Customer service management device, customer service management system, customer service management method, and program

Abstract

This invention provides a customer service management device, a customer service management system, a customer service management method, and a program that can improve customer satisfaction with customer service. [Solution] The cloud server 1, which functions as a customer response means management device, includes an inquiry information identification unit 114 that identifies inquiry content information indicating the content of an inquiry from a customer and customer attribute information indicating the attributes of the customer, and an answer means estimation unit 116 that estimates an answer means suitable for the customer from among a plurality of pre-set answer means based on the identified inquiry content information and customer attribute information.

Description

Technical Field

[0001] The present disclosure relates to a customer support means management device, a customer support means management system, a customer support means management method, and a program.

Background Art

[0002] There has been proposed a customer support system used in call center operations that reduces the occurrence of claims by switching the response by a virtual operator to a response by an actual operator based on at least one of customer attribute information, information on the matters desired by the customer, and the status of an operator who responds to the customer (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] By the way, in recent years, as customer support means, in addition to direct customer support means by an operator and so-called automatic voice response customer support means by a virtual operator, customer support means utilizing chat, email, etc. are increasingly being adopted. And the direct customer support means by an operator does not necessarily lead to an early resolution of the customer's problem, and in many cases, for some customers, the so-called text-based customer support means utilizing chat, email, etc. can resolve the problem earlier. For this reason, there is a demand to improve the customer satisfaction with customer support by selecting and adopting the most suitable customer support means for each customer from a plurality of customer support means.

[0005] This disclosure is made in view of the above circumstances and aims to provide a customer service means management device, a customer service means management system, a customer service means management method, and a program that can improve customer satisfaction with customer service. [Means for solving the problem]

[0006] To achieve the above objectives, the customer response means management device relating to this disclosure is An inquiry information identification unit identifies inquiry content information that shows the content of the customer's inquiry and customer attribute information that shows the attributes of the customer, The system includes an answer means estimation unit that estimates an appropriate answer means for the customer from among a plurality of pre-set answer means based on the identified inquiry content information and customer attribute information. [Effects of the Invention]

[0007] According to this disclosure, the response means estimation unit estimates a response means appropriate to the customer from among a number of pre-set response means, based on inquiry content information indicating the content of the customer's inquiry and customer attribute information indicating the customer's attributes. This makes it easier to respond to customer inquiries with a response means appropriate to the customer, thereby improving customer satisfaction with customer service. [Brief explanation of the drawing]

[0008] [Figure 1] Schematic diagram of the customer support system according to the embodiment of this disclosure [Figure 2] Block diagram showing the hardware configuration of the customer support system according to the embodiment. [Figure 3] Block diagram showing the functional configuration of the customer support system according to the embodiment. [Figure 4] A diagram showing an example of the information stored by the inquiry history storage unit according to the embodiment. [Figure 5] A diagram illustrating the response means estimation model according to the embodiment. [Figure 6]A diagram illustrating the operation of the answer means estimation unit according to the embodiment. [Figure 7] Sequence diagram showing the operation of the customer support system according to the embodiment. [Figure 8] Sequence diagram showing the operation of the customer support system according to the embodiment. [Figure 9] Sequence diagram showing the operation of the customer support system according to the embodiment. [Figure 10] (A) is a diagram showing an example of a part of the chat screen image displayed on the display unit of a terminal device owned by a user according to the embodiment, and (B) is a diagram showing another example of a part of the chat screen image displayed on the display unit of a terminal device owned by a user according to the embodiment. [Figure 11] Sequence diagram showing the operation of the customer support system according to the embodiment. [Figure 12] Sequence diagram showing the operation of the customer support system according to the embodiment. [Figure 13] Sequence diagram showing the operation of the customer support system according to the embodiment. [Figure 14] Sequence diagram showing the operation of the customer support system according to the embodiment. [Figure 15] A flowchart showing an example of the flow of customer support management processing performed by the cloud server according to the embodiment. [Figure 16] A flowchart showing an example of the flow of customer support management processing performed by the cloud server according to the embodiment. [Figure 17] A flowchart showing an example of the flow of customer support management processing performed by the cloud server according to the embodiment. [Figure 18] A flowchart showing an example of the flow of customer support management processing performed by the cloud server according to the embodiment. [Figure 19] Block diagram showing the functional configuration of the customer support system in a modified version. [Figure 20] Block diagram showing the functional configuration of the customer support system in a modified version. [Modes for carrying out the invention]

[0009] Hereinafter, a monitoring system according to an embodiment of the present disclosure will be described with reference to the drawings. The customer support means management device according to the present embodiment includes an inquiry information specifying unit that specifies inquiry content information indicating the content of an inquiry from a customer and customer attribute information indicating the attributes of the customer, and an answer means estimating unit that estimates an answer means suitable for the customer from among a plurality of preset types of answer means based on the specified inquiry content information and customer attribute information.

[0010] The customer support system according to the present embodiment is used, for example, by a customer support service provider that responds to inquiries aimed at consulting about the purchase of home appliances, functions, and operations from customers. As shown in FIG. 1, for example, this customer support system includes a cloud server 1, a terminal device 2 possessed by a customer U1, a terminal device 3 and an IP telephone 4 installed in the customer support service provider. The terminal device 3 and the IP telephone 4 are used, for example, by an operator who answers an inquiry from the customer U1 at the customer support service provider. The customer support system also includes a SIP (Session Initiation Protocol) server 5 and a mail server 6 that manages e-mails addressed to the customer U1. The cloud server 1, the terminal devices 2 and 3, the IP telephone 4, the SIP server 5, and the mail server 6 can communicate with each other via a wide area network NW1. The wide area network NW1 is, for example, the Internet.

[0011] The terminal device 2 is a portable terminal device having a telephone function such as a smartphone, and includes an input unit 205, a display unit 204, and an IP address storage unit (not shown) that temporarily stores the IP address information of the person being called. Here, the display unit 204 is a display device such as a liquid crystal display, and the input unit 205 is a transparent touch pad arranged on top of the display unit 204, for example. Assume that customer U1 uses the telephone function of the terminal device 2 to specify the telephone number of the customer support service provider and performs a telephone inquiry start operation for an inquiry about home appliances by phone. In this case, the terminal device 2 generates call message information including the telephone number information of the customer support service provider and the IP address information of the terminal device 2, and transmits it to the SIP server 5. Then, when the terminal device 2 acquires response message information including IP address information indicating the IP address managed by the cloud server 1, the terminal device 2 extracts the IP address information included in the acquired response message information and stores it in the IP address storage unit. As a result, the terminal device 2 is in a state where voice communication using RTP (Real-time Transport Protocol) with the cloud server 1 or the IP phone 4 is possible. Then, the terminal device 2 performs voice communication using RTP with the cloud server 1 or the IP phone 4 using the IP address information stored in the IP address storage unit. Further, when the terminal device 2 acquires the operator connection voice information, which will be described later, transmitted from the cloud server 1, the terminal device 2 reproduces the operator connection voice based on the acquired operator connection voice information. At this time, the terminal device 2 reproduces a voice such as "Connecting to the operator, please wait a moment."

[0012] Furthermore, suppose that customer U1 accesses a website for chatting provided by cloud server 1 by operating the input unit 205 of terminal device 2, and then performs a chat start operation to initiate a chat. In this case, terminal device 2 generates response request information requesting a response to an inquiry via chat from cloud server 1 and sends it to cloud server 1. Also, when terminal device 2 receives the question message information sent from cloud server 1 (described later), it forms a message image based on the received question message information and displays it on the display unit 204. Here, when customer U1 performs a response message input operation by entering a response message via the input unit 205 of terminal device 2, terminal device 2 forms a response message image showing the entered response message and displays it on the display unit 204, and also generates response message information showing the said response message and sends it to cloud server 1. Furthermore, when terminal device 2 receives the inquiry reminder message information sent from cloud server 1 (described later), it forms an inquiry reminder message image based on the received inquiry reminder message information and displays it on the display unit 204. Here, when customer U1 performs an inquiry input operation by inputting inquiry content via the input unit 205 of terminal device 2, terminal device 2 forms an inquiry content message image representing the input inquiry content message and displays it on the display unit 204, and also generates inquiry content message information representing the said inquiry content message and sends it to cloud server 1. Furthermore, when terminal device 2 receives the response message information described later from cloud server 1, it forms a response message image based on the acquired response message information and displays it on the display unit 204. In addition, when terminal device 2 receives the survey information described later from cloud server 1, it forms a survey image based on the acquired survey information and displays it on the display unit 204. Here, when customer U1 performs a survey response operation by inputting the answers to the survey via the input unit 205 of terminal device 2, terminal device 2 forms a response content image representing the input answers and displays it on the display unit 204, and also generates survey result information representing the survey results and sends it to cloud server 1.

[0013] Furthermore, terminal device 2 has the function of sending emails to mail server 6, for example, via SMTP (Simple Mail Transfer Protocol). When customer U1 creates an inquiry email on terminal device 2 according to the inquiry format published on the website of the customer support service provider, terminal device 2 sends the created inquiry email to cloud server 1. This inquiry email includes customer attribute information indicating the customer's attributes and inquiry content information indicating the content of the inquiry. Furthermore, when customer U1 performs an email viewing operation via input unit 205 to view the content of the email, terminal device 2 sends email acquisition request information, including the email address information assigned to customer U1, to mail server 6, for example, via POP (Post Office Protocol) or IMAP (Internet Message Access Protocol). As a result, terminal device 2 acquires the email sent from mail server 6, forms an email viewing image for viewing the content of the acquired email, and displays it on display unit 204.

[0014] The IP phone 4 is equipped with an IP address storage unit (not shown) that temporarily stores the IP address information of the person it is calling. When the IP phone 4 receives a response request message information from the cloud server 1, which includes the IP address information of the terminal device 2 and requests a response from the customer, it extracts the IP address information of the terminal device 2 contained in the received response request message information and stores it in the IP address storage unit. This makes the IP phone 4 capable of voice communication with the terminal device 2 via RTP. Then, when the operator answers the call, the IP phone 4 uses the IP address information stored in the IP address storage unit to perform voice communication with the terminal device 2 via RTP.

[0015] SIP server 5 includes a location storage unit (not shown) that stores location registration information that associates telephone number information and IP address information of nodes with telephone functionality on the wide area network NW1. SIP server 5 stores the telephone number information and IP address information assigned to the terminal device 2 owned by customer U1 in the location storage unit, associating them with each other, and also stores the telephone number information of the customer support service provider and the IP address information managed by cloud server 1 in the location storage unit. When SIP server 5 receives outgoing message information sent from terminal device 2, which includes the telephone number information of the customer support service provider and the IP address information of terminal device 2, it refers to the location registration information stored in the location storage unit to identify the IP address information corresponding to the telephone number information contained in the outgoing message information. SIP server 5 then forwards the outgoing message information to cloud server 1, which is identified by the IP address indicated by the identified IP address information. Next, SIP server 5 obtains response message information from cloud server 1, including the IP address information of cloud server 1, and forwards it to terminal device 2, which is identified by the IP address indicated by the source IP address information of the outgoing message information contained in the obtained response message information.

[0016] The mail server 6 manages emails addressed to or originating from terminal device 2 owned by customer U1. When mail server 6 receives an email addressed to cloud server 1 sent from terminal device 2, it forwards the received email to cloud server 1. Mail server 6 also has an email storage unit (not shown) that stores emails addressed to terminal device 2 owned by customer U1 for each customer U1. When mail server 6 receives an email addressed to customer U1 sent from cloud server 1, it stores the received email in the email storage unit, associating it with customer U1's email address information. Then, when mail server 6 receives email retrieval request information sent from terminal device 2, it selects an email corresponding to the email address information contained in the retrieved email retrieval request information stored in the email storage unit and sends it to terminal device 2.

[0017] Terminal device 3 is, for example, a personal computer and includes a display unit 304 and an input unit 305. When terminal device 3 receives an inquiry notification email from customer U1 sent from cloud server 1 containing the content of the inquiry, it forms an inquiry notification image to notify the operator of the content of customer U1's inquiry and displays it on the display unit 304. After the operator confirms the inquiry content displayed in the inquiry notification image, they perform a response email creation operation via the input unit 305 to create a response email containing the answer to customer U1. Terminal device 3 then generates the response email and sends it to mail server 6, which manages emails addressed to customer U1. At this time, terminal device 3 also sends a response completion notification to cloud server 1 to notify that the response to customer U1 has been completed.

[0018] As shown in Figure 2, the cloud server 1 comprises a CPU 101, a main memory unit 102, an auxiliary memory unit 103, a communication unit 106, a timing unit 108, and a bus 109 connecting these units. The CPU 101 is, for example, a multi-core processor. The main memory unit 102 consists of volatile memory and is used as a workspace for the CPU 101. The auxiliary memory unit 103 consists of large-capacity non-volatile memory and stores programs for realizing various functions of the cloud server 1. The communication unit 106 is connected to a wide-area network NW1. The timing unit 108 has, for example, a real-time clock.

[0019] The CPU 101 reads the program stored in the auxiliary storage unit 103 into the main storage unit 102 and executes it, so as shown in Figure 3, it functions as a voice transmission / reception unit 111, a chat information transmission / reception unit 112, an email transmission / reception unit 113, an inquiry information identification unit 114, a voice recognition unit 115, an answer means estimation unit 116, a telephone answer request unit 117, a chat answer unit 118, an inquiry content notification unit 119, a teacher information selection unit 120, a model generation unit 121 and a contact notification unit 122, a question message notification unit 123 and a resolution time update unit 124. Furthermore, the auxiliary storage unit 103 shown in Figure 2 has an inquiry history storage unit 131, a resolution time estimation model storage unit 132, an answer information storage unit 133 and a teacher information storage unit 134, as shown in Figure 3. Furthermore, the main memory unit 102 includes an IP address storage unit (not shown) that temporarily stores the IP address information of the communication partner used when the voice transmission / reception unit 111 performs voice communication or when the chat information transmission / reception unit 112 sends and receives chat information.

[0020] The inquiry history storage unit 131 stores, for example as shown in Figure 4, initial inquiry means information indicating the initial inquiry means from customer U1, initial inquiry date and time information, final response means information indicating the final response means, number of days required to resolve the inquiry information indicating the number of days required to resolve the content of the inquiry, age information indicating the age of customer U1, gender information indicating the gender of customer U1, target product information indicating the product that is the subject of the inquiry, and category information indicating the category of the inquiry content, in association with customer identification information that identifies the customer. Here, age information and gender information correspond to customer attribute information indicating the customer's attributes, and target product information and category information correspond to inquiry content information indicating the content of the customer's inquiry. The customer identification information includes at least one of the following: telephone number information of customer U1's terminal device 2 or landline telephone (not shown), IP address information of terminal device 2, and email address information indicating the email address assigned to customer U1. In the example shown in Figure 4, the customer identified by customer identifier IDU[0] is a woman in her 20s who initially contacted us via chat regarding the purchase of an air conditioner. This inquiry was ultimately resolved via chat, and the resolution took less than one day. Here, the final response method information is set to NULL after the initial inquiry from customer U1 until the inquiry is resolved.

[0021] The Resolution Time Estimation Model Storage Unit 132 stores Resolution Time Estimation Model information, which shows a Resolution Time Estimation Model that calculates the number of days required to resolve a customer's inquiry, corresponding to each of several types of response methods, based on the aforementioned inquiry content information and customer attribute information. The Resolution Time Estimation Model is, for example, a feedforward neural network. In this case, the Resolution Time Estimation Model Storage Unit 132 stores information indicating the structure of the neural network and information indicating the weight coefficients in that neural network. Here, the information indicating the structure of the neural network includes information indicating the number of nodes, the number of layers, the weight coefficients corresponding to each node, and the activation function. As shown in Figure 5, this neural network has an input layer L10, a hidden layer L20, and an output layer L30. The input layer L10 is input with inquiry content information and customer attribute information corresponding to the most recently received inquiry stored in the inquiry history storage unit 131.

[0022] The hidden layer L20 consists of N (N is a positive integer) layers, each containing a predetermined number M[j] of nodes x[j,i] (1≦i≦M[j], where M[j] is a positive integer). That is, the hidden layer L20 has a structure in which each node sequence is connected to the others. Here, the output y[j,i] of each node x[j,i] is expressed by the following relationship (1).

number

[0023] Returning to Figure 3, the response information storage unit 133 stores information used when responding to customer U1's inquiry via chat. The response information storage unit 133 stores, for example, specification information and operation manual information for each of the multiple types of target products handled by the customer support service provider, as well as price information for each of the multiple types of target products.

[0024] The training information storage unit 134 stores combinations of response content information and customer attribute information stored in the inquiry history storage unit 131, and information on the number of days required for resolution, which are selected as training information for training the model on the number of days required for resolution.

[0025] When the voice transmission / reception unit 111 receives outgoing message information transmitted from the SIP server 5, it extracts the IP address information of the terminal device 2 contained in the received outgoing message information and stores it in the IP address storage unit. The voice transmission / reception unit 111 then generates response message information including the IP address information managed by the cloud server 1 and the IP address information of the terminal device 2 that sent the outgoing message information, and sends it to the SIP server 5. Subsequently, when the voice transmission / reception unit 111 receives response voice information from the terminal device 2, it notifies the voice recognition unit 115 of the received response voice information and the telephone number information and IP address information of the terminal device 2 that sent the response voice information. Furthermore, when the voice transmission / reception unit 111 receives question message information from the question message notification unit 123, it generates question voice information based on the notified question message information and sends the generated question voice information to the terminal device 2. Furthermore, when the voice transmission / reception unit 111 receives telephone operator connection message information (described later) from the telephone answer request unit 117, it generates telephone operator connection voice information based on the notified telephone operator connection message information and transmits the generated telephone operator connection voice information to the terminal device 2.

[0026] When the voice recognition unit 115 receives response voice information, telephone number information, and IP address information from the voice transmission / reception unit 111, it performs voice recognition processing to convert the received response voice information into response message information indicating the content of the response. The voice recognition unit 115 then notifies the inquiry information identification unit 114 of the response message information along with the telephone number reception notification and IP address information.

[0027] When the chat information transmission / reception unit 112 receives response request information transmitted from terminal device 2, it extracts the IP address information of terminal device 2 contained in the received response request information and stores it in the IP address storage unit. Subsequently, when the chat information transmission / reception unit 112 receives question message information from question message notification unit 123, it transmits the notified question message information to terminal device 2. Furthermore, when the chat information transmission / reception unit 112 receives response message information from terminal device 2, it notifies the inquiry information identification unit 114 of the received response message information and the IP address information of terminal device 2 that sent the response message information. In addition, when the chat information transmission / reception unit 112 receives inquiry reminder message information and response message information from chat response unit 118, it transmits the notified inquiry reminder message information and response message information to terminal device 2. Furthermore, when the chat information transmission / reception unit 112 receives inquiry content message information indicating the aforementioned inquiry content transmitted from terminal device 2, it notifies the chat response unit 118 of the received inquiry content message information. Furthermore, when the chat information transmission / reception unit 112 receives survey information from the chat response unit 118, it transmits the received survey information to the terminal device 2. Subsequently, when the chat information transmission / reception unit 112 receives survey result information transmitted from the terminal device 2, it notifies the chat response unit 118 of the acquired survey result information. In addition, when the chat information transmission / reception unit 112 receives contact information from the contact notification unit 122, including telephone number information indicating the telephone number of the contact person for the inquiry, or URL information indicating the URL of a website (Uniform Resource Locator) that publishes a format for sending inquiry emails, it generates contact notification message information including the telephone number information or URL information based on the notified contact information and transmits it to the terminal device 2.

[0028] The question message notification unit 123 sequentially generates question message information indicating a question message to customer U1 in order to identify the attributes of customer U1 and the content of the inquiry, and notifies the voice transmission / reception unit 111 or the chat information transmission / reception unit 112. Furthermore, once the generation of all question message information for identifying the attributes of customer U1 and the content of the inquiry is complete, the question message notification unit 123 notifies the answer means estimation unit 116 of the question completion notification information.

[0029] When the email sending / receiving unit 113 obtains the aforementioned inquiry email from the email server 6, it notifies the inquiry information identification unit 114 of the inquiry information indicating the content of the obtained inquiry email, as well as the email address information and IP address information of the terminal device 2 that sent the inquiry email. Furthermore, when the email sending / receiving unit 113 receives contact information including the aforementioned telephone number information or URL information from the contact notification unit 122, it generates a contact notification email including the telephone number information or URL information based on the notified contact information and sends it to the terminal device 2.

[0030] The inquiry information identification unit 114 identifies the aforementioned inquiry content information and customer attribute information based on the response message information notified from the speech recognition unit 115, the response message information notified from the chat information transmission / reception unit 112, or the inquiry information notified from the email transmission / reception unit 113. The inquiry information identification unit 114 then stores the identified inquiry content information and customer attribute information in the inquiry history storage unit 131, associating them with customer identification information that includes at least one of telephone number information, IP address information, and email address information. Furthermore, if the inquiry information identification unit 114 receives inquiry information from the email transmission / reception unit 113, it notifies the response means estimation unit 116 of the question completion notification information.

[0031] The response means estimation unit 116 estimates a response means suitable for the customer from among a plurality of pre-set response means, based on the inquiry content information and customer attribute information identified by the inquiry information identification unit 114. Specifically, when the aforementioned question completion notification information is received, the response means estimation unit 116 estimates the number of days required to resolve each of the plurality of response means corresponding to the newly stored inquiry content information and customer attribute information, using the resolution time estimation model indicated by the resolution time estimation model information stored in the resolution time estimation model storage unit 132, based on the inquiry content information and customer attribute information newly stored in the inquiry history storage unit 131. Here, the response means estimation unit 116 first identifies the inquiry content information and customer attribute information newly stored in the inquiry history storage unit 131. Then, based on the identified inquiry content information and customer attribute information, the response means estimation unit 116 generates a vector Vin consisting of, for example, elements representing the inquiry content, i.e., the target equipment and category, and elements representing the customer's age and gender, as shown in Figure 6. In the example shown in Figure 6, the vector Vin is set so that the element IN[0] representing the target device is identified as "air conditioner" by identification information IDE[0], the element IN[1] representing the category is identified as "purchase consultation" by identification information IDC[0], the element IN[2] representing the customer's age is identified as "20s" by identification information IDA[1], and the element IN[3] representing gender is identified as "male" by identification information IDG[0]. The response means estimation unit 116 then inputs the vector Vin into the aforementioned resolution time estimation model and calculates the aforementioned expected value vector Vout, which has the resolution time corresponding to each of the pre-set multiple types of response means as its elements. The response means estimation unit 116 then estimates the response means corresponding to the element with the smallest expected value among the multiple elements of the calculated expected value vector Vout as the appropriate response means for customer U1. In the example shown in Figure 6, the three elements constituting the expected value vector Vout correspond to chat response means, telephone response means, and email response means, respectively. Then, if the response means estimation unit 116 estimates chat response means as a suitable response means for customer U1, it notifies the chat response unit 118 of the newly stored inquiry content information and customer identification information in the inquiry history storage unit 131.Furthermore, if the response means estimation unit 116 estimates telephone response as a suitable response means for customer U1, it notifies the telephone response request unit 117 and the contact notification unit 122 of the customer identification information newly stored in the inquiry history storage unit 131. At this time, the response means estimation unit 116 notifies the response means information indicating the estimated telephone response means and the customer identification information of the day required for resolution update unit 124. Moreover, if the response means estimation unit 116 estimates email response as a suitable response means for customer U1, it notifies the inquiry content notification unit 119 of the inquiry content information and customer identification information newly stored in the inquiry history storage unit 131. At this time, the response means estimation unit 116 notifies the response means information indicating the estimated email response means and the customer identification information of the day required for resolution update unit 124.

[0032] Returning to Figure 3, when the telephone response request unit 117 receives customer identification information from the response means estimation unit 116, it generates a response request message information including the IP address information contained in the customer identification information and sends it to the IP telephone 4. The telephone response request unit 117 also notifies the voice transmission / reception unit 111 of a telephone operator connection message information indicating to customer U1 that the call will be connected to an operator.

[0033] When the chat response unit 118 obtains inquiry content information and customer identification information from the response means estimation unit 116, it generates question message information to ask for more detailed information about the target device of the inquiry, such as the model number of the target device, and notifies the chat information transmission / reception unit 112. Subsequently, when the chat response unit 118 receives response message information from the chat information transmission / reception unit 112, it identifies the response content indicated by the response message information. Then, using the RAG (Retrieval-Augmented Generation) system, the chat response unit 118 selects the information necessary for the response from the information stored in the response information storage unit 133 based on the inquiry content information notified from the response means estimation unit 116 and the identified response content. Furthermore, the chat response unit 118 generates inquiry reminder message information and notifies the chat information transmission / reception unit 112. Subsequently, when the chat information transmission / reception unit 112 receives inquiry content message information, it uses a large language model (LLM) to estimate the response content based on the selected information and the notified inquiry content message information. The chat response unit 118 then notifies the chat information transmission / reception unit 112 of response message information indicating the estimated response content. After generating the response message information, the chat response unit 118 also generates survey information and notifies the chat information transmission / reception unit 112 of it. Subsequently, when the chat response unit 118 receives survey result information from the chat information transmission / reception unit 112, it determines whether or not the inquiry has been resolved based on the notified survey result information. If the chat response unit 118 determines that the inquiry has been resolved, it notifies the customer identification information of the corresponding customer U1 and the response means information indicating the chat response means to the resolution time update unit 124.

[0034] When the inquiry content notification unit 119 receives inquiry content information and customer identification information from the response means estimation unit 116, it generates an inquiry content notification email to convey the inquiry content to the operator based on the notified inquiry content information and customer identification information, and sends it to the terminal device 3.

[0035] When the contact information notification unit 122 receives customer identification information from the response means estimation unit 116, it notifies the chat information sending / receiving unit 112 and the email sending / receiving unit 113 of contact information including telephone number information indicating the telephone number of the customer support service provider to which the inquiry is directed, or URL information of the website where the inquiry format is published.

[0036] When the Resolution Time Update Unit 124 receives customer identification information and response means information indicating either telephone or email response from the Response Means Estimation Unit 116, it determines whether a predetermined number of days has elapsed since the notification. Here, the standard number of days is set to, for example, one day. If the Resolution Time Update Unit 124 determines that a predetermined number of days has elapsed since the notification, it calculates the number of days that have elapsed from the initial inquiry date and time to the time of determination. The Resolution Time Update Unit 124 then sets the final response means information stored in the inquiry history storage unit 131 to the telephone or email response means indicated by the notified response means information, and sets the Resolution Time Update Unit 124 to the calculated number of days that have elapsed since the notification. Then, the Resolution Time Update Unit 124 sets the final response means information stored in the inquiry history storage unit 131 to the chat response means indicated by the notified response means information, and also sets the Resolution Time Information to the calculated number of elapsed days.

[0037] The training information selection unit 120 selects combinations of response content information and customer attribute information stored in the inquiry history storage unit 131 and the number of days required for resolution information that can be used as training information for training the aforementioned number of days required for resolution model. Here, the training information selection unit 120 excludes combinations of response content information and customer attribute information stored in the inquiry history storage unit 131 and the number of days required for resolution information in which the final response method information is set to NULL, and combinations in which the number of days required for resolution exceeds a predetermined upper limit. The training information selection unit 120 then stores the selected combinations in the training information storage unit 134.

[0038] The model generation unit 121 generates new estimated solution time model information using the updated combinations of the response content information and customer attribute information stored in the training information storage unit 134 and the solution time information, whenever the combinations are updated. It then updates the estimated solution time model information stored in the solution time estimate model storage unit 132 with the generated estimated solution time model information. Specifically, the model generation unit 121 first calculates the average number of solution time corresponding to each of the multiple types of response methods corresponding to each of the multiple types of response content information and customer attribute information combinations after the update. Here, for response methods for which there is no actual solution time data, the average number of solution time is set to a predetermined upper limit. Then, the expected value corresponding to each of the response methods that constitute the aforementioned expected value vector is set to the calculated average number of solution time. The model generation unit 121 then calculates the error between the expected value vector calculated using the pre-update solution time estimation model and the set expected value vector, and determines the weight coefficients of the new solution time estimation model using the error propagation method (backpropagation) based on the calculated error. The model generation unit 121 then updates the solution time estimation model information stored in the solution time estimation model storage unit 132 with the solution time estimation model information indicating the determined weight coefficients.

[0039] Next, the operation of the customer support system according to this embodiment will be explained with reference to Figures 7 to 14. First, as shown in Figure 7, suppose customer U1 uses the telephone function of terminal device 2 to initiate a telephone inquiry for a home appliance inquiry by specifying the telephone number of the customer support service provider. In this case, terminal device 2 generates outgoing message information including the telephone number information of the customer support service provider and the IP address information of terminal device 2 (step S1), and the generated outgoing message information is sent from terminal device 2 to SIP server 5 (step S2). Meanwhile, when SIP server 5 receives the outgoing message information, it refers to the location registration information stored in the location storage unit and identifies the IP address information corresponding to the telephone number information of the customer support service provider included in the outgoing message information (step S3). Next, the outgoing message information is transferred from SIP server 5 to cloud server 1, which is identified by the IP address indicated by the IP address information identified by SIP server 5 (step S4). Meanwhile, when cloud server 1 receives the outgoing message information, it extracts the IP address information of terminal device 2 included in the received outgoing message information and stores it in the IP address storage unit (step S5). Next, Cloud Server 1 generates a response message information that includes the IP address information managed by Cloud Server 1 and the IP address information of Terminal Device 2, the source of the outgoing message information (Step S6), and the generated response message information is sent from Cloud Server 1 to SIP Server 5 (Step S7). Meanwhile, SIP Server 5, upon receiving the response message information, identifies the IP address information of Terminal Device 2, the source of the outgoing message information, which is included in the response message information (Step S8). Next, the response message information is forwarded from SIP Server 5 to Terminal Device 2, which is identified by the IP address indicated by the IP address information identified (Step S9). Meanwhile, Terminal Device 2, upon receiving the response message information, extracts the IP address information managed by Cloud Server 1 included in the received response message information and stores it in its IP address storage unit (Step S10). This enables voice communication via RTP between Terminal Device 2 and Cloud Server 1.

[0040] Next, Cloud Server 1 generates question message information indicating a question message to Customer U1 in order to identify Customer U1's attributes and inquiry content, and generates question audio information based on the generated question message information (Step S11). Subsequently, the generated question audio information is sent from Cloud Server 1 to Terminal Device 2 (Step S12). Meanwhile, when Terminal Device 2 receives the question audio information, it plays the question audio based on the received question audio information (Step S13). Now, suppose Customer U1 inputs an answer voice using the telephone function of Terminal Device 2. In this case, Terminal Device 2 generates answer audio information indicating the input voice (Step S14), and the generated answer audio information is sent from Terminal Device 2 to Cloud Server 1 (Step S15). Meanwhile, when Cloud Server 1 receives the answer audio information, it performs speech recognition processing to convert the notified answer audio information into answer message information indicating the answer content. Furthermore, Cloud Server 1 identifies the aforementioned inquiry content information and customer attribute information based on the answer message information. Then, the cloud server 1 stores the identified inquiry content information and customer attribute information in the inquiry history storage unit 131, associating it with customer identification information that includes at least one of telephone number information, IP address information, and email address information (step S16). Next, the series of processes from steps S11 to S16 are repeated, so that the cloud server 1 acquires the answer content information and customer attribute information necessary for estimating the answer means. Then, the cloud server 1 determines that it has finished asking questions to acquire the answer content information and customer attribute information necessary for estimating the answer means, as shown in Figure 8 (step S17). In this case, the cloud server 1 performs a process to estimate the answer means suitable for customer U1 based on the acquired answer content information and customer attribute information. Here, the cloud server 1 estimates that the answer means suitable for customer U1 is a telephone answer means (step S18). In this case, the cloud server 1 generates the aforementioned telephone operator connection voice information (step S19), and the generated telephone operator connection voice information is transmitted from the cloud server 1 to the terminal device 2 (step S20).Meanwhile, when terminal device 2 acquires telephone operator connection voice information, it plays the telephone operator connection voice based on the acquired telephone operator connection voice information (step S21). Next, cloud server 1 generates the aforementioned response request message information (step S22), and the generated response request message information is sent from cloud server 1 to IP phone 4 (step S23). Meanwhile, when IP phone 4 acquires the response request message information, it extracts the IP address information of terminal device 2 contained in the acquired response request message information and stores it in the IP address storage unit (step S24). As a result, IP phone 4 becomes capable of voice communication with terminal device 2 via RTP. Then, IP phone 4 rings (step S25). When the operator answers the call on IP phone 4, IP phone 4 uses the IP address information stored in the IP address storage unit to send and receive voice call information with terminal device 2 (step S26), thereby making a telephone call with customer U1.

[0041] Subsequently, as shown in Figure 9, when the cloud server 1 determines that a predetermined number of days has elapsed since the start of the telephone call (step S27), it updates the final response means information stored in the inquiry history storage unit 131 by setting it as the telephone response means (step S28). The cloud server 1 also calculates the number of days elapsed from the initial inquiry date and time to the point when it determines that the predetermined number of days has elapsed, and updates the number of days required to resolve information stored in the inquiry history storage unit 131 by setting it to the calculated number of elapsed days (step S29). At this time, the cloud server 1 updates the combination of the response content information and customer attribute information stored in the training information storage unit 134 and the number of days required to resolve information based on the updated response content information and customer attribute information and the number of days required to resolve information stored in the inquiry history storage unit 131. Next, the cloud server 1 generates new estimated solution time model information using the combination of the updated response content information and customer attribute information stored in the training information storage unit 134 and the estimated solution time information, and updates the estimated solution time model information stored in the estimated solution time model storage unit 132 with the generated estimated solution time model information (step S30).

[0042] Furthermore, suppose that customer U1 accesses a website for chatting provided by cloud server 1 by operating the input unit 205 of terminal device 2, and then performs a chat start operation to initiate a chat. In this case, terminal device 2 generates response request information requesting a response to an inquiry via chat from cloud server 1 (step S31), and the generated response request information is sent from terminal device 2 to cloud server 1 (step S32). Meanwhile, cloud server 1, upon receiving the response request information, generates question message information indicating a question message to customer U1 in order to identify customer U1's attributes and the content of the inquiry (step S33). Subsequently, the generated question message information is sent from cloud server 1 to terminal device 2 (step S34). Meanwhile, terminal device 2, upon receiving the question message information, forms a message image based on the acquired question message information and displays it on the display unit 204 (step S35). Here, when customer U1 performs a response message input operation by inputting a response message via the input unit 205 of terminal device 2, terminal device 2 forms a response message image representing the input response message and displays it on the display unit 204 (step S36). Terminal device 2 generates response message information representing the input response message (step S37). At this time, the generated response message information is transmitted from terminal device 2 to cloud server 1 (step S38). Meanwhile, when cloud server 1 obtains the response message information, it identifies the aforementioned inquiry content information and customer attribute information based on the acquired response message information. Then, cloud server 1 stores the identified inquiry content information and customer attribute information in the inquiry history storage unit 131 in association with customer identification information including at least one of telephone number information, IP address information, and email address information (step S39). Subsequently, the series of processes from steps S33 to S39 are repeated, so that cloud server 1 obtains the response content information and customer attribute information necessary for estimating the response means.

[0043] At this time, the terminal device 2 displays a chat screen image GA1 on the display unit 204, for example, as shown in Figure 10(A). The chat screen image GA1 includes question message images M111, M112, and M113, and answer message images M121, M122, and M123, which represent the customer U1's answer messages to the questions.

[0044] Next, let's assume that the cloud server 1 determines that it has finished asking questions to obtain the answer content information and customer attribute information necessary for estimating the answer means, as shown in Figure 11 (step S40). In this case, the cloud server 1 performs a process to estimate the answer means suitable for customer U1 based on the acquired answer content information and customer attribute information. Now, let's assume that the cloud server 1 estimates that the answer means suitable for customer U1 is the chat answer means (step S41). In this case, the cloud server 1 generates question message information to ask for more detailed information about the target device of the inquiry necessary for estimating the answer content (step S42), and the generated question message information is sent from the cloud server 1 to the terminal device 2 (step S43). Meanwhile, when the terminal device 2 obtains the question message information, it forms a question message image based on the obtained question message information and displays it on the display unit 204 (step S44). Here, when customer U1 performs a response message input operation to input a response message via the input unit 205 of the terminal device 2, the terminal device 2 forms a response message image representing the input response message and displays it on the display unit 204 (step S45), and also generates response message information representing the said response message (step S46).

[0045] At this time, the terminal device 2 displays a chat screen image GA2 on the display unit 204, for example, as shown in Figure 10(B). The chat screen image GA2 includes question message images M111, M112, M113, and M211, and answer message images M121, M122, M123, and M221, which represent the customer U1's answer messages to the questions. In the example shown in Figure 10(B), a message image M212 is included that asks customer U1 whether it is necessary to make an inquiry by phone before displaying the inquiry reminder message image M212, which will be described later.

[0046] Returning to Figure 11, the generated response message information is then sent from terminal device 2 to cloud server 1 (step S47). Meanwhile, as shown in Figure 12, when cloud server 1 receives the response message information, it identifies more detailed response information regarding the target device of the inquiry, which is necessary to estimate the response content, based on the received response message information (step S48). Subsequently, cloud server 1 generates inquiry reminder message information that indicates an inquiry reminder message requesting specific details of the inquiry from customer U1 (step S49), and the generated inquiry reminder message information is sent from cloud server 1 to terminal device 2 (step S50). Meanwhile, when terminal device 2 receives the inquiry reminder message information, it forms an inquiry reminder message image based on the received inquiry reminder message information and displays it on display unit 204 (step S51).

[0047] Here, when customer U1 performs an inquiry input operation by inputting the inquiry content via the input unit 205 of terminal device 2, terminal device 2 forms an inquiry content message image showing the input inquiry content and displays it on the display unit 204 (step S52), and generates inquiry content message information showing the said inquiry content (step S53). Subsequently, the generated inquiry content message information is transmitted from terminal device 2 to cloud server 1 (step S54). Next, cloud server 1 selects the information necessary for the answer from the information stored in the answer information storage unit 133 based on the inquiry content information and the answer content identified in step S48. Then, cloud server 1 estimates the answer content based on the selected information and the acquired inquiry content message information (step S55). Subsequently, answer message information showing the estimated answer content is transmitted from cloud server 1 to terminal device 2 (step S56). Meanwhile, when terminal device 2 acquires the answer message information, it forms an answer message image based on the acquired answer message information and displays it on the display unit 204 (step S57).

[0048] Furthermore, after generating response message information, the cloud server 1 generates survey information (step S58). The generated survey information is then transmitted from the cloud server 1 to the terminal device 2 (step S59). Meanwhile, when the terminal device 2 acquires the survey information, it forms a survey image based on the acquired survey information and displays it on the display unit 204 (step S60). Now, suppose customer U1 performs a survey response operation by inputting answers to the survey via the input unit 205 of the terminal device 2. In this case, the terminal device 2 forms a survey response content image showing the input answers and displays it on the display unit 204 (step S61), and generates survey result information showing the survey results, as shown in Figure 13 (step S62). The generated survey result information is then transmitted from the terminal device 2 to the cloud server 1 (step S63).

[0049] On the other hand, suppose the cloud server 1 acquires the survey results information and determines that the inquiry has been resolved based on the acquired survey results information (step S64). In this case, the cloud server 1 updates the final response means information stored in the inquiry history storage unit 131 by setting it to the chat response means (step S65). Subsequently, the cloud server 1 calculates the number of days elapsed from the initial inquiry date and time to the time when it was determined that the inquiry had been resolved, and updates the number of days required for resolution information stored in the inquiry history storage unit 131 by setting it to the calculated number of elapsed days (step S66). At this time, the cloud server 1 updates the combination of the response content information and customer attribute information stored in the training information storage unit 134 and the number of days required for resolution information based on the updated response content information and customer attribute information and the number of days required for resolution information stored in the inquiry history storage unit 131. Next, the cloud server 1 generates new solution time estimation model information using the updated response content information and customer attribute information stored in the training information storage unit 134, and the solution time estimation model information, and updates the solution time estimation model information stored in the solution time estimation model storage unit 132 with the generated solution time estimation model information (step S67).

[0050] Furthermore, suppose customer U1 performs an inquiry input operation via the input unit 205 of terminal device 2, entering the inquiry content in accordance with the inquiry format published on the website by the customer support service provider. In this case, terminal device 2 generates an inquiry email based on the entered inquiry content (step S68), and the generated inquiry email is sent from terminal device 2 to mail server 6 (step S69). The inquiry email sent to mail server 6 is then forwarded from mail server 6 to cloud server 1 (step S70). Meanwhile, when cloud server 1 receives the inquiry email, it identifies the response content information and customer attribute information necessary for estimating the response method based on the content of the received inquiry email. Then, cloud server 1 stores the identified inquiry content information and customer attribute information in the inquiry history storage unit 131, associating it with customer identification information including at least one of telephone number information, IP address information, and email address information (step S71). Next, cloud server 1 performs processing to estimate a response method suitable for customer U1 based on the response content information and customer attribute information newly stored in the inquiry history storage unit 131. Here, suppose that the cloud server 1 estimates that the appropriate response method for customer U1 is email response (step S72). In this case, the cloud server 1 generates an inquiry content notification email to convey the inquiry content to the operator based on the inquiry content information and customer identification information (step S73), and the generated inquiry content notification email is sent from the cloud server 1 to the terminal device 3 (step S74).

[0051] Meanwhile, when terminal device 3 receives an inquiry notification email, it forms the aforementioned inquiry notification image and displays it on the display unit 304 (step S75). At this point, the operator confirms the inquiry content displayed in the inquiry notification image and then, via the input unit 305 of terminal device 3, performs a response email creation operation to create a response email containing the answer to customer U1, as shown in Figure 14, and terminal device 3 generates the response email (step S76). Then, when the operator performs a mail sending operation to send the response email to customer U1, the response email is sent from terminal device 3 to mail server 6 (step S77). At this time, response completion notification information notifying that the response to customer U1 has been completed is sent from terminal device 3 to cloud server 1 (step S78). Subsequently, when customer U1 performs a mail viewing operation via the input unit 205 of terminal device 2 to view the contents of the email, the aforementioned mail acquisition request information is sent from terminal device 2 to mail server 6 (step S79). As a result, terminal device 2 acquires emails sent from mail server 6, forms an email viewing image for viewing the content of the acquired emails, and displays it on display unit 204. Meanwhile, when mail server 6 acquires email acquisition request information, a reply email is sent from mail server 6 to terminal device 2 (step S80). Meanwhile, when terminal device 2 acquires the reply email, it forms an email viewing image for viewing the content of the acquired reply email and displays it on display unit 204 (step S81).

[0052] Subsequently, after obtaining the response completion notification information, if the cloud server 1 determines that a predetermined number of days has elapsed (step S82), it updates the final response method information stored in the inquiry history storage unit 131 by setting it to the email response method (step S83). The cloud server 1 also calculates the number of days elapsed from the initial inquiry date and time to the point in time when it determines that the number of days elapsed has elapsed, and updates the number of days required to resolve information stored in the inquiry history storage unit 131 by setting it to the calculated number of elapsed days (step S84). At this time, the cloud server 1 updates the combination of the response content information and customer attribute information stored in the training information storage unit 134 and the number of days required to resolve information based on the updated response content information and customer attribute information and the number of days required to resolve information stored in the inquiry history storage unit 131. Next, the cloud server 1 generates new solution time estimation model information using the updated response content information and customer attribute information stored in the training information storage unit 134, along with the solution time estimation model information. It then updates the solution time estimation model information stored in the solution time estimation model storage unit 132 with the generated solution time estimation model information (step S85).

[0053] Next, the customer service management process executed by the cloud server 1 according to this embodiment will be explained with reference to Figures 15 to 18. This customer service management process is started, for example, when a program for executing the customer service management process is started on the cloud server 1. First, as shown in Figure 15, the voice transmission / reception unit 111 determines whether or not it has acquired the outgoing message information transmitted from the terminal device 2 (step S101). If the voice transmission / reception unit 111 determines that it has not acquired the outgoing message information (step S101: No), the process in step S117, which will be described later, is executed. On the other hand, if the voice transmission / reception unit 111 determines that it has acquired the outgoing message information (step S101: Yes), it extracts the IP address information of the terminal device 2 included in the acquired outgoing message information and stores it in the IP address storage unit (step S102). Next, the voice transmission / reception unit 111 generates response message information including the IP address information managed by the cloud server 1 and the IP address information of the terminal device 2 that sent the outgoing message information, and sends it to the SIP server 5 (step S103).

[0054] Next, the question message notification unit 123 generates the aforementioned question message information and notifies the voice transmission / reception unit 111, and the voice transmission / reception unit 111 generates question voice information based on the question message information notified by the question message notification unit 123. Then, the voice transmission / reception unit 111 transmits the generated question voice information to the terminal device 2 (step S104). After that, the voice transmission / reception unit 111 determines whether or not it has acquired the answer voice information transmitted from the terminal device 2 (step S105). If the voice transmission / reception unit 111 determines that it was not able to acquire the answer voice information within the predetermined waiting time (step S105: No), the process in step S101 is executed again. On the other hand, if the voice transmission / reception unit 111 determines that it has acquired the answer voice information (step S105: Yes), it notifies the voice recognition unit 115 of the acquired answer voice information and the telephone number information and IP address information of the terminal device 2 that sent the answer voice information. The speech recognition unit 115 then performs speech recognition processing to convert the notified response voice information into response message information indicating the content of the response, and notifies the inquiry information identification unit 114 of the response message information along with the telephone number information and IP address information. The inquiry information identification unit 114 then identifies the aforementioned inquiry content information and customer attribute information based on the response message information notified by the speech recognition unit 115, and stores the identified inquiry content information and customer attribute information in the inquiry history storage unit 131 in association with customer identification information including at least one of the telephone number information, IP address information, and email address information (step S106). Next, the question message notification unit 123 determines whether all questions have been completed, that is, whether the generation of all question message information has been completed (step S107). If the question message notification unit 123 determines that there is question message information that has not yet been generated (step S107: No), the process in step S104 is executed again. On the other hand, when the question message notification unit 123 determines that the generation of all question message information has been completed (step S107: Yes), it notifies the answer means estimation unit 116 of the question completion notification information.Then, when the question completion notification information is received, the response means estimation unit 116 performs a process to estimate a response means suitable for the customer from among a plurality of pre-set response means, based on the newly stored inquiry content information and customer attribute information, each time new inquiry content information and customer attribute information are stored in the inquiry history storage unit 131. Here, the response means estimation unit 116 determines whether or not the estimated response means is a telephone response means (step S108). If the response means estimation unit 116 determines that it has estimated a response means other than a telephone response means (step S108: No), the process in step S123 described below is executed.

[0055] On the other hand, if the response means estimation unit 116 determines that it has estimated the telephone response means (step S108: Yes), it notifies the telephone response request unit 117 and the contact notification unit 122 of the customer identification information newly stored in the inquiry history storage unit 131, and also notifies the response means information indicating the estimated telephone response means and the customer identification information of the customer identification information of the response means update unit 124. The contact notification unit 122 then determines whether customer U1 has made an inquiry by "telephone" (step S109). If the contact notification unit 122 determines that customer U1 has made an inquiry by "chat" or "email" (step S109: No), it uses the customer identification information notified by the response means estimation unit 116 to generate contact information including telephone number information indicating the telephone number of the customer support service provider, or URL information of the website where the inquiry format is published, and notifies the chat information sending / receiving unit 112 and the email sending / receiving unit 113. Then, if customer U1 makes an inquiry via "chat" or "email," the chat information transmission / reception unit 112 generates contact notification information including the notified contact information and sends it to the terminal device 2. Also, if customer U1 makes an inquiry via "chat" or "email," the email transmission / reception unit 113 generates a contact notification email including the notified contact information and sends it to the email server 6 (step S110). Subsequently, the processing in step S113 is executed.

[0056] On the other hand, if the contact notification unit 122 determines that customer U1 is making an inquiry by "telephone" (step S109: Yes), the telephone response request unit 117 notifies the voice transmission / reception unit 111 of telephone operator connection message information indicating that the call will be connected to an operator for customer U1. The voice transmission / reception unit 111 then generates telephone operator connection voice information based on the notified telephone operator connection message information and transmits the generated telephone operator connection voice information to the terminal device 2 (step S111). Subsequently, the telephone response request unit 117 generates response request message information including IP address information included in customer U1's customer identification information and transmits it to the IP telephone 4 (step S112).

[0057] Next, the Resolution Time Update Unit 124 determines whether a predetermined number of days has elapsed since the customer identification information and the telephone response information were notified from the Response Method Estimation Unit 116 (Step S113). If the Resolution Time Update Unit 124 determines that the standard number of days has not yet elapsed (Step S113: No), the process in Step S101 is executed again. On the other hand, if the Resolution Time Update Unit 124 determines that the standard number of days has elapsed since the notification (Step S113: Yes), it calculates the number of days elapsed from the initial inquiry date and time to the determined time. Then, the Resolution Time Update Unit 124 updates the final response information stored in the inquiry history storage unit 131 by setting it to the telephone response information indicated by the notified response method information (Step S114). Next, the Resolution Time Estimate Update Unit 124 updates the Resolution Time Estimate Information by setting the Resolution Time Estimate Information stored in the Inquiry History Storage Unit 131 to the calculated number of elapsed days (Step S115). Subsequently, each time the combination of the response content information and customer attribute information stored in the Training Information Storage Unit 134 and the Resolution Time Estimate Information is updated, the Model Generation Unit 121 generates new Resolution Time Estimate Model Information using the updated combination, and updates the Resolution Time Estimate Model Information stored in the Resolution Time Estimate Model Storage Unit 132 with the generated Resolution Time Estimate Model Information (Step S116). After that, the process in Step S101 is executed again.

[0058] Furthermore, if the voice transmission / reception unit 111 determines in step S101 that it has not acquired the outgoing message information (step S101: No), the chat information transmission / reception unit 112 determines whether or not it has acquired the response request information transmitted from the terminal device 2 (step S117). If the chat information transmission / reception unit 112 determines that it has not acquired the response request information (step S117: No), the process in step S136 described later is executed. On the other hand, if the chat information transmission / reception unit 112 determines that it has acquired the response request information (step S117: Yes), it extracts the IP address information of the terminal device 2 included in the acquired response request information and stores it in the IP address storage unit (step S118). Next, the question message notification unit 123 generates the aforementioned question message information and notifies the chat information transmission / reception unit 112, and the chat information transmission / reception unit 112 transmits the question message information notified by the question message notification unit 123 to the terminal device 2 (step S119). Subsequently, the chat information transmission / reception unit 112 determines whether or not it has received the response message information transmitted from the terminal device 2 (step S120). If the chat information transmission / reception unit 112 determines that it was unable to receive the response message information within a preset waiting time (step S120: No), the process in step S117 is executed again. On the other hand, if the chat information transmission / reception unit 112 determines that it has received the response message information (step S120: Yes), it notifies the inquiry information identification unit 114 of the received response message information along with the IP address information. The inquiry information identification unit 114 then identifies the aforementioned inquiry content information and customer attribute information based on the response message information notified by the chat information transmission / reception unit 112, and stores the identified inquiry content information and customer attribute information in the inquiry history storage unit 131 in association with customer identification information including at least one of telephone number information, IP address information, and email address information (step S121). Next, the question message notification unit 123 determines whether or not all questions have been completed (step S122). If the question message notification unit 123 determines that not all questions have been completed yet (step S122: No), the process in step S119 is executed again.On the other hand, when the question message notification unit 123 determines that all questions have been completed (step S122: Yes), it notifies the answer means estimation unit 116 of the question completion notification information. When the answer means estimation unit 116 receives the question completion notification information, it performs a process to estimate an appropriate answer means for the customer from among a number of pre-set answer means, based on the newly stored inquiry content information and customer attribute information, each time new inquiry content information and customer attribute information are stored in the inquiry history storage unit 131. Here, the answer means estimation unit 116 determines whether the estimated answer means is a chat answer means or not (step S123). Here, when the answer means estimation unit 116 determines that it has estimated an answer means other than a chat answer means (step S123: No), the process in step S139 described below is executed.

[0059] On the other hand, if the response means estimation unit 116 determines that it has estimated a chat response means (step S123: Yes), it notifies the chat response unit 118 of the newly stored inquiry content information and customer identification information in the inquiry history storage unit 131, and also notifies the response means information indicating the estimated chat response means and customer identification information of the day required for resolution update unit 124. Here, when the chat response unit 118 obtains the inquiry content information and customer identification information from the response means estimation unit 116, it generates the aforementioned question message information and notifies the chat information transmission / reception unit 112, and the chat information transmission / reception unit 112 transmits the notified question message information to the terminal device 2 (step S124). Subsequently, the chat information transmission / reception unit 112 determines whether or not it has obtained the response message information transmitted from the terminal device 2 (step S125). Here, if the chat information transmission / reception unit 112 determines that it was not able to obtain the response message information within the predetermined waiting time (step S125: No), the process in step S117 is executed again. Meanwhile, when the chat information transmission / reception unit 112 determines that it has acquired response message information (step S125: Yes), it notifies the chat response unit 118 of the acquired response message information. Then, when the chat response unit 118 acquires the response message information, it identifies the content of the response indicated by the acquired response message information (step S126).

[0060] Subsequently, the chat response unit 118 generates inquiry reminder message information and notifies the chat information transmission / reception unit 112, which then transmits the notified inquiry reminder message information to the terminal device 2 (step S127). Next, the chat information transmission / reception unit 112 determines whether or not it has received the inquiry content message information transmitted from the terminal device 2 (step S128). If the chat information transmission / reception unit 112 determines that it was not able to receive the inquiry content message information within the pre-set waiting time (step S128: No), the process in step S117 is executed again. On the other hand, if the chat information transmission / reception unit 112 determines that it has received the inquiry content message information (step S128: Yes), it notifies the chat response unit 118 of the received inquiry content message information. Then, the chat response unit 118 estimates the response content based on the notified inquiry content message information, generates response message information indicating the estimated response content, and notifies the chat information transmission / reception unit 112, which then transmits the notified response message information to the terminal device 2 (step S129).

[0061] Next, the chat response unit 118 generates survey information and notifies the chat information transmission / reception unit 112, which then transmits the notified survey information to the terminal device 2 (step S130). Subsequently, the chat information transmission / reception unit 112 acquires the survey result information transmitted from the terminal device 2 (step S131) ​​and notifies the chat response unit 118 of the acquired survey result information. Next, as shown in Figure 17, the chat response unit 118 determines whether or not the inquiry has been resolved based on the notified survey result information (step S132). If the chat response unit 118 determines that the inquiry has not been resolved (step S132: No), the process in step S117 is executed again. On the other hand, if the chat response unit 118 determines that the inquiry has been resolved (step S132: Yes), it notifies the customer identification information of the corresponding customer U1 and the response means information indicating the chat response means to the resolution required days update unit 124.

[0062] Next, the Resolution Time Update Unit 124 calculates the number of days elapsed from the initial inquiry date and time to the determined time. Then, the Resolution Time Update Unit 124 updates the final response means information stored in the inquiry history storage unit 131 by setting it to the chat response means indicated by the notified response means information (step S133). After that, the Resolution Time Update Unit 124 updates the Resolution Time Information by setting the Resolution Time Information stored in the inquiry history storage unit 131 to the calculated number of elapsed days (step S134). Next, the Model Generation Unit 121 generates new Resolution Time Estimation Model Information using the updated combination of the response content information and customer attribute information stored in the training information storage unit 134 and the Resolution Time Information, and updates the Resolution Time Estimation Model Information stored in the Resolution Time Estimation Model Storage Unit 132 with the generated Resolution Time Estimation Model Information (step S135). After that, the process in step S117 is executed again.

[0063] Furthermore, if the email sending / receiving unit 113 determines in step S117 that it has not obtained the response request information (step S117: No), the email sending / receiving unit 113 determines whether or not it has obtained the inquiry email sent from the terminal device 2 (step S136). If the email sending / receiving unit 113 determines that it has not obtained the inquiry email (step S136: No), the process in step S101 is executed again. On the other hand, if the email sending / receiving unit 113 determines that it has obtained the inquiry email (step S136: Yes), it extracts the IP address information of the terminal device 2 contained in the obtained response request information and stores it in the IP address storage unit (step S137). Then, the email sending / receiving unit 113 notifies the inquiry information identification unit 114 of the inquiry information indicating the content of the obtained inquiry email. Next, the inquiry information identification unit 114 identifies the aforementioned inquiry content information and customer attribute information based on the notified inquiry information. Then, the inquiry information identification unit 114 stores the identified inquiry content information and customer attribute information in the inquiry history storage unit 131, associating them with customer identification information that includes at least one of telephone number information, IP address information, and email address information (step S138). At this time, the inquiry information identification unit 114 notifies the answer means estimation unit 116 of the aforementioned question completion notification information.

[0064] Next, when the question completion notification information is received, the response means estimation unit 116, each time new inquiry content information and customer attribute information are stored in the inquiry history storage unit 131, performs a process to estimate a response means suitable for the customer from among a plurality of pre-set types of response means based on the newly stored inquiry content information and customer attribute information. Here, the response means estimation unit 116 determines whether or not it has estimated the response means to be an email response (step S139). If the response means estimation unit 116 determines that it has estimated a response means other than an email response (step S139: No), the process in step S108 is executed again.

[0065] On the other hand, if the response means estimation unit 116 determines that it has estimated the email response means (step S139: Yes), it notifies the inquiry content notification unit 119 of the newly stored inquiry content information and customer identification information in the inquiry history storage unit 131. At this time, the response means estimation unit 116 notifies the response means information indicating the estimated email response means and customer identification information of the resolution required days update unit 124. Subsequently, the inquiry content notification unit 119 generates an inquiry content notification email to inform the operator of the inquiry content based on the notified inquiry content information and customer identification information and sends it to the terminal device 3 (step S140). Next, after sending the inquiry content notification email, the inquiry content notification unit 119 determines whether or not it has obtained the aforementioned response completion notification information sent from the terminal device 3 within a predetermined waiting time (step S141). If the inquiry content notification unit 119 determines that it was not able to obtain the response completion notification information (step S141: No), the process in step S140 is executed again.

[0066] On the other hand, suppose the inquiry content notification unit 119 determines that it has obtained the response completion notification information (Step S141: Yes). In this case, the response time update unit 124 determines whether a predetermined number of days has elapsed since the customer identification information and the response means information indicating the email response means were notified from the response means estimation unit 116 (Step S142). If the response time update unit 124 determines that the predetermined number of days has not yet elapsed (Step S142: No), the process in Step S136 is executed again. On the other hand, if the response time update unit 124 determines that the predetermined number of days has elapsed since the notification (Step S142: Yes), it calculates the number of days that have elapsed from the initial inquiry date and time to the time of determination. Then, the response time update unit 124 updates the final response means information stored in the inquiry history storage unit 131 by setting it to the telephone response means indicated by the notified response means information (Step S143). Next, the Resolution Time Estimate Update Unit 124 updates the Resolution Time Estimate Information by setting the Resolution Time Estimate Information stored in the Inquiry History Storage Unit 131 to the calculated number of elapsed days (Step S144). Subsequently, each time the combination of the response content information and customer attribute information stored in the Training Information Storage Unit 134 and the Resolution Time Estimate Information is updated, the Model Generation Unit 121 generates new Resolution Time Estimate Model Information using the updated combination, and updates the Resolution Time Estimate Model Information stored in the Resolution Time Estimate Model Storage Unit 132 with the generated Resolution Time Estimate Model Information (Step S145). After that, the process in Step S136 is executed again.

[0067] As described above, in the cloud server 1 according to this embodiment, the response means estimation unit 116 estimates a response means suitable for the customer from among a plurality of pre-set response means, based on inquiry content information indicating the content of the inquiry from customer U1 and customer attribute information indicating the attributes of customer U1. This makes it easier to respond to customer inquiries from customer U1 with a response means suitable for customer U1, thereby improving customer U1's satisfaction with customer service.

[0068] Incidentally, companies like the customer support service provider described in the embodiment use various communication methods such as inquiry desks for content on websites and emails, call centers, chat systems for handling inquiries, and social media to respond to inquiries from customer U1. Furthermore, the means by which customer U1 makes inquiries are diverse, including telephone and email. When an inquiry comes from customer U1, it is often because customer U1 is facing a problem, and if the inquiry is not resolved smoothly, customer U1 is likely to become dissatisfied. One reason why inquiries cannot be resolved smoothly and take a long time to resolve is that the response may not be provided using the most suitable means for customer U1. For example, for customer U1 who is not comfortable operating a portable terminal device 2 such as a smartphone, response methods such as chat and email that can be viewed on the terminal device 2 may not be the most suitable means of response. Also, the most suitable means of response for customer U1 may differ depending on customer U1's age, gender, and the subject of the inquiry. In contrast, in this embodiment, it is possible to estimate an appropriate response method for customer U1 based on the content of the inquiry from customer U1 and the attributes of customer U1. Therefore, it is possible to suppress customer dissatisfaction caused by responding with an inappropriate response method.

[0069] Although embodiments of the present disclosure have been described above, the present disclosure is not limited to the embodiments described above. For example, as shown in Figure 19, the cloud server 2001 may include a telephone transfer determination unit 2125 that determines whether or not to transfer a call from a customer to an operator's telephone based on the congestion status of the telephone line used by the operator who will answer the inquiry, when the telephone answering means estimation unit 116 estimates that the telephone answering means is an appropriate answering means for the customer. In Figure 19, components similar to those in the embodiments are denoted by the same reference numerals as in Figure 3. In addition, the cloud server 2001 has all the functional configurations of the cloud server 1, but some of its functional configurations are omitted in Figure 19. Here, when a call is started with the terminal device 2, the IP telephone 4 sends call start notification information to the cloud server 2001 indicating that a call is in progress, including call start time information indicating the time the call started, and when the call ends, it sends call end notification information to the cloud server 2001 including call end time information indicating the time the call ended.

[0070] Furthermore, the cloud server 2001 includes a call status storage unit 2134, a call status information acquisition unit 2126, and a website location notification unit 2127. The call status storage unit 2134 stores call status information, call start time information, call end time information, and call duration information, which indicates the call status of each IP phone 4, in association with telephone identification information that identifies the IP phone 4.

[0071] When the call status information acquisition unit 2126 acquires call start information transmitted from the IP phone 4, it sets the call status information corresponding to the phone identification information of the IP phone 4 that sent the acquired call start notification information, which is stored in the call status storage unit 2134, to "on a call," and stores the call start time information included in the acquired call start time information in the call status storage unit 2134. On the other hand, when the call status information acquisition unit 2126 acquires call end notification information transmitted from the IP phone 4, it sets the call status information corresponding to the phone identification information of the IP phone 4 that sent the acquired call end notification information, which is stored in the call status storage unit 2134, to "waiting," and stores the call end time information included in the acquired call end time information in the call status storage unit 2134. At this time, the call status information acquisition unit 2126 calculates the elapsed time from the call start time indicated by the call start time information to the call end time indicated by the call end time information, and stores the call duration information indicating the calculated elapsed time in the call status storage unit 2134.

[0072] When the call forwarding determination unit 2125 determines that the call is a call forwarding means by the answering means estimation unit 116, it refers to the call start time information stored in the call status storage unit 2134 and calculates the average arrival interval, which indicates the time interval at which the IP phone 4 receives a call from customer U1. The call forwarding determination unit 2125 also refers to the call duration information stored in the call status storage unit 2134 and calculates the average service duration, which indicates the average call duration across all IP phones 4. Next, the call forwarding determination unit 2125 calculates the utilization rate from the calculated average arrival interval and average service duration, and calculates the average waiting time from the calculated utilization rate and average service duration based on the M / M / 1 queuing theory. If the calculated average waiting time is less than or equal to a preset waiting time threshold, the call forwarding determination unit 2125 determines that the call from the customer should be forwarded to the operator's phone. In this case, the call answering request unit 117 sends answering request message information to the IP phone 4, as described in the embodiment. On the other hand, the call forwarding determination unit 2125 determines that if the calculated average waiting time is longer than a preset waiting time threshold, it will not forward the customer's call to the operator's phone. In this case, the website location notification unit 2127 generates website location notification information that notifies the location of the website where the FAQ (Frequently Asked Questions) is posted and notifies the voice transmission / reception unit 111. The voice transmission / reception unit 111 then generates voice information to notify the website's location by voice based on the notified website location notification information and transmits it to the terminal device 2.

[0073] This configuration helps to prevent customers U1 from being unnecessarily kept waiting for their calls to be connected to an operator when telephone lines are congested.

[0074] In the embodiment, for example, as shown in Figure 20, the cloud server 3001 may include a telephone response necessity determination unit 3128 that determines whether or not it is necessary to respond by telephone based on the inquiry content information and customer attribute information identified by the inquiry information identification unit 114. In Figure 20, components similar to those in the embodiment are denoted by the same reference numerals as in Figure 3. Furthermore, the cloud server 3001 has all the functional configurations of the cloud server 1, but some of its functional configurations are omitted in Figure 20. The telephone response necessity determination unit 3128 determines whether or not it is necessary to respond by telephone based on the inquiry content information and customer attribute information newly stored in the inquiry history storage unit 131. For example, if the inquiry concerns a recall, the telephone response necessity determination unit 3128 determines that it is necessary to respond by telephone. Then, if the telephone response request unit 117 determines that a response is required via telephone response means, it sends the response request message information to the IP telephone 4, regardless of the response means estimated by the response means estimation unit 116.

[0075] With this configuration, in the case of urgent inquiries, such as a recall of electrical equipment, the call can be connected to an operator quickly, thus contributing to alleviating customer U1's anxiety and facilitating an early resolution.

[0076] In the embodiment, the response means may further include a website-presentation response means that presents a website containing FAQs.

[0077] In this embodiment, we described an example in which, when customer U1 makes an inquiry by "telephone," the response means estimation unit 116 estimates a telephone response means and connects directly to an operator. However, the system is not limited to this, and for example, the telephone number of the operator to be connected to may be notified by voice.

[0078] Furthermore, the various functions of Cloud Server 1 related to this disclosure may be implemented by software, firmware, or a combination of software and firmware. In this case, the software or firmware may be written as a program, and the program may be stored on a computer-readable recording medium such as a flexible disk, CD-ROM (Compact Disc Read Only Memory), DVD (Digital Versatile Disc), and MO (Magneto-Optical Disc) and distributed, and a computer capable of implementing the aforementioned functions may be configured by loading and installing the program on a computer. In cases where each function is implemented by a division of labor between the OS (Operating System) and applications, or by cooperation between the OS and applications, only the parts other than the OS may be stored on the recording medium.

[0079] Furthermore, it is possible to superimpose each program onto the carrier wave and distribute it over a network. For example, the program could be posted on a bulletin board system (BBS) on the network and distributed over the network. These programs could then be launched and executed under the control of the OS, just like other application programs, to perform the aforementioned processing.

[0080] The various aspects of this disclosure are summarized below as an appendix.

[0081] (Note 1) An inquiry information identification unit identifies inquiry content information that shows the content of the customer's inquiry and customer attribute information that shows the attributes of the customer, The system includes an answer means estimation unit that estimates an appropriate answer means for the customer from among a plurality of pre-set answer means based on the identified inquiry content information and customer attribute information, Customer service management device. (Note 2) The response means estimation unit estimates the number of days required to resolve the customer's inquiry using a resolution time estimation model that determines the number of days required to resolve the customer's inquiry for each of the multiple types of response means corresponding to the identified inquiry content information and customer attribute information, and estimates the response means with the shortest estimated number of days required to resolve as the response means suitable for the customer. A customer service management device as described in Appendix 1. (Note 3) The aforementioned response means includes at least a chat response means for responding via chat, If the response means estimation unit estimates that the chat response means is a suitable response means for the customer, the system further includes a contact notification unit that notifies the customer of the URL of a website for making an inquiry to a contact via chat. A customer service management device as described in Appendix 1 or 2. (Note 4) The aforementioned response means includes, at a minimum, an email response means for responding by email. If the response means estimation unit estimates that the email response means is a suitable response means for the customer, the system further includes a contact notification unit that notifies the customer of the URL of a website where an inquiry format for making an inquiry to the contact person via email is published. A customer service management device as described in any one of the appendices 1 to 3. (Note 5) The aforementioned response means includes, at a minimum, a telephone response means for providing a response by telephone. If the response means estimation unit estimates that the telephone response means is a suitable response means for the customer, the system further includes a telephone response request unit that requests an operator to respond by telephone from a pre-set contact. A customer service management device as described in any one of the appendices 1 to 4. (Note 6) The aforementioned response means includes, at a minimum, a telephone response means for providing a response by telephone. If the response means estimation unit estimates that the telephone response means is a suitable response means for the customer, the system further includes a telephone transfer determination unit that determines whether or not to transfer the call from the customer to the operator's telephone based on the congestion status of the telephone line used by the operator who will answer the inquiry. A customer service management device as described in any one of the appendices 1 through 5. (Note 7) The system further includes a telephone response necessity determination unit that determines whether or not it is necessary to respond by telephone based on the identified inquiry content information and customer attribute information. A customer service management device as described in any one of the appendices 1 through 6. (Note 8) An inquiry information identification unit identifies inquiry content information that shows the content of the customer's inquiry and customer attribute information that shows the attributes of the customer, The system includes an answer means estimation unit that estimates an appropriate answer means for the customer from among a plurality of pre-set answer means based on the identified inquiry content information and customer attribute information, Customer service management system. (Note 9) The customer response management device includes the steps of identifying inquiry content information indicating the content of the customer's inquiry and customer attribute information indicating the customer's attributes, The customer response means management device includes the step of estimating a response means suitable for the customer from among a plurality of pre-set response means based on the identified inquiry content information and customer attribute information, Customer service management methods. (Note 10) Computers, An inquiry information identification unit identifies inquiry content information that shows the content of the customer's inquiry and customer attribute information that shows the attributes of the customer. A response means estimation unit estimates a response means suitable for the customer from among a plurality of pre-set response means, based on the identified inquiry content information and customer attribute information. A program designed to function as such. [Industrial applicability]

[0082] This disclosure is suitable as a system for providing customer support services. [Explanation of Symbols]

[0083] 1,2001,3001 Cloud Server, 2,3 Terminal Device, 4 IP Phone, 5 SIP Server, 6 Mail Server, 101 CPU, 102 Main Memory Unit, 103 Auxiliary Memory Unit, 106 Communication Unit, 108 Timing Unit, 109 Bus, 111 Voice Transmission / Reception Unit, 112 Chat Information Transmission / Reception Unit, 113 Mail Transmission / Reception Unit, 114 Inquiry Information Identification Unit, 115 Voice Recognition Unit, 116 Response Method Estimation Unit, 117 Telephone Response Request Unit, 118 Chat Response Unit, 119 Inquiry Content Notification Unit, 120 Teacher Information Selection Unit, 121 Model Generation Unit, 122 Contact Information Notification Unit, 123 Question Message Notification Unit, 124 Resolution Time Update Unit, 131 Inquiry History Storage Unit, 132 Resolution Time Estimation Model Storage Unit, 133 Response Information Storage Unit, 134 Teacher information storage unit, 204, 304 Display unit, 205, 305 Input unit, 2125 Call forwarding determination unit, 2126 Call status information acquisition unit, 2127 Website location notification unit, 2134 Call status storage unit, 3128 Call response necessity determination unit, GA1, GA2 Chat screen image, L10 Input layer, L20 Hidden layer, L30 Output layer, M111, M112, M113 Question message image, M121, M122, M123 Answer message image, NW1 Wide area network, U1 Customer

Claims

1. An inquiry information identification unit identifies inquiry content information that shows the content of the customer's inquiry and customer attribute information that shows the attributes of the customer, The system includes an answer means estimation unit that estimates an appropriate answer means for the customer from among a plurality of pre-set answer means based on the identified inquiry content information and customer attribute information, Customer service management device.

2. The response means estimation unit estimates the number of days required to resolve the customer's inquiry using a resolution time estimation model that determines the number of days required to resolve the customer's inquiry for each of the multiple types of response means corresponding to the identified inquiry content information and customer attribute information, and estimates the response means with the shortest estimated number of days required to resolve as the response means suitable for the customer. A customer service means management device according to claim 1.

3. The aforementioned response means includes at least a chat response means for responding via chat, If the response means estimation unit estimates that the chat response means is a suitable response means for the customer, the system further includes a contact notification unit that notifies the customer of the URL of a website for making an inquiry to a contact via chat. A customer service means management device according to claim 1 or 2.

4. The aforementioned response means includes, at a minimum, an email response means for responding by email. If the response means estimation unit estimates that the email response means is a suitable response means for the customer, the system further includes a contact notification unit that notifies the customer of the URL of a website where an inquiry format for making an inquiry to the contact person via email is published. A customer service means management device according to claim 1 or 2.

5. The aforementioned response means includes, at a minimum, a telephone response means for providing a response by telephone. If the response means estimation unit estimates that the telephone response means is a suitable response means for the customer, the system further includes a telephone response request unit that requests an operator to respond by telephone from a pre-set contact. A customer service means management device according to claim 1 or 2.

6. The aforementioned response means includes, at a minimum, a telephone response means for providing a response by telephone. If the response means estimation unit estimates that the telephone response means is a suitable response means for the customer, the system further includes a telephone transfer determination unit that determines whether or not to transfer the call from the customer to the operator's telephone based on the congestion status of the telephone line used by the operator who will answer the inquiry. A customer service means management device according to claim 1 or 2.

7. The system further includes a telephone response necessity determination unit that determines whether or not it is necessary to respond by telephone based on the identified inquiry content information and customer attribute information. A customer service means management device according to claim 1 or 2.

8. An inquiry information identification unit identifies inquiry content information that shows the content of the customer's inquiry and customer attribute information that shows the attributes of the customer, The system includes an answer means estimation unit that estimates an appropriate answer means for the customer from among a plurality of pre-set answer means based on the identified inquiry content information and customer attribute information, Customer service management system.

9. The customer response management device includes the steps of identifying inquiry content information indicating the content of the customer's inquiry and customer attribute information indicating the customer's attributes, The customer response means management device includes the step of estimating a response means suitable for the customer from among a plurality of pre-set response means based on the identified inquiry content information and customer attribute information, Customer service management methods.

10. Computers, An inquiry information identification unit identifies inquiry content information that shows the content of the customer's inquiry and customer attribute information that shows the attributes of the customer. A response means estimation unit estimates a response means suitable for the customer from among a plurality of pre-set response means, based on the identified inquiry content information and customer attribute information. A program designed to function as such.

Images