, and the emotion category reflects the emotion of the client;
determining the emotion category of the client corresponding to each service in the at least one service according to the emotion category probability of the second target sound characteristic information of the client corresponding to each service in the at least one service;
inputting a business category corresponding to each business in at least one business of the teller and an emotion category of a client corresponding to each business in the at least one business into a genetic algorithm model to obtain performance corresponding to each business in the at least one business; wherein the genetic algorithm model is obtained by pre-training;
and determining the total performance of the teller according to the performance corresponding to each service in the at least one service.
Specifically, the determining the service class of the teller according to the service class probability of the teller includes:
and determining the teller with the highest business class probability as the business class of the teller.
Specifically, the sound characteristics include one or more of the following:
audio features, pitch features, volume features, and speech rate features.
An embodiment of the present application further provides a device for determining a service class, where the device includes:
the teller terminal comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring target voice information corresponding to each service in at least one service of a teller;
the processing unit is used for respectively executing the following processing aiming at the target voice information corresponding to each service to obtain the service type of the teller corresponding to each service;
the processing unit includes:
the system comprises a characteristic acquisition unit, a voice recognition unit and a voice recognition unit, wherein the characteristic acquisition unit is used for acquiring first target voice characteristic information of a teller according to target voice information of the teller, and the first target voice characteristic information embodies voice characteristics of the teller;
a first determining unit, configured to determine a probability that the first target sound characteristic information corresponds to at least one traffic class;
the conversion unit is used for converting the target voice information into text information, and the text information comprises an answer sentence of the customer question replied by the teller;
the matching unit is used for matching the text information with preset answer text information to obtain a matching degree, wherein the preset answer text information comprises preset answer sentences, and the matching degree reflects the matching degree of the answer sentences of the teller and the preset answer sentences;
the first input unit is used for inputting the probability of the first target sound characteristic information corresponding to at least one business category and the matching degree into a first Bayesian model to obtain the business category probability of the teller;
and the second determining unit is used for determining the business category of the teller according to the business category probability of the teller.
Optionally, the apparatus further comprises:
the third determining unit is used for determining the performance corresponding to each business according to the business category of each business of the teller and the corresponding relation between the business category and the performance of the teller;
and the fourth determining unit is used for determining the total performance of the teller according to the performance corresponding to each service.
Optionally, the apparatus further comprises:
a second obtaining unit, configured to obtain second target sound feature information of a client corresponding to each service in at least one service, where the second target sound feature information represents a sound feature of the client;
a fifth determining unit, configured to classify the second target sound feature information according to a preset interval, and determine an interval probability corresponding to the second target sound feature information; the interval probability is the probability of the interval in which the second target sound characteristic information is located;
the second input unit is used for respectively inputting the interval probabilities corresponding to the second target sound characteristic information into a second Bayesian model to obtain emotion category probabilities corresponding to the second target sound characteristic information; the emotion category probability is the probability that the second target sound characteristic information corresponds to an emotion category, and the emotion category reflects the emotion of the client;
a sixth determining unit, configured to determine, according to an emotion category probability of second target sound feature information of a client corresponding to each service in the at least one service, an emotion category of the client corresponding to each service in the at least one service;
a third input unit, configured to input a service category corresponding to each service in at least one service of the teller and an emotion category of a client corresponding to each service in the at least one service into a genetic algorithm model, so as to obtain a performance corresponding to each service in the at least one service; wherein the genetic algorithm model is obtained by pre-training;
and the seventh input unit is used for determining the total performance of the teller according to the performance corresponding to each service in the at least one service.
Specifically, the determining, by the second determining unit, the service class of the teller according to the service class probability of the teller includes:
the second determination unit determines the teller as the business class with the highest probability of the business class.
Specifically, the sound characteristics include one or more of the following:
audio features, pitch features, volume features, and speech rate features.
Compared with the prior art, the invention has at least the following advantages:
the embodiment of the application provides an emotion classification determination method, which comprises the following steps: acquiring target voice information corresponding to each service in at least one service of a teller; respectively executing the following processing aiming at the target voice information corresponding to each service to obtain the service category of the teller corresponding to each service: obtaining first target sound characteristic information of the teller according to the target voice information of the teller, wherein the first target sound characteristic information embodies the sound characteristic of the teller; determining a probability that the first target sound characteristic information corresponds to at least one traffic class; converting the target voice information into text information, wherein the text information comprises an answer sentence of the teller for replying to a customer question; matching the text information with preset answer text information to obtain a matching degree, wherein the preset answer text information comprises preset answer sentences, and the matching degree reflects the matching degree of the answer sentences of the teller and the preset answer sentences; inputting the probability of the first target sound characteristic information corresponding to at least one service class and the matching degree into a first Bayesian model to obtain the service class probability of the teller; and determining the business category of the teller according to the business category probability of the teller. Therefore, the method and the device have the advantages that the voice information of each service teller is obtained to obtain the voice characteristic information of each service teller and the probability of the service category corresponding to the voice characteristic information, the voice information is converted into the text information to obtain the matching degree of the teller, the probability and the matching degree of the service category corresponding to the voice characteristic information are input into the Bayesian model to obtain the probability of the service category of each service of the teller, the service category of the teller is determined according to the probability of the service category of the teller, the bank can accurately determine the service handling category of the teller, the service capability of the teller can be better known, and corresponding training is conducted on the service category of the teller to improve the service level of the teller.
Drawings
In order to more clearly illustrate the embodiments of the present application, the drawings needed to be used in the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only embodiments of the present application, and it is obvious to those skilled in the art that other drawings can be obtained from the provided drawings without inventive effort.
Fig. 1 is a schematic flowchart of a service class determination method according to an embodiment of the present application;
fig. 2 is a schematic flowchart of a service class determination method according to a second embodiment of the present application;
fig. 3 is a schematic flowchart of a service class determination method according to a third embodiment of the present application;
fig. 4 is a schematic structural diagram of a service class determination device according to an embodiment of the present application.
Detailed Description
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 is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
As known from the background art, the prior bank teller needs to have qualified business handling capacity, so the bank needs to carry out standardized identification on the business class of the teller, but the current business class for the bank teller is artificially and subjectively judged, and the classification result of the business class of the teller is not fair and accurate. How to accurately determine the transaction category of the teller is a specific problem faced by banks.
Therefore, in the embodiment of the application, the voice information of the teller of each service is obtained to obtain the sound characteristic information of the teller and the probability of the service class corresponding to the sound characteristic information, the voice information is converted into the text information to obtain the matching degree of the teller, the probability and the matching degree of the service class corresponding to the sound characteristic information are input into the Bayesian model to obtain the probability of the service class of each service of the teller, and the service class of the teller is determined according to the probability of the service class of the teller, so that a bank can accurately determine the service handling class of the teller, and the service capability of the teller can be better known.
As shown in fig. 1, a schematic flow chart of a service class determination method provided in an embodiment of the present application is shown, where the method includes the following steps:
step 101: and acquiring target voice information corresponding to each service in at least one service of the teller.
In the embodiment of the application, the target voice information is the conversation voice when the teller handles the business for the client, and the performance of the teller in handling the business can be reflected. The teller transacted business comprises one or more businesses.
In the specific implementation process of step 101, the voice information of the teller may be recorded by the sound recording device, and the voice information of the teller is obtained by obtaining the dialogue voice in the sound recording device.
Step 102: and respectively executing the following processing aiming at the target voice information corresponding to each service to obtain the service category of the teller corresponding to each service.
In the embodiment of the application, the voice information of each service can be processed, and finally the service type of the service handled by the teller is obtained.
In the process of implementing step 102, the following steps may be included:
step A1: and obtaining first target sound characteristic information of the teller according to the target voice information of the teller, wherein the first target sound characteristic information embodies the sound characteristic of the teller.
In the embodiment of the application, the target sound characteristic information represents the sound characteristic of the teller during a certain business transaction. The method comprises the steps that a bank teller handles business for a client, at least one business is included, each business is provided with a voice conversation between the teller and the client, the voice conversation is voice information, the voice information of the teller can be recorded through a recording device, and target voice characteristic information of the teller, namely voice characteristics of the teller, is obtained through a voice analysis model. Optionally, the sound characteristics include one or more of: audio features, pitch features, volume features, and speech rate features. For example, the teller may be able to make a volume of over 80 db while transacting a particular transaction for a customer. In another example, the teller transacts another transaction for the customer at a volume of 50 to 60 decibels and a speech rate of 80 words per minute.
Step B1: determining a probability that the first target sound characteristic information corresponds to at least one traffic class.
In the embodiment of the application, the sound characteristic information of the teller can be classified according to the business class, that is, the sound characteristic of the teller can be classified according to the business class. Specifically, in the volume characteristics of the teller, classification may be performed according to the business category. The business class represents the attributes of the teller, and the business classes are different when the teller presents different attributes. The service type may be 6 types, optionally, the service type may be a sixth type, the representative teller attribute is a higher service capability and a higher service processing speed, and may also indicate that the teller has a mild emotion in the service processing process and provides a higher level of service for the client, the service type may also be a first type, the representative teller attribute is a lower service capability and provides a lower speed of service processing, and may also indicate that the teller may have emotional fluctuation in the service processing process and provide a lower level of service for the client. According to the above, in the volume characteristics of the teller, the teller may be classified according to 6 service classes, for example, the teller volume is classified into the first class of the service classes in the interval of 80 db to 100 db, that is, the teller volume is between 80 db to 100 db, and the teller is considered to have a lower service class for the transaction.
After the sound characteristic information of the teller is classified according to the business class, the probability of the business class corresponding to the sound characteristic information of the teller, namely the probability of the business class corresponding to the sound characteristic of the teller can be determined. Specifically, the volume characteristic of the teller may be between 80 db and 100 db, the corresponding service class is the first class, and the probability of the service class may be 0.1, that is, the probability of the service class corresponding to the volume characteristic of the teller may be 0.1.
It should be noted that, in the embodiment of the present application, the probability of the business class corresponding to the sound feature information of the teller is obtained through the historical sound data of the teller, that is, the training sound feature information of the teller:
in the embodiment of the application, the historical sound data of the teller, namely the training sound characteristic information of the teller, can be obtained by the teller through the teller historical service processing of the voice information recorded by the recording device. The training voice characteristic information of the teller can reflect the voice characteristics of the teller and can also reflect the business class of the teller.
In the embodiment of the application, the training sound characteristic information of the teller is classified according to the preset interval corresponding to the business class. Specifically, the classification method is generally the same as the classification method of the target sound feature information. For example, for the volume characteristics, the service categories are 2 categories, which are respectively the first category and the second category, and then the volume characteristics are also divided into 2 intervals, which may be the first interval when the volume is between 40 db and 70 db, and the second interval corresponding to the service categories when the volume is between 0 db and 40 db or above 70 db, and the first category corresponding to the service categories.
After the training voice characteristic information of the teller is classified according to the preset interval corresponding to the business class, counting the times of the training voice characteristic information of the teller in the preset interval corresponding to a certain business class and the times of the total sample data of the training voice characteristic information of the teller, wherein the ratio of the times of the training voice characteristic information of the teller in the preset interval corresponding to a certain business class to the times of the total sample data of the training voice characteristic information of the teller is the probability of the voice characteristic information of the teller corresponding to a certain business class. Specifically, the training volume characteristics of the teller may be classified according to business categories, for example, the training volume of the teller is classified into a first category of the business categories from 80 db to 100 db, the number of times that the training volume of the teller is in this interval is 3 times, the total sample data of the training volume characteristics of the teller is 100 times, and the probability of the first category corresponding to the volume characteristics is 0.03.
Step C1: and converting the target voice information into text information, wherein the text information comprises an answer sentence of the teller replying to the customer question.
In the embodiment of the application, the voice conversation of the teller, namely the voice information, acquired by the recording device is converted into the text information, wherein the text information is mainly the business communication information between the teller and the client, and the text information comprises the answer sentence of the teller replying the business handling related question proposed by the client.
Step D1: and matching the text information with preset answer text information to obtain a matching degree, wherein the preset answer text information comprises preset answer sentences, and the matching degree reflects the matching degree of the answer sentences of the teller and the preset answer sentences.
In the embodiment of the application, the preset answer text information is a standard answer template aiming at relevant questions which are possibly proposed by a client and related to business handling, and the questions of the client to the business questions can be solved. The preset answer text information comprises preset answer sentences aiming at the business questions. The method has the advantages that the text information of the teller can be matched with the preset answer text information, namely, the answer sentence replied by the teller to a customer question is matched with the preset answer sentence to obtain the matching degree, the matching degree refers to the matching degree of the answer sentence of the teller and the preset answer sentence, the matching degree can reflect the specialty of teller service handling, the higher the matching degree is, the more professional the teller service handling is, and the stronger the service handling capacity is. For example, the business category can be divided into a first category and a second category, and if the matching degree is 0.6 to 1, the business category can be considered as the first category, which represents that the business handling of the teller has strong professional.
Step E1: and inputting the probability of the first target sound characteristic information corresponding to at least one service class and the matching degree into a first Bayesian model to obtain the service class probability of the teller.
In the embodiment of the present application, the bayesian model, i.e., the bayesian formula, is a mathematical formula for calculating the probability. The service class probability is the probability of the service class corresponding to the sound feature information and the matching degree of the teller, optionally, the service class probability may be the probability that the volume feature and the speed feature of the teller are both the first class of the service class, the service class probability may also be the probability that the tone feature, the audio feature and the matching degree of the teller are all the sixth class of the service class, and the service class probability may also be the probability that the volume feature, the speed feature, the tone feature, the audio feature and the matching degree of the teller are all the second class of the service class. The teller service category can reflect accurate classification of the teller service capacity, specifically, the teller service category can be 6 categories, optionally, the teller service category can be a sixth category, the category to which the teller service capacity belongs is stronger, the service handling speed is higher, the teller service category can also be shown to be mild in mood and stronger in professional in the service handling process, the teller service category can also be a first category, the category to which the teller service capacity belongs is weaker, the service handling speed is lower, the teller service category can also be shown to be possibly subjected to mood fluctuation in the service handling process, and the professional performance is weaker.
In the process of implementing step 103 specifically, the probability corresponding to the sound feature of the teller classified according to the service class and the probability of classifying the matching degree according to the service class can be calculated as the service class probability of the teller through a bayesian formula. Specifically, for the volume feature, the probability corresponding to the first service category is 0.1, for the speech rate feature, the probability corresponding to the first service category is 0.2, for the matching degree, the probability corresponding to the first category is 0.3, and the three probabilities are input into a bayesian formula to obtain the probabilities of the first service category relative to the volume feature, the speech rate feature and the matching degree, that is, the probability of the first service category.
It should be noted that the matching degree is classified according to the service category and the probability is calculated, which can also be obtained through the historical matching degree data of the teller:
in the embodiment of the application, the historical text information of the teller, namely training text information, can be obtained through the voice information recorded by the recording equipment in the historical transaction service of the teller, and the training text information is matched with the preset answer text information to obtain the matching degree. The matching degree of the teller can reflect the professional of the teller service and can also reflect the service capability of the teller.
In the embodiment of the application, the matching degree of the teller can be classified according to the preset interval corresponding to the business category. Specifically, the classification method is generally the same as the classification method of the training voice feature information of the teller.
After the matching degrees of the tellers are classified according to the preset intervals, counting the times that the matching degrees of the tellers are in a certain preset interval and the total sample data times of the matching degrees of the tellers, wherein the ratio of the times that the matching degrees of the tellers are in the certain preset interval to the total sample data times of the matching degrees of the tellers is the probability of a certain service class corresponding to the matching degrees. Specifically, the matching degree of the teller may be classified according to the business category, for example, the section where the matching degree of the teller is between 0.6 and 1 is classified as a first category of the business category, the number of times that the matching degree of the teller is in this section is 3 times, the total sample data of the matching degree of the teller is 100 times, and the section probability is 0.03, that is, the probability of the matching degree with respect to the first category of the business category is 0.03.
Step F1: and determining the business category of the teller according to the business category probability of the teller.
In the embodiment of the application, the business class probability of the teller can reflect the business capability of the teller, specifically, the business class of the teller may be 3 classes, the first business class probability is 0.5, the second business class probability is 0.3, and the third business class probability is 0.2. The service class of the teller can be determined by comparing the sizes of the service class probabilities of the three classes, specifically, the maximum service class probability can be determined as the service class of the teller, and the first service class is 0.5 and is the maximum value, so that the service class of the teller can be determined as the first service class.
According to the service type determining method provided by the embodiment of the application, target voice information corresponding to each service in at least one service of a teller is obtained; respectively executing the following processing aiming at the target voice information corresponding to each service to obtain the service category of the teller corresponding to each service: obtaining first target sound characteristic information of the teller according to the target voice information of the teller, wherein the first target sound characteristic information embodies the sound characteristic of the teller; determining a probability that the first target sound characteristic information corresponds to at least one traffic class; converting the target voice information into text information, wherein the text information comprises an answer sentence of the teller for replying to a customer question; matching the text information with preset answer text information to obtain a matching degree, wherein the preset answer text information comprises preset answer sentences, and the matching degree reflects the matching degree of the answer sentences of the teller and the preset answer sentences; inputting the probability of the first target sound characteristic information corresponding to at least one service class and the matching degree into a first Bayesian model to obtain the service class probability of the teller; and determining the business category of the teller according to the business category probability of the teller. Therefore, the method and the device have the advantages that the voice information of each service teller is obtained to obtain the voice characteristic information of each service teller and the probability of the service category corresponding to the voice characteristic information, the voice information is converted into the text information to obtain the matching degree of the teller, the probability and the matching degree of the service category corresponding to the voice characteristic information are input into the Bayesian model to obtain the probability of the service category of each service of the teller, the service category of the teller is determined according to the probability of the service category of the teller, the bank can accurately determine the service handling category of the teller, the service capability of the teller can be better known, and corresponding training is conducted on the service category of the teller to improve the service level of the teller.
As shown in fig. 2, a schematic flow chart of another service class determining method provided in the embodiment of the present application is shown, where the method includes the following steps:
step 201: and acquiring target voice information corresponding to each service in at least one service of the teller.
Step 202: and respectively executing the following processing aiming at the target voice information corresponding to each service to obtain the service category of the teller corresponding to each service.
In the process of implementing step 202, the following steps may be included:
step A2: and obtaining first target sound characteristic information of the teller according to the target voice information of the teller, wherein the first target sound characteristic information embodies the sound characteristic of the teller.
Step B2: determining a probability that the first target sound characteristic information corresponds to at least one traffic class.
Step C2: and converting the target voice information into text information, wherein the text information comprises an answer sentence of the teller replying to the customer question.
Step D2: and matching the text information with preset answer text information to obtain a matching degree, wherein the preset answer text information comprises preset answer sentences, and the matching degree reflects the matching degree of the answer sentences of the teller and the preset answer sentences.
Step E2: and inputting the probability of the first target sound characteristic information corresponding to at least one service class and the matching degree into a first Bayesian model to obtain the service class probability of the teller.
Step F2: and determining the business category of the teller according to the business category probability of the teller.
The execution principle of step 201 to step 202 is consistent with that of step 101 to step 102, and is not described herein again.
Step 203: and determining the performance corresponding to each service according to the service category of each service of the teller and the corresponding relation between the service category and the performance of the teller.
In the embodiment of the application, the teller service category and the teller performance have a corresponding relationship, that is, the teller performance can be obtained by using the teller service category and the corresponding relationship. Specifically, the teller service category and the teller performance may be in a linear correspondence relationship, for example, the teller service category may be 3 categories, the teller performance corresponding to the first service category may be the highest performance, and may be 1.2, the teller performance corresponding to the second service category may be 1, the teller performance corresponding to the third service category may be the lowest performance, and may be 0.8, for example, if the teller service category is the first category, the teller performance of the service is 1.2.
Step 204: and determining the total performance of the teller according to the performance corresponding to each service.
In the embodiment of the application, the total performance of the teller represents the business service level of the teller, and the total performance is higher, which shows that the business service level of the teller is higher, and better user experience can be brought to a client. The total performance may be a sum of the performance corresponding to each business, or may be an average of the performance corresponding to each business.
According to the service type determining method provided by the embodiment of the application, target voice information corresponding to each service in at least one service of a teller is obtained; respectively executing the following processing aiming at the target voice information corresponding to each service to obtain the service category of the teller corresponding to each service: obtaining first target sound characteristic information of the teller according to the target voice information of the teller, wherein the first target sound characteristic information embodies the sound characteristic of the teller; determining a probability that the first target sound characteristic information corresponds to at least one traffic class; converting the target voice information into text information, wherein the text information comprises an answer sentence of the teller for replying to a customer question; matching the text information with preset answer text information to obtain a matching degree, wherein the preset answer text information comprises preset answer sentences, and the matching degree reflects the matching degree of the answer sentences of the teller and the preset answer sentences; inputting the probability of the first target sound characteristic information corresponding to at least one service class and the matching degree into a first Bayesian model to obtain the service class probability of the teller; and determining the business category of the teller according to the business category probability of the teller. Therefore, the method and the device have the advantages that the voice information of each service teller is obtained to obtain the voice characteristic information of each service teller and the probability of the service category corresponding to the voice characteristic information, the voice information is converted into the text information to obtain the matching degree of the teller, the probability and the matching degree of the service category corresponding to the voice characteristic information are input into the Bayesian model to obtain the probability of the service category of each service of the teller, the service category of the teller is determined according to the probability of the service category of the teller, the bank can accurately determine the service handling category of the teller, the service capability of the teller can be better known, and corresponding training is conducted on the service category of the teller to improve the service level of the teller.
As shown in fig. 3, a schematic flow chart of another service class determining method provided in the embodiment of the present application is shown, where the method includes the following steps:
step 301: and acquiring target voice information corresponding to each service in at least one service of the teller.
Step 302: and respectively executing the following processing aiming at the target voice information corresponding to each service to obtain the service category of the teller corresponding to each service.
In the process of implementing step 302, the following steps may be included:
step A3: and obtaining first target sound characteristic information of the teller according to the target voice information of the teller, wherein the first target sound characteristic information embodies the sound characteristic of the teller.
Step B3: determining a probability that the first target sound characteristic information corresponds to at least one traffic class.
Step C3: and converting the target voice information into text information, wherein the text information comprises an answer sentence of the teller replying to the customer question.
Step D3: and matching the text information with preset answer text information to obtain a matching degree, wherein the preset answer text information comprises preset answer sentences, and the matching degree reflects the matching degree of the answer sentences of the teller and the preset answer sentences.
Step E3: and inputting the probability of the first target sound characteristic information corresponding to at least one service class and the matching degree into a first Bayesian model to obtain the service class probability of the teller.
Step F3: and determining the business category of the teller according to the business category probability of the teller.
The execution principle of steps 301 to 302 is consistent with that of steps 101 to 102, and is not described herein again.
Step 303: and acquiring second target sound characteristic information of a client corresponding to each service in at least one service, wherein the second target sound characteristic information embodies the sound characteristics of the client.
In the embodiment of the application, each client has own unique sound characteristic information in the process of transacting business for the client by a bank teller. The target sound characteristic information represents the sound characteristic of a certain client in a certain business transaction. Optionally, the sound characteristics include one or more of: audio features, pitch features, volume features, and speech rate features. For example, a client may have a volume in excess of 80 db during a particular transaction. For another example, a client may have a volume of 50 to 60 db and a speech rate of 80 words per minute during another transaction.
In the process of implementing step 303, the voice information of the client may be recorded by a recording device, and the target sound characteristic information of the client, that is, the sound characteristic of the client, may be obtained by using a sound analysis model.
Step 304: classifying the second target sound characteristic information according to a preset interval, and determining an interval probability corresponding to the second target sound characteristic information; the interval probability is the probability of the interval in which the second target sound characteristic information is located.
In the embodiment of the present application, the preset interval refers to classifying the sound feature information according to emotion categories, that is, the sound features of the client may be classified according to the emotion categories of the client. Specifically, among the volume characteristics of the client, classification may be made according to emotion categories. The emotion category may be 6 categories, and optionally, the emotion category may be a sixth category representing high emotion and high emotional performance, and the emotion category may be the first category representing low emotion and disappointing emotional performance. According to the above, in the volume characteristics of the client, the client may be classified according to 6 emotion categories, for example, the client volume is classified into the first category of emotion categories in the interval of 80 db to 100 db, that is, the client volume is between 80 db to 100 db, and the client is considered to be disappointed in the business handling experience.
After the voice feature information of the client is classified according to the preset interval, the interval probability corresponding to the voice feature information of the client, namely the interval probability corresponding to the voice feature of the client, can be determined. Specifically, the volume characteristic of the client may be in a preset interval between 80 db and 100 db, and the interval probability of the interval may be 0.1, that is, the interval probability corresponding to the volume characteristic of the client may be 0.1.
It should be noted that, the interval probability in the embodiment of the present application is obtained through the historical voice data of the client, that is, the training voice feature information of the client, and the specific steps include:
and acquiring training sound characteristic information of a client corresponding to each service in at least one service, wherein the training sound characteristic information reflects the sound characteristics of the client.
In the embodiment of the application, the historical sound data of the client, namely the training sound characteristic information of the client, can be obtained through the voice information recorded by the recording equipment of the historical transaction service of the client. The training voice characteristic information of the client can reflect the voice characteristics of the client and can also reflect the emotion types, namely the emotion levels, of the client.
Classifying the training sound characteristic information of the client according to a preset interval, and calculating the interval probability of the preset interval.
In the embodiment of the application, the training sound characteristic information of the client is classified according to the preset interval corresponding to the emotion type. After the training voice feature information of the client is classified according to the preset interval, counting the times of the training voice feature information of the client in a certain preset interval and the times of the total sample data of the training voice feature information of the client, wherein the ratio of the times of the training voice feature information of the client in the certain preset interval to the times of the total sample data of the training voice feature information of the client is the interval probability of the preset interval. Specifically, the training volume characteristics of the client may be classified according to the emotion category, for example, the interval where the training volume of the client is 80 db to 100 db is classified as the first category of the emotion category, the number of times that the training volume of the client is in this interval is 3 times, the total sample data of the training volume characteristics of the client is 100 times, and the interval probability is 0.03.
Step 305: respectively inputting the interval probabilities corresponding to the second target sound characteristic information into a second Bayesian model to obtain emotion category probabilities corresponding to the second target sound characteristic information; the emotion category probability is the probability that the second target sound characteristic information corresponds to an emotion category, and the emotion category reflects the emotion of the client.
In the embodiment of the present application, the probability of the section corresponding to the target sound feature information of the client may be determined. The Bayes model, i.e. Bayes formula, is a mathematical formula for calculating probability. The emotion category probability is a probability of an emotion category corresponding to the voice feature information of the client, optionally, the emotion category probability may be a probability that both the volume feature and the speech rate feature of the client are a first category of the emotion category, the emotion category probability may also be a probability that both the tone feature and the audio feature of the client are a sixth category of the emotion category, and the emotion category probability may also be a probability that the volume feature, the speech rate feature, the tone feature and the audio feature of the client are a second category of the emotion category. The emotion classification of the client can reflect the emotion of the client, specifically, the emotion classification can be 6 classifications, optionally, the emotion classification can be a sixth classification, the represented emotion is high, the emotional performance can be happy, the emotion classification can also be a first classification, the represented emotion is low, and the emotional performance can be disappointed.
In the process of implementing step 305 specifically, the interval probability corresponding to the voice feature of the client classified according to the emotion category may be calculated as the emotion category probability corresponding to the target voice feature information of the client by using a bayesian formula. Specifically, for the volume feature, the interval probability corresponding to the first emotion category is 0.1, for the speech rate feature, the interval probability corresponding to the first emotion category is 0.2, and the two interval probabilities are input into a bayesian formula to obtain the probability of the first emotion category relative to the volume feature and the speech rate feature, that is, the first emotion category probability.
Step 306: and determining the emotion category of the client corresponding to each service in the at least one service according to the emotion category probability of the second target sound characteristic information of the client corresponding to each service in the at least one service.
In the embodiment of the application, the emotion category probability of the client can reflect the emotion level of the client, specifically, the emotion category of the client can be 3 categories, the first category emotion represents that the emotion of the client is high, and the emotional performance can be happy. The second category of emotions represents a more peaceful client emotion and the emotional expression may be smiling, and the third category of emotional expression represents a less emotional expression and the emotional expression may be disappointing. The first emotion category probability is 0.5, the second emotion category is 0.3, and the third emotion category is 0.2. The emotion classification of the client can be determined by comparing the emotion classification probabilities of the three classes, specifically, the emotion classification with the highest emotion classification probability can be determined as the emotion classification of the client, and the first emotion classification is 0.5 and the maximum emotion classification of the client can be determined as the first emotion classification.
Step 307: inputting a business category corresponding to each business in at least one business of the teller and an emotion category of a client corresponding to each business in the at least one business into a genetic algorithm model to obtain performance corresponding to each business in the at least one business; wherein the genetic algorithm model is obtained by pre-training.
In the embodiment of the application, the teller performance is obtained by inputting the business class of a certain business of the teller and the emotion class of a client of the business into a genetic algorithm model. The genetic algorithm model is obtained through pre-training.
Step 308: and determining the total performance of the teller according to the performance corresponding to each service in the at least one service.
In the embodiment of the application, the total performance of the teller represents the business service level of the teller, and the total performance is higher, which shows that the business service level of the teller is higher, and better user experience can be brought to a client. The total performance may be a sum of the performance corresponding to each business, or may be an average of the performance corresponding to each business.
According to the service type determining method provided by the embodiment of the application, target voice information corresponding to each service in at least one service of a teller is obtained; respectively executing the following processing aiming at the target voice information corresponding to each service to obtain the service category of the teller corresponding to each service: obtaining first target sound characteristic information of the teller according to the target voice information of the teller, wherein the first target sound characteristic information embodies the sound characteristic of the teller; determining a probability that the first target sound characteristic information corresponds to at least one traffic class; converting the target voice information into text information, wherein the text information comprises an answer sentence of the teller for replying to a customer question; matching the text information with preset answer text information to obtain a matching degree, wherein the preset answer text information comprises preset answer sentences, and the matching degree reflects the matching degree of the answer sentences of the teller and the preset answer sentences; inputting the probability of the first target sound characteristic information corresponding to at least one service class and the matching degree into a first Bayesian model to obtain the service class probability of the teller; and determining the business category of the teller according to the business category probability of the teller. Therefore, the method and the device have the advantages that the voice information of each service teller is obtained to obtain the voice characteristic information of each service teller and the probability of the service category corresponding to the voice characteristic information, the voice information is converted into the text information to obtain the matching degree of the teller, the probability and the matching degree of the service category corresponding to the voice characteristic information are input into the Bayesian model to obtain the probability of the service category of each service of the teller, the service category of the teller is determined according to the probability of the service category of the teller, the bank can accurately determine the service handling category of the teller, the service capability of the teller can be better known, and corresponding training is conducted on the service category of the teller to improve the service level of the teller.
Based on the service class determination method provided in the embodiment of the present application, an embodiment of the present application further provides a service class determination device 400, and as shown in fig. 4, a schematic structural diagram of a service class determination device provided in the embodiment of the present application is shown, where the method includes:
a first obtaining unit 410, configured to obtain target voice information corresponding to each service in at least one service of a teller;
a processing unit 420, configured to perform the following processing on the target voice information corresponding to each service, respectively, to obtain a service category of the teller corresponding to each service;
the processing unit 420 includes:
the characteristic obtaining unit 421 is configured to obtain first target sound characteristic information of the teller according to the target voice information of the teller, where the first target sound characteristic information represents a sound characteristic of the teller;
a first determining unit 422, configured to determine a probability that the first target sound characteristic information corresponds to at least one traffic class;
a conversion unit 423, configured to convert the target voice information into text information, where the text information includes an answer sentence for the teller to reply to a customer question;
the matching unit 424 is configured to match the text information with preset answer text information to obtain a matching degree, where the preset answer text information includes preset answer sentences, and the matching degree represents a matching degree between the answer sentences of the teller and the preset answer sentences;
a first input unit 425, configured to input, to a first bayesian model, a probability that the first target sound feature information corresponds to at least one service class and the matching degree, so as to obtain a service class probability of the teller;
a second determining unit 426, configured to determine the business category of the teller according to the business category probability of the teller.
The device also includes:
the third determining unit is used for determining the performance corresponding to each business according to the business category of each business of the teller and the corresponding relation between the business category and the performance of the teller;
and the fourth determining unit is used for determining the total performance of the teller according to the performance corresponding to each service.
The device also includes:
a second obtaining unit, configured to obtain second target sound feature information of a client corresponding to each service in at least one service, where the second target sound feature information represents a sound feature of the client;
a fifth determining unit, configured to classify the second target sound feature information according to a preset interval, and determine an interval probability corresponding to the second target sound feature information; the interval probability is the probability of the interval in which the second target sound characteristic information is located;
the second input unit is used for respectively inputting the interval probabilities corresponding to the second target sound characteristic information into a second Bayesian model to obtain emotion category probabilities corresponding to the second target sound characteristic information; the emotion category probability is the probability that the second target sound characteristic information corresponds to an emotion category, and the emotion category reflects the emotion of the client;
a sixth determining unit, configured to determine, according to an emotion category probability of second target sound feature information of a client corresponding to each service in the at least one service, an emotion category of the client corresponding to each service in the at least one service;
a third input unit, configured to input a service category corresponding to each service in at least one service of the teller and an emotion category of a client corresponding to each service in the at least one service into a genetic algorithm model, so as to obtain a performance corresponding to each service in the at least one service; wherein the genetic algorithm model is obtained by pre-training;
and the seventh input unit is used for determining the total performance of the teller according to the performance corresponding to each service in the at least one service.
The second determining unit 426, according to the teller's business class probability, determining the teller's business class includes:
the second determining unit 426 determines the teller with the highest probability of the business class as the business class of the teller.
The device sound characteristics include one or more of the following:
audio features, pitch features, volume features, and speech rate features.
The terms "first," "second," "third," "fourth," and the like in the description and in the claims of the present application and in the drawings described above, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the several embodiments provided in the present application, it should be understood that the disclosed system and method may be implemented in other ways. For example, the division of the unit is only one logical service division, and there may be other division ways in actual implementation, for example, a plurality of units or components may be combined or may be 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, each service unit in the embodiments of the present application 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 hardware form, and can also be realized in a software service unit form.
The above-described embodiments are intended to explain the objects, aspects and advantages of the present invention in further detail, and it should be understood that the above-described embodiments are merely exemplary embodiments of the present invention.
The above embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present application.

Claims (10)

1. A method for determining a traffic class, the method comprising:
acquiring target voice information corresponding to each service in at least one service of a teller;
respectively executing the following processing aiming at the target voice information corresponding to each service to obtain the service category of the teller corresponding to each service:
obtaining first target sound characteristic information of the teller according to the target voice information of the teller, wherein the first target sound characteristic information embodies the sound characteristic of the teller;
determining a probability that the first target sound characteristic information corresponds to at least one traffic class;
converting the target voice information into text information, wherein the text information comprises an answer sentence of the teller for replying to a customer question;
matching the text information with preset answer text information to obtain a matching degree, wherein the preset answer text information comprises preset answer sentences, and the matching degree reflects the matching degree of the answer sentences of the teller and the preset answer sentences;
inputting the probability of the first target sound characteristic information corresponding to at least one service class and the matching degree into a first Bayesian model to obtain the service class probability of the teller;
and determining the business category of the teller according to the business category probability of the teller.
2. The method of claim 1, further comprising:
determining the performance corresponding to each business according to the business category of each business of the teller and the corresponding relation between the business category and the performance of the teller;
and determining the total performance of the teller according to the performance corresponding to each service.
3. The method of claim 2, further comprising:
acquiring second target sound characteristic information of a client corresponding to each service in at least one service, wherein the second target sound characteristic information embodies the sound characteristics of the client;
classifying the second target sound characteristic information according to a preset interval, and determining an interval probability corresponding to the second target sound characteristic information; the interval probability is the probability of the interval in which the second target sound characteristic information is located;
respectively inputting the interval probabilities corresponding to the second target sound characteristic information into a second Bayesian model to obtain emotion category probabilities corresponding to the second target sound characteristic information; the emotion category probability is the probability that the second target sound characteristic information corresponds to an emotion category, and the emotion category reflects the emotion of the client;
determining the emotion category of the client corresponding to each service in the at least one service according to the emotion category probability of the second target sound characteristic information of the client corresponding to each service in the at least one service;
inputting a business category corresponding to each business in at least one business of the teller and an emotion category of a client corresponding to each business in the at least one business into a genetic algorithm model to obtain performance corresponding to each business in the at least one business; wherein the genetic algorithm model is obtained by pre-training;
and determining the total performance of the teller according to the performance corresponding to each service in the at least one service.
4. The method of claim 1, wherein determining the teller's business class based on the teller's business class probability comprises:
and determining the teller with the highest business class probability as the business class of the teller.
5. The method of claim 1, wherein the sound features comprise one or more of:
audio features, pitch features, volume features, and speech rate features.
6. An apparatus for determining a traffic class, the apparatus comprising:
the teller terminal comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring target voice information corresponding to each service in at least one service of a teller;
the processing unit is used for respectively executing the following processing aiming at the target voice information corresponding to each service to obtain the service type of the teller corresponding to each service;
the processing unit includes:
the system comprises a characteristic acquisition unit, a voice recognition unit and a voice recognition unit, wherein the characteristic acquisition unit is used for acquiring first target voice characteristic information of a teller according to target voice information of the teller, and the first target voice characteristic information embodies voice characteristics of the teller;
a first determining unit, configured to determine a probability that the first target sound characteristic information corresponds to at least one traffic class;
the conversion unit is used for converting the target voice information into text information, and the text information comprises an answer sentence of the customer question replied by the teller;
the matching unit is used for matching the text information with preset answer text information to obtain a matching degree, wherein the preset answer text information comprises preset answer sentences, and the matching degree reflects the matching degree of the answer sentences of the teller and the preset answer sentences;
the first input unit is used for inputting the probability of the first target sound characteristic information corresponding to at least one business category and the matching degree into a first Bayesian model to obtain the business category probability of the teller;
and the second determining unit is used for determining the business category of the teller according to the business category probability of the teller.
7. The apparatus of claim 6, further comprising:
the third determining unit is used for determining the performance corresponding to each business according to the business category of each business of the teller and the corresponding relation between the business category and the performance of the teller;
and the fourth determining unit is used for determining the total performance of the teller according to the performance corresponding to each service.
8. The apparatus of claim 7, further comprising:
a second obtaining unit, configured to obtain second target sound feature information of a client corresponding to each service in at least one service, where the second target sound feature information represents a sound feature of the client;
a fifth determining unit, configured to classify the second target sound feature information according to a preset interval, and determine an interval probability corresponding to the second target sound feature information; the interval probability is the probability of the interval in which the second target sound characteristic information is located;
the second input unit is used for respectively inputting the interval probabilities corresponding to the second target sound characteristic information into a second Bayesian model to obtain emotion category probabilities corresponding to the second target sound characteristic information; the emotion category probability is the probability that the second target sound characteristic information corresponds to an emotion category, and the emotion category reflects the emotion of the client;
a sixth determining unit, configured to determine, according to an emotion category probability of second target sound feature information of a client corresponding to each service in the at least one service, an emotion category of the client corresponding to each service in the at least one service;
a third input unit, configured to input a service category corresponding to each service in at least one service of the teller and an emotion category of a client corresponding to each service in the at least one service into a genetic algorithm model, so as to obtain a performance corresponding to each service in the at least one service; wherein the genetic algorithm model is obtained by pre-training;
and the seventh input unit is used for determining the total performance of the teller according to the performance corresponding to each service in the at least one service.
9. The apparatus of claim 6, wherein the second determining unit determines the business class of the teller based on the business class probability of the teller comprises:
the second determination unit determines the teller as the business class with the highest probability of the business class.
10. The apparatus of claim 6, wherein the sound features comprise one or more of:
audio features, pitch features, volume features, and speech rate features.
CN202011049699.2A 2020-09-29 2020-09-29 Service class determining method and device Active CN112131369B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011049699.2A CN112131369B (en) 2020-09-29 2020-09-29 Service class determining method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011049699.2A CN112131369B (en) 2020-09-29 2020-09-29 Service class determining method and device

Publications (2)

Publication Number Publication Date
CN112131369A true CN112131369A (en) 2020-12-25
CN112131369B CN112131369B (en) 2024-02-02

Family

ID=73844680

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011049699.2A Active CN112131369B (en) 2020-09-29 2020-09-29 Service class determining method and device

Country Status (1)

Country Link
CN (1) CN112131369B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113192511A (en) * 2021-04-27 2021-07-30 中国工商银行股份有限公司 Information input method, information input device, electronic device, and storage medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107316643A (en) * 2017-07-04 2017-11-03 科大讯飞股份有限公司 Voice interactive method and device
CN107818798A (en) * 2017-10-20 2018-03-20 百度在线网络技术（北京）有限公司 Customer service quality evaluating method, device, equipment and storage medium
CN108228622A (en) * 2016-12-15 2018-06-29 平安科技（深圳）有限公司 The sorting technique and device of traffic issues
CN108764753A (en) * 2018-06-06 2018-11-06 平安科技（深圳）有限公司 Test method, apparatus, computer equipment and the storage medium of business personnel's ability
CN112116165A (en) * 2020-09-28 2020-12-22 中国银行股份有限公司 Service performance determination method and device

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108228622A (en) * 2016-12-15 2018-06-29 平安科技（深圳）有限公司 The sorting technique and device of traffic issues
CN107316643A (en) * 2017-07-04 2017-11-03 科大讯飞股份有限公司 Voice interactive method and device
CN107818798A (en) * 2017-10-20 2018-03-20 百度在线网络技术（北京）有限公司 Customer service quality evaluating method, device, equipment and storage medium
CN108764753A (en) * 2018-06-06 2018-11-06 平安科技（深圳）有限公司 Test method, apparatus, computer equipment and the storage medium of business personnel's ability
CN112116165A (en) * 2020-09-28 2020-12-22 中国银行股份有限公司 Service performance determination method and device

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113192511A (en) * 2021-04-27 2021-07-30 中国工商银行股份有限公司 Information input method, information input device, electronic device, and storage medium
CN113192511B (en) * 2021-04-27 2023-01-06 中国工商银行股份有限公司 Information input method, information input device, electronic device, and storage medium

Also Published As

Publication number Publication date
CN112131369B (en) 2024-02-02

Similar Documents

Publication Publication Date Title
US7912714B2 (en) Method for segmenting communication transcripts using unsupervised and semi-supervised techniques
CN111081279A (en) Voice emotion fluctuation analysis method and device
CN111694938B (en) Emotion recognition-based reply method and device, computer equipment and storage medium
CN110390946A (en) A kind of audio signal processing method, device, electronic equipment and storage medium
CN111428017B (en) Human-computer interaction optimization method and related device
CN111243602A (en) Voiceprint recognition method based on gender, nationality and emotional information
CN111539221B (en) Data processing method and system
CN111182162A (en) Telephone quality inspection method, device, equipment and storage medium based on artificial intelligence
CN108228622A (en) The sorting technique and device of traffic issues
CN107766560B (en) Method and system for evaluating customer service flow
CN110955770A (en) Intelligent dialogue system
CN111274390A (en) Emotional reason determining method and device based on dialogue data
CN112131369B (en) Service class determining method and device
CN112116165B (en) Business performance determining method and device
CN112667792A (en) Man-machine conversation data processing method and device, computer equipment and storage medium
CN116127011A (en) Intention recognition method, device, electronic equipment and storage medium
CN114822557A (en) Method, device, equipment and storage medium for distinguishing different sounds in classroom
CN113705199A (en) Work order priority confirmation method, work order priority confirmation device, work order priority confirmation electronic equipment, work order priority confirmation medium and work order priority confirmation product
CN113111855A (en) Multi-mode emotion recognition method and device, electronic equipment and storage medium
KR101066228B1 (en) Emotion classification system and method thereof
CN111178982A (en) Customer satisfaction analysis method, storage medium and computer device
CN112132477A (en) Service performance determination method and device
CN110807370A (en) Multimode-based conference speaker identity noninductive confirmation method
CN117131183B (en) Customer service automatic reply method and system based on session simulation
CN107798480B (en) Service quality evaluation method and system for customer service

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