A customer service seat allocation method, device, computer equipment and storage medium

Abstract

The embodiment of the present application discloses a method, device, computer equipment and storage medium for allocating customer service seats, wherein the method includes: obtaining historical call data and customer information of customers of the banking system, identifying personality types based on the customer information to obtain the customer personality types of the customers, extracting audio features and text features of the historical call data, identifying the friendliness of the audio features and text features through a friendliness prediction model to obtain the friendliness level of the customer, the banking system receiving a customer service seat service call dialed by a target customer, obtaining the customer personality type and friendliness level of the target customer, and allocating customer service seat personnel who answer the call from the target customer based on the customer personality type and friendliness level. Through the above method, when a customer calls, the customer service seat personnel who answer the call are allocated according to the customer personality type and friendliness level pre-obtained by the platform for the customer, so as to improve the call quality.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a customer service seat allocation method, device, computer equipment and storage medium. Background Art

[0002] With the development of social economy and financial technology, people have higher and higher requirements for the services of financial enterprises. In order to better serve customers, enterprises have set up customer service seats accordingly. Customer service seats can answer manual calls made by customers and solve problems for customers.

[0003] In the prior art, when a customer calls, a corresponding customer service agent is assigned to the customer based on the customer service agent's free time to answer the call. The allocation method is simple. If the current customer is difficult to communicate with and the assigned free customer service agent is a new employee who lacks answering experience, the customer service agent will be unable to respond to emergencies, thereby affecting the current call quality. Summary of the invention

[0004] The embodiments of the present application provide a customer service seat allocation method, apparatus, computer equipment and storage medium to solve the problems raised in the above-mentioned background technology.

[0005] In a first aspect, an embodiment of the present application provides a method for allocating customer service seats, the method comprising:

[0006] Acquire historical call data and customer information of customers of the banking system, wherein the historical call data includes at least one historical call audio and its corresponding historical call audio text, and the customer information includes basic customer information and historical business information of the customer;

[0007] Performing personality type identification based on the basic information of the customer and the historical business handling information to obtain the customer personality type of the customer;

[0008] Acquire the audio features of the historical call audio based on the audio feature extraction model, and acquire the text features of the historical call audio text based on the text feature extraction model;

[0009] Performing friendliness identification on the audio features and the text features using a friendliness prediction model to obtain the friendliness level of the customer;

[0010] The bank system receives a customer service agent service call from a target customer, obtains the customer personality type and the friendliness level of the target customer, and assigns a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level.

[0011] In some embodiments, the step of acquiring the audio features of the historical call audio based on the audio feature extraction model, and acquiring the text features of the historical call audio text based on the text feature extraction model, includes:

[0012] Segmenting the historical call audio to obtain multiple audio segments;

[0013] Performing text translation on each of the audio segments to obtain a text segment;

[0014] Extracting audio features of each audio segment based on an audio feature extraction model;

[0015] Based on the text feature extraction model, the text features of each text segment are extracted.

[0016] In some embodiments, the performing friendliness identification on the audio features and the text features using a friendliness prediction model to obtain the friendliness level of the customer includes:

[0017] Using the friendliness prediction model, emotion recognition is performed on the audio features and text features of each audio segment and its corresponding text segment to obtain a corresponding emotion type, wherein the emotion type at least includes a negative emotion type;

[0018] Selecting the audio segment whose emotion type is a negative emotion type as a negative emotion audio segment, and calculating the ratio of the negative emotion audio segment to all the audio segments;

[0019] Based on the ratio, a predicted friendliness level is obtained.

[0020] In some embodiments, the performing friendliness identification on the audio features and the text features using a friendliness prediction model to obtain the friendliness level of the customer includes:

[0021] By using the friendliness prediction model, the audio features and text features of each of the historical call audios and the corresponding historical call audio texts are identified to obtain the corresponding friendliness level to be predicted;

[0022] Based on the level proportions of all the friendliness levels to be predicted, a predicted friendliness level is obtained.

[0023] In some embodiments, the bank system receives a customer service agent service call from a target customer, obtains the customer personality type and the friendliness level of the target customer, and assigns a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level, including:

[0024] Selecting historical call audios corresponding to the predicted friendliness level higher than the preset level from all the predicted friendliness levels;

[0025] Confirm the gender of the customer service agent of the selected historical call audio;

[0026] Based on the male-female ratio of the personnel, determine the gender of the customer service agent preferred by the customer;

[0027] The customer service agent who answers the incoming call is assigned based on the gender of the customer service agent preferred by the customer, the customer's personality type and the friendliness level.

[0028] In some embodiments, the bank system receives a customer service agent service call from a target customer, obtains the customer personality type and the friendliness level of the target customer, and assigns a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level, including:

[0029] The bank system receives a customer service agent service call from a target customer, and obtains the customer personality type and the friendliness level of the target customer;

[0030] Identify available customer service agents who are idle;

[0031] According to the personality type of the customer, a customer service agent to be determined is selected from the idle customer service agents;

[0032] A customer service agent who answers the incoming call is selected from the customer service agents to be determined according to the friendliness level.

[0033] In some embodiments, after the banking system receives a customer service agent service call from a target customer, obtains the customer personality type and the friendliness level of the target customer, and assigns a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level, the method further includes:

[0034] Determining a matching customer emotion management solution based on the friendliness level;

[0035] The customer personality type and the customer emotion management plan are displayed on a display interface of a customer service agent answering the incoming call.

[0036] In a second aspect, an embodiment of the present application provides a customer service seat allocation device, the device comprising:

[0037] An information acquisition unit, used to acquire historical call data and customer information of customers of the banking system, wherein the historical call data includes at least one historical call audio and its corresponding historical call audio text, and the customer information includes basic customer information and historical business information of the customer;

[0038] A personality type acquisition unit, configured to identify the personality type of the customer based on the basic information of the customer and the historical business information, and obtain the personality type of the customer;

[0039] A feature acquisition unit, used to acquire audio features of the historical call audio based on an audio feature extraction model, and to acquire text features of the historical call audio text based on a text feature extraction model;

[0040] A friendliness acquisition unit, configured to identify the friendliness of the audio feature and the text feature using a friendliness prediction model to obtain the friendliness level of the customer;

[0041] The personnel allocation unit is used for the bank system to receive a customer service agent service call from a target customer, obtain the customer personality type and the friendliness level of the target customer, and allocate a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level.

[0042] In a third aspect, an embodiment of the present application further provides a computer device, including a memory and a processor, wherein the memory is used to store instructions and data, and the processor is used to execute the customer service seat allocation method described above.

[0043] In a fourth aspect, an embodiment of the present application further provides a storage medium, wherein a plurality of instructions are stored in the storage medium, and the instructions are suitable for being loaded by a processor to execute the customer service seat allocation method described above.

[0044] The customer service agent allocation method in the embodiment of the present application obtains the customer's personality type and friendliness level in advance based on the customer's customer information and historical call data through the bank platform. When the customer calls, the customer service agent who answers the call is allocated based on the customer's personality type and friendliness level to improve the call quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1It is a flow chart of a customer service seat allocation method provided in an embodiment of the present application.

[0047] Figure 2 This is another flow chart of a customer service seat allocation method provided in an embodiment of the present application.

[0048] Figure 3 It is a structural schematic diagram of a customer service seat allocation device provided in an embodiment of the present application.

[0049] Figure 4 It is a structural diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0051] The embodiments of the present application provide a customer service seat allocation method, device, computer equipment and storage medium, which can allocate suitable customer service seats to callers based on their historical data, thereby improving the quality of incoming calls. The customer service seat allocation method, device, computer equipment and storage medium will be described in detail below.

[0052] See also Figure 1 , Figure 1 A flowchart of a customer service seat allocation method provided in an embodiment of the present application, the method includes the following contents:

[0053] 101. Obtain historical call data and customer information of customers of the banking system, wherein the historical call data includes at least one historical call audio and its corresponding historical call audio text, and the customer information includes basic customer information and historical business information of the customer.

[0054] The bank system can obtain and update the customer's historical call data and customer information by setting up scheduled processing tasks, where the historical call data is the bank customer service agent's manual phone calls made by the customer before the bank system updates the data.

[0055] The historical call data includes at least one historical call audio and its corresponding historical call audio text. The historical call audio is the audio data of the call between the customer and the customer service agent in the historical call, and the historical call audio text is the text data corresponding to the historical call audio. It can be considered that the bank system obtains the historical call audio, and obtains the historical call audio text by text translation of the historical call audio.

[0056] Customer information is information related to personal customers, which can be considered as personal information entered by customers into the bank system when handling banking business, or as information related to customers obtained by the bank system from other platforms through the customer's personal related information.

[0057] The basic information of the customer in the customer information may include name, age, ID number, education, work and work experience, etc. The historical business information in the customer information includes the types of business that the customer has handled in the bank system, and the investment selection preference questionnaire related to business risks that the customer has done in the bank system, etc. The use of the basic information of the customer in the embodiment of the present application complies with the requirements of laws and regulations.

[0058] 102. Perform personality type identification based on the basic information of the customer and the historical business handling information to obtain the customer personality type of the customer.

[0059] In an embodiment of the present application, the basic information of the customer and the historical business information can be input into the deep neural model to obtain the customer's personality type.

[0060] In the deep neural model, the model can comprehensively analyze the customer's personality type based on information such as the customer's job, age, type of business handled, and investment choice preference questionnaire results related to business risks.

[0061] For example, if a customer works in sales and, through analysis of the results of an investment choice preference questionnaire related to the types of business the customer handles and business risks, it is concluded that the customer is a risk-seeking type, then it can be predicted that the customer's personality type is an extrovert.

[0062] Optionally, all basic customer information and historical business information can be input into the deep learning model. The deep learning model selects reference data according to a preset selection method, and then determines the final data for predicting the customer's personality type based on the reference data. The customer's personality type is predicted based on the final data.

[0063] When setting the customer's personality type, you can set it to include extroverted personality, neutral personality and introverted personality. Among them, the introverted personality can be divided into ordinary introverted personality and extreme introverted personality, etc. There is no restriction on the setting method of the customer's personality type.

[0064] 103. Acquire audio features of the historical call audio based on an audio feature extraction model, and acquire text features of the historical call audio text based on a text feature extraction model.

[0065] The extraction of audio features of historical call audio can be achieved through an audio feature extraction model. The audio feature extraction model can be a MATLAB model, a Python model, or other models that can realize audio feature extraction.

[0066] The text features of the historical call audio texts may be extracted through a text feature extraction model, which may be a bag-of-words model or other models that can implement text feature extraction.

[0067] In the embodiments of the present application, there is no restriction on the model types of the audio feature extraction model and the text feature extraction model.

[0068] In some embodiments, historical call audio is segmented to obtain multiple audio segments, each audio segment is translated into text to obtain a text segment, audio features of each audio segment are extracted based on an audio feature extraction model, and text features of each text segment are extracted based on a text feature extraction model.

[0069] For example, the historical call audio is segmented according to the blank time of the blank audio segment. The blank audio segment is an audio segment without sound. The blank time can be considered as the stop time after the customer or customer service agent finishes a sentence or a paragraph.

[0070] Determine the position and blank time of the blank audio segment in the historical call audio, compare the blank time with the preset time, and when the blank time is greater than the preset time, it can be considered that the blank audio segment at that location is the stop time after the customer or customer service agent finishes a sentence or a paragraph, select the blank audio segment with a blank time greater than the preset time, and divide the historical call audio based on the selected blank audio segment to obtain multiple audio segments.

[0071] The above-mentioned segmentation of the historical call audio based on the selected blank audio segments to obtain multiple audio segments may include: based on the selected blank audio segments, confirming the positions of these blank audio segments in the historical call audio, and preprocessing the historical call audio, the preprocessing method includes noise reduction processing and silence filtering processing, in the silence filtering processing, all blank audio segments in the historical call audio will be filtered out, and the historical call audio after preprocessing is segmented based on the positions of the selected blank audio segments to obtain multiple audio segments, then, the obtained audio segments are audible audio segments.

[0072] After obtaining the audio segments, each audio segment can be translated into text to obtain the corresponding text segment. The text translation can be achieved through the corresponding model.

[0073] It should be noted that when extracting features from historical call audio and historical call audio text, feature extraction can be performed directly on all historical call audio and historical call audio text of the customer, and the obtained audio features and text features correspond to the entire historical call data. Alternatively, when extracting features from historical call audio and historical call audio text, feature extraction can be performed on each historical call audio and its corresponding historical call audio text, and the obtained audio features and text features correspond to a historical call audio and its corresponding historical call audio text. Then, the audio features and text features of the historical call data are the fusion of the audio features and text features of all historical call audio and its corresponding historical call audio text.

[0074] 104. Perform friendliness identification on the audio features and the text features using a friendliness prediction model to obtain the friendliness level of the customer.

[0075] After obtaining the audio features and text features of the customer's historical call data, the audio features and text features are input into the friendliness prediction model for friendliness identification to obtain the friendliness level corresponding to the customer. The friendliness level can be set to the first level, the second level, the third level, the fourth level and the fifth level, where the first level is very friendly, the second level is friendly, the third level is neutral, the fourth level is unfriendly, and the fifth level is very unfriendly.

[0076] Optionally, the setting of the friendliness level may include other methods, which are not limited here.

[0077] In some embodiments, through the friendliness prediction model, emotion recognition is performed on the audio features and text features of each audio segment and its corresponding text segment to obtain the corresponding emotion type, and the emotion type at least includes a negative emotion type. The audio segment with a negative emotion type is selected as the negative emotion audio segment, and the ratio of the negative emotion audio segment to all audio segments is calculated. Based on the ratio, a predicted friendliness level is obtained.

[0078] After the historical call audio is segmented into multiple audio segments and the corresponding text segments are obtained through the audio segments, the audio features of each audio segment and the text features of the text segment corresponding to the audio segment are obtained, and emotion recognition is performed on the audio features and text features of each audio segment and its corresponding text segment to obtain the corresponding emotion type.

[0079] The above-mentioned emotion recognition is performed on the audio features and text features of each audio segment and its corresponding text segment to obtain the corresponding emotion type, which may include: obtaining the emotion type of the audio segment according to the audio feature recognition, obtaining the emotion type of the text segment according to the text feature recognition, and comparing the two emotion types. If they are consistent, the emotion type is confirmed as the emotion type of the corresponding audio segment; if they are inconsistent, the emotion type of the corresponding audio segment is confirmed according to the preset emotion type priority.

[0080] The emotion type can be set to include negative emotion type, neutral emotion type, positive emotion type, etc.

[0081] Based on the recognition results of the emotion type of each audio segment, audio segments with negative emotion types are selected, the number of these audio segments is calculated, and the total number of all audio segments is calculated. The number of audio segments with negative emotion types is compared with the total number of segments, that is, the proportion of audio segments with negative emotion types in all audio segments is confirmed, and the friendliness level is determined according to the ratio.

[0082] The above-mentioned confirmation of the proportion of audio segments of negative emotion types in all audio segments, and determining the friendliness level according to the ratio, may include: setting a friendliness reference table, in which the relationship between the ratio and the friendliness level is indicated, for example, the friendliness level includes the first level, the second level, the third level, the fourth level and the fifth level, the first level is very friendly, the second level is friendly, the third level is neutral, the fourth level is unfriendly, and the fifth level is very unfriendly. When the ratio is lower than 1 / 10, the friendliness level is the second level, when the ratio is lower than 0, the friendliness level is the first level, when the ratio is higher than 1 / 3 but lower than 1 / 2, the friendliness level is the fourth level, when the ratio is higher than 1 / 2, the friendliness level is the fifth level, and so on.

[0083] In other embodiments, the friendliness prediction model is used to identify the audio features and text features of each historical call audio and its corresponding historical call audio text to obtain the corresponding friendliness level to be predicted, and the predicted friendliness level is obtained based on the level ratio of all friendliness levels to be predicted.

[0084] The audio features and text features of each historical call audio and its corresponding historical call audio text are obtained, and then based on the friendliness prediction model, the audio features and text features are used to obtain the predicted friendliness level of each historical call audio and its corresponding historical call audio text.

[0085] All the friendliness levels to be predicted of the customer's historical call data are divided into the same level, and the friendliness levels to be predicted with lower friendliness are selected, and the ratio of these friendliness levels to be predicted and all friendliness levels to be predicted is calculated to obtain the friendliness level of the customer.

[0086] It can be understood that the friendliness level to be predicted is the friendliness level corresponding to each historical call audio and its corresponding historical call audio text.

[0087] For example, the friendliness level includes the first level, the second level, the third level, the fourth level and the fifth level, the first level is very friendly, the second level is friendly, the third level is neutral, the fourth level is unfriendly, and the fifth level is very unfriendly. The total number of all friendliness levels to be predicted of the customer's historical call data is calculated, and all friendliness levels to be predicted are divided according to the same friendliness level to obtain multiple friendliness level groups to be predicted, including the fourth level group and the fifth level group, and the number of friendliness groups to be predicted included in the fourth level group and the fifth level group is calculated, and the number of groups of the fourth level group and the fifth level group is added, and then the ratio is calculated with the total number to obtain the proportion of the friendliness levels to be predicted of the fourth level group and the fifth level group, and the friendliness level of the customer is obtained based on the proportion.

[0088] Furthermore, for example, a proportion greater than or equal to 40% is set as the fifth level, corresponding to very unfriendly. If the total number of friendliness levels to be predicted in the customer's historical call data is 20, of which the fourth level grouping includes 5 friendliness levels to be predicted and the fifth level grouping includes 5 friendliness levels to be predicted, then the number of groups in the fourth level grouping and the fifth level grouping is added together to obtain 10 friendliness levels to be predicted. The proportion of 10 and 20 is 50%, so it can be judged that the customer's friendliness is very unfriendly, and the customer's friendliness level is the fifth level.

[0089] Optionally, the above example is only an example provided in the embodiment of the present application, and the setting method between the friendliness level and the proportion can also be other settings, which are not limited here.

[0090] 105. The bank system receives a customer service agent service call from a target customer, obtains the customer personality type and the friendliness level of the target customer, and assigns a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level.

[0091] Since obtaining the customer's personality type and friendliness level requires a certain amount of processing time, and the waiting time between the customer dialing in and the call being answered is limited, the customer's personality type and friendliness level are usually completed before the customer dials in.

[0092] After receiving a customer service call from a customer service agent, the corresponding customer personality type and friendliness level can be obtained according to the dialed phone number, and the customer service agent who answers the call is assigned according to the customer personality type and friendliness level. For example, when the customer personality type is extroverted, and the friendliness level corresponding to the customer friendliness level is friendly, the customer service agent can be randomly assigned. When the customer personality type is introverted, and the friendliness level corresponding to the customer friendliness level is unfriendly, a customer service agent with higher experience is assigned.

[0093] In some embodiments, historical call audio corresponding to a predicted friendliness level higher than a preset level is selected from all friendliness levels to be predicted, the gender of the customer service agent of the selected historical call audio is confirmed, and based on the male-female ratio of the agent, the gender of the customer service agent preferred by the customer is determined; based on the gender of the customer service agent preferred by the customer, the customer's personality type and friendliness level, the customer service agent who answers the incoming call is assigned.

[0094] After obtaining the predicted friendliness level corresponding to each historical call audio and its corresponding historical call audio text, the historical call audios corresponding to the predicted friendliness level with higher friendliness are selected, and the gender of the customer service agents therein are identified through the selected historical call audios, thereby knowing the gender of the customer service agent corresponding to the customer under more friendly circumstances.

[0095] For example, the customer's personality type is introverted, and the friendliness level corresponding to the customer's friendliness level is unfriendly. However, in the customer's historical call data, the customer service agent who answered the call in the historical call audio with friendliness levels of friendly and very friendly was female. Therefore, it can be considered that the customer is relatively friendly to female customer service agents. When assigning customer service agents to answer the calls, customer service agents with high experience and female gender can be assigned.

[0096] In other embodiments, the banking system receives a customer service agent service call from a target customer, obtains the target customer's personality type and friendliness level, determines idle customer service agents who are idle, selects customer service agents to be determined from the idle customer service agents based on the customer's personality type, and selects a customer service agent who answers the call from the customer service agents to be determined based on the friendliness level.

[0097] Usually, banks have multiple customer service agents who solve customers' problems through manual conversations with them. Since the time when customer service agents answer the phone is uncertain, it is necessary to confirm the status of the customer service agents before assigning them according to the customer's personality type and friendliness level.

[0098] The customer service agent allocation method of the embodiment of the present application includes obtaining the historical call data and customer information of the customers of the banking system, the historical call data includes at least one historical call audio and its corresponding historical call audio text, and the customer information includes the basic information of the customer and the historical business information of the customer; based on the basic information of the customer and the historical business information, the personality type is identified to obtain the customer's personality type; based on the audio feature extraction model, the audio features of the historical call audio are obtained, and based on the text feature extraction model, the text features of the historical call audio text are obtained; the friendliness of the audio features and the text features are identified by the friendliness prediction model to obtain the customer's friendliness level; the banking system receives the customer service agent service call dialed by the target customer, obtains the customer personality type and friendliness level of the target customer, and allocates the customer service agent who answers the call of the target customer based on the customer personality type and friendliness level. Through the above method, when the customer calls, the customer service agent who answers the call is allocated according to the customer personality type and friendliness level pre-obtained by the platform to improve the call quality.

[0099] See also Figure 2 , Figure 2 Another flow chart of a customer service seat allocation method provided in an embodiment of the present application, the method includes the following contents:

[0100] 101. Obtain historical call data and customer information of customers of the banking system, wherein the historical call data includes at least one historical call audio and its corresponding historical call audio text, and the customer information includes basic customer information and historical business information of the customer.

[0101] 102. Perform personality type identification based on the basic information of the customer and the historical business handling information to obtain the customer personality type of the customer.

[0102] 103. Acquire audio features of the historical call audio based on an audio feature extraction model, and acquire text features of the historical call audio text based on a text feature extraction model.

[0103] 104. Perform friendliness identification on the audio features and the text features using a friendliness prediction model to obtain the friendliness level of the customer.

[0104] 105. The bank system receives a customer service agent service call from a target customer, obtains the customer personality type and the friendliness level of the target customer, and assigns a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level.

[0105] 106. Confirm a matching customer emotion management solution based on the friendliness level.

[0106] The customer emotion management plan can be pre-set by the bank platform according to the friendliness level. It is a plan corresponding to the friendliness level. The plan may include ways to alleviate customer emotions and communication methods set according to customer personal information, etc.

[0107] When setting up a customer emotion management plan, the banking platform can set it based on the friendliness level, that is, one friendliness level corresponds to one customer emotion management plan.

[0108] When setting up a customer emotion management plan, the banking platform may also set it based on the customer's personality type and friendliness level, that is, one friendliness level combined with the customer's personality type corresponds to one customer emotion management plan.

[0109] Optionally, when setting up the customer emotion management plan, the bank platform can also set it based on the friendliness level, customer personality type and other content, which will not be elaborated here.

[0110] 107. Display the customer personality type and the customer emotion management solution on a display interface of the customer service agent answering the call.

[0111] A display can be installed at the answering position of the customer service agents of the bank to display the customer's personality type and customer emotion management plan, which is conducive to the communication between the customer service agents and the customers.

[0112] The customer service agent allocation method of the embodiment of the present application includes obtaining the historical call data and customer information of the customers of the banking system, the historical call data includes at least one historical call audio and its corresponding historical call audio text, and the customer information includes the basic information of the customer and the historical business information of the customer; based on the basic information of the customer and the historical business information, the personality type is identified to obtain the customer's personality type; based on the audio feature extraction model, the audio features of the historical call audio are obtained, and based on the text feature extraction model, the text features of the historical call audio text are obtained; the friendliness of the audio features and the text features are identified by the friendliness prediction model to obtain the customer's friendliness level; the banking system receives the customer service agent service call dialed by the target customer, obtains the customer personality type and friendliness level of the target customer, and allocates the customer service agent who answers the call of the target customer based on the customer personality type and friendliness level. Through the above method, when the customer calls, the customer service agent who answers the call is allocated according to the customer personality type and friendliness level pre-obtained by the platform to improve the call quality.

[0113] See also Figure 3 , Figure 3 This is a schematic diagram of the structure of a customer service seat allocation device provided in an embodiment of the present application. The customer service seat allocation device 200 includes the following units:

[0114] The information acquisition unit 201 is used to acquire the historical call data and customer information of the customers of the banking system. The historical call data includes at least one historical call audio and its corresponding historical call audio text. The customer information includes basic customer information and historical business information of the customer.

[0115] The personality type acquisition unit 202 is used to identify the personality type of the customer based on the basic information of the customer and the historical business information, and obtain the personality type of the customer.

[0116] The feature acquisition unit 203 is used to acquire the audio features of the historical call audio based on the audio feature extraction model, and to acquire the text features of the historical call audio text based on the text feature extraction model.

[0117] The friendliness acquisition unit 204 is used to perform friendliness identification on the audio features and the text features through a friendliness prediction model to obtain the friendliness level of the customer.

[0118] The personnel allocation unit 205 is used for the bank system to receive a customer service agent service call from a target customer, obtain the customer personality type and the friendliness level of the target customer, and allocate a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level.

[0119] Optionally, in some embodiments, the feature acquisition unit 203 may include the following subunits:

[0120] The audio segmentation subunit is used to segment the historical call audio into multiple audio segments.

[0121] The text translation subunit is used to translate each audio segment into a text segment.

[0122] The audio feature extraction subunit is used to extract the audio features of each audio segment based on the audio feature extraction model.

[0123] The text feature extraction subunit is used to extract the text features of each text segment based on the text feature extraction model.

[0124] Optionally, in some embodiments, the friendliness acquisition unit 204 may include the following subunits:

[0125] The emotion recognition subunit is used to perform emotion recognition on the audio features and text features of each audio segment and its corresponding text segment through a friendliness prediction model to obtain a corresponding emotion type, where the emotion type at least includes a negative emotion type.

[0126] The ratio calculation subunit is used to select the audio segment with negative emotion type as the negative emotion audio segment, and calculate the ratio of the negative emotion audio segment to all audio segments.

[0127] The friendliness level acquisition subunit is used to obtain a predicted friendliness level based on the ratio.

[0128] Optionally, the customer service seat allocation device 200 of the embodiment of the present application also includes other units and sub-units, which are not described in detail here.

[0129] The customer service agent allocation device 200 of the embodiment of the present application includes an information acquisition unit 201, which is used to acquire historical call data and customer information of customers of the banking system, the historical call data including at least one historical call audio and its corresponding historical call audio text, and the customer information including basic customer information and historical business information of the customer; a personality type acquisition unit 202, which is used to identify the personality type based on the basic customer information and the historical business information to obtain the customer personality type of the customer; a feature acquisition unit 203, which is used to acquire the audio features of the historical call audio based on the audio feature extraction model, and to acquire the text features of the historical call audio text based on the text feature extraction model; a friendliness acquisition unit 204, which is used to identify the friendliness of the audio features and the text features through the friendliness prediction model to obtain the friendliness level of the customer; a personnel allocation unit 205, which is used for the banking system to receive the customer service agent service call dialed by the target customer, to acquire the customer personality type and friendliness level of the target customer, and to allocate the customer service agent who answers the call of the target customer based on the customer personality type and the friendliness level. Through the above device, when a customer calls, the customer service agent who answers the call is assigned according to the customer's personality type and friendliness level obtained in advance by the platform, so as to improve the call quality.

[0130] See also Figure 4 , Figure 4 A schematic diagram of the structure of a computer device provided in an embodiment of the present application, wherein the computer device 300 includes a processor 301 having one or more processing cores, a memory 302 having one or more computer-readable storage media, and a computer program stored in the memory 302 and executable on the processor 301. The processor 301 is electrically connected to the memory 302. It can be understood by those skilled in the art that the computer device structure shown in the figure does not constitute a limitation on the computer device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange components differently.

[0131] The processor 301 is the control center of the computer device 300. It uses various interfaces and lines to connect the various parts of the entire computer device 300, executes various functions of the computer device 300 and processes data by running or loading software programs and / or modules stored in the memory 302, and calling data stored in the memory 302, thereby monitoring the computer device 300 as a whole.

[0132] In the embodiment of the present application, the processor 301 in the computer device 300 will load instructions corresponding to the processes of one or more application programs into the memory 302 according to the following steps, and the processor 301 will run the application programs stored in the memory 302 to implement various functions:

[0133] Obtain historical call data and customer information of customers in the banking system, where the historical call data includes at least one historical call audio and its corresponding historical call audio text, and the customer information includes basic customer information and historical business information of the customer;

[0134] Identify the customer's personality type based on the customer's basic information and historical business information to obtain the customer's personality type;

[0135] Acquire audio features of historical call audios based on an audio feature extraction model, and acquire text features of historical call audio texts based on a text feature extraction model;

[0136] The friendliness prediction model is used to identify the friendliness of audio features and text features to obtain the customer's friendliness level.

[0137] The banking system receives customer service agent service calls from target customers, obtains the target customer's personality type and friendliness level, and assigns customer service agents to answer calls from target customers based on the personality type and friendliness level.

[0138] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0139] Optionally, the computer device 300 further includes a data transmission module 303, which is electrically connected to the processor to implement data transmission. Those skilled in the art will appreciate that Figure 4 The computer device structure shown in the figure does not constitute a limitation on the computer device, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0140] although Figure 4 Not shown, the computer device 300 may also include a speaker, a display and other electronic structures, which will not be described in detail here.

[0141] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0142] As can be seen from the above, the computer device provided in this embodiment obtains historical call data and customer information of customers of the banking system, the historical call data includes at least one historical call audio and its corresponding historical call audio text, the customer information includes basic customer information and historical business information of the customer, personality type identification is performed based on the basic customer information and the historical business information to obtain the customer personality type of the customer, audio features of the historical call audio are obtained based on an audio feature extraction model, and text features of the historical call audio text are obtained based on a text feature extraction model, friendliness identification is performed on the audio features and text features through a friendliness prediction model to obtain the customer's friendliness level, the banking system receives a customer service agent service call dialed by a target customer, obtains the target customer's customer personality type and friendliness level, and assigns a customer service agent to answer the target customer's call based on the customer personality type and friendliness level.

[0143] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0144] To this end, an embodiment of the present application provides a computer-readable storage medium, in which multiple computer programs are stored, and the computer program can be loaded by a processor to execute the steps in any emotion recognition method provided in the embodiment of the present application. For example, the computer program can execute the following steps:

[0145] Obtain historical call data and customer information of customers in the banking system, where the historical call data includes at least one historical call audio and its corresponding historical call audio text, and the customer information includes basic customer information and historical business information of the customer;

[0146] Identify the customer's personality type based on the customer's basic information and historical business information to obtain the customer's personality type;

[0147] Acquire audio features of historical call audios based on an audio feature extraction model, and acquire text features of historical call audio texts based on a text feature extraction model;

[0148] The friendliness prediction model is used to identify the friendliness of audio features and text features to obtain the customer's friendliness level.

[0149] The banking system receives customer service agent service calls from target customers, obtains the target customer's personality type and friendliness level, and assigns customer service agents to answer calls from target customers based on the personality type and friendliness level.

[0150] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0151] The storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0152] Since the computer program stored in the storage medium can execute the steps in any one of the customer service seat allocation methods provided in the embodiments of the present application, the beneficial effects that can be achieved by any one of the customer service seat allocation methods provided in the embodiments of the present application can be achieved. Please refer to the previous embodiments for details and will not be repeated here.

[0153] The above is a detailed introduction to a customer service seat allocation method, device, computer equipment and storage medium provided in an embodiment of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A customer service seat allocation method, characterized in that: The method comprises: Acquire historical call data and customer information of customers of the banking system, wherein the historical call data includes at least one historical call audio and its corresponding historical call audio text, and the customer information includes basic customer information and historical business information of the customer; Performing personality type identification based on the basic information of the customer and the historical business handling information to obtain the customer personality type of the customer; Segmenting the historical call audio to obtain multiple audio segments; Performing text translation on each of the audio segments to obtain a text segment; Extracting audio features of each audio segment based on an audio feature extraction model; Based on the text feature extraction model, extract the text features of each text segment; The friendliness level of the customer is obtained by performing friendliness identification on the audio features and the text features through a friendliness prediction model, including: performing emotion identification on the audio features and text features of each audio segment and its corresponding text segment through the friendliness prediction model to obtain a corresponding emotion type, wherein the emotion type at least includes a negative emotion type; selecting an audio segment whose emotion type is a negative emotion type as a negative emotion audio segment, and calculating a ratio of the negative emotion audio segment to all the audio segments; and obtaining a predicted friendliness level based on the ratio; The bank system receives a customer service agent service call from a target customer, obtains the customer personality type and the friendliness level of the target customer, and assigns a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level.

2. The customer service seat allocation method according to claim 1, characterized in that: The performing friendliness identification on the audio features and the text features by using a friendliness prediction model to obtain the friendliness level of the customer further includes: By using the friendliness prediction model, the audio features and text features of each of the historical call audios and the corresponding historical call audio texts are identified to obtain the corresponding friendliness level to be predicted; Based on the level proportions of all the friendliness levels to be predicted, a predicted friendliness level is obtained.

3. The customer service seat allocation method according to claim 2, characterized in that: The bank system receives a customer service agent service call from a target customer, obtains the customer personality type and the friendliness level of the target customer, and assigns a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level, including: Selecting historical call audios corresponding to the predicted friendliness level higher than the preset level from all the predicted friendliness levels; Confirm the gender of the customer service agent of the selected historical call audio; Based on the male-female ratio of the personnel, determine the gender of the customer service agent preferred by the customer; The customer service agent who answers the incoming call is assigned based on the gender of the customer service agent preferred by the customer, the customer's personality type and the friendliness level.

4. The customer service seat allocation method according to claim 1, characterized in that: The bank system receives a customer service agent service call from a target customer, obtains the customer personality type and the friendliness level of the target customer, and assigns a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level, including: The bank system receives a customer service agent service call from a target customer, and obtains the customer personality type and the friendliness level of the target customer; Identify available customer service agents who are idle; According to the personality type of the customer, a customer service agent to be determined is selected from the idle customer service agents; A customer service agent who answers the incoming call is selected from the customer service agents to be determined according to the friendliness level.

5. The customer service seat allocation method according to claim 1, characterized in that: After the bank system receives a customer service agent service call from a target customer, obtains the customer personality type and the friendliness level of the target customer, and assigns a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level, the method further includes: Determining a matching customer emotion management solution based on the friendliness level; The customer personality type and the customer emotion management plan are displayed on a display interface of a customer service agent answering the incoming call.

6. A customer service seat allocation device, characterized in that: The device comprises: An information acquisition unit, used to acquire historical call data and customer information of customers of the banking system, wherein the historical call data includes at least one historical call audio and its corresponding historical call audio text, and the customer information includes basic customer information and historical business information of the customer; A personality type acquisition unit, configured to identify the personality type of the customer based on the basic information of the customer and the historical business information, and obtain the personality type of the customer; A feature acquisition unit, used for segmenting the historical call audio to obtain multiple audio segments; performing text translation on each of the audio segments to obtain text segments; extracting audio features of each of the audio segments based on an audio feature extraction model; and extracting text features of each of the text segments based on a text feature extraction model; The friendliness acquisition unit is used to identify the friendliness of the audio features and the text features through a friendliness prediction model to obtain the friendliness level of the customer, including: performing emotion recognition on the audio features and text features of each audio segment and its corresponding text segment through the friendliness prediction model to obtain the corresponding emotion type, wherein the emotion type at least includes a negative emotion type; selecting the audio segment whose emotion type is a negative emotion type as a negative emotion audio segment, and calculating the ratio of the negative emotion audio segment to all the audio segments; and obtaining the predicted friendliness level based on the ratio; The personnel allocation unit is used for the bank system to receive a customer service agent service call from a target customer, obtain the customer personality type and the friendliness level of the target customer, and allocate a customer service agent to answer the call from the target customer based on the customer personality type and the friendliness level.

7. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory is used to store instructions and data, and the processor is used to execute the customer service seat allocation method according to any one of claims 1 to 5.

8. A storage medium, characterized in that: The storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the customer service seat allocation method according to any one of claims 1 to 5.

Images