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

Abstract

To provide a customer service support system, a customer service support device, a customer service support method, and a customer service support program that perform notification according to customer service situations.SOLUTION: A customer service support system 1 according to one embodiment of the present invention includes: emotion analysis means 10 for analyzing respective emotions of a first salesclerk and a customer; and notification instruction means 20 for instructing the output of a customer service support notification to a second salesclerk based on an emotion analysis result of the first salesclerk and that of the customer.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a customer service support system, a customer service support device, a customer service support method, and a customer service support program. [Background technology]

[0002] In customer service work, it is necessary to properly assess the customer's situation and provide high-quality customer service. In addition, to provide high-quality customer service, it may be necessary to have another store employee assist the customer depending on the situation.

[0003] As a technology that meets the above-mentioned needs, Patent Document 1 discloses a technology that captures images and audio of a customer and an attendant, and evaluates the customer's level of understanding from the captured images and audio, thereby helping to improve customer satisfaction.

[0004] Furthermore, Patent Document 2 discloses a technology that acquires the content of the conversation between the customer and the customer service representative and the status of the customer service representative, and generates information indicating the degree of need to support the customer service representative based on the similarity with past cases and the status of the customer service representative. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Publication No. 2018-124604 [Patent Document 2] International Publication No. 2019 / 163700 Summary of the Invention [Problem to be solved by the invention]

[0006] Depending on the customer-to-customer interaction, a customer may need assistance from another employee. However, it is difficult for the other employee to recognize that assistance is needed. It is also not easy for an employee currently serving a customer to request assistance from another employee. Therefore, a system is needed that can appropriately notify an employee other than the employee currently serving a customer that assistance is needed.

[0007] Therefore, one of the objects of the present invention is to provide a customer service support system, a customer service support device, a customer service support method, and a customer service support program that provide notifications according to the customer service situation based on the results of emotion analysis between customers and store clerks. [Means for solving the problem]

[0008] One aspect of the customer service support system of the present invention includes an emotion analysis means for analyzing the emotions between a first store clerk and a customer based on an image of the first store clerk and the customer, and a notification instruction means for instructing the output of a customer service support notification to a second store clerk different from the first store clerk based on the emotion analysis results of the first store clerk and the emotion analysis results of the customer.

[0009] One aspect of the customer service support device of the present invention includes an emotion analysis unit that analyzes the emotions of a first store clerk and a customer based on an image of the first store clerk and the customer, and a notification instruction unit that instructs the output of a customer service support notification to a second store clerk different from the first store clerk based on the emotion analysis results of the first store clerk and the emotion analysis results of the customer.

[0010] One aspect of the customer service support method of the present invention analyzes the emotions of a first store clerk and a customer based on an image of the first store clerk and the customer, and instructs the output of a customer service support notification to a second store clerk different from the first store clerk based on the emotion analysis results of the first store clerk and the emotion analysis results of the customer.

[0011] One aspect of the customer service support program of the present invention causes a computer to analyze the emotions of a first store clerk and a customer based on an image of the first store clerk and the customer, and to instruct a second store clerk different from the first store clerk to output a customer service support notification based on the emotion analysis results of the first store clerk and the emotion analysis results of the customer. [Effects of the Invention]

[0012] According to the present invention, it is possible to notify a store clerk other than the one currently serving the customer that customer service assistance is required. [Brief explanation of the drawings]

[0013] [Figure 1] 1 is a block diagram showing an example of the configuration of a customer service support system according to a first embodiment. [Figure 2] 4 is a flowchart showing an example of the operation of the customer service support system according to the first embodiment. [Figure 3] FIG. 10 is a diagram showing an overview of a customer service support system according to a second embodiment. [Figure 4] FIG. 1 is a block diagram illustrating an example of the configuration of an analysis system. [Figure 5] This is an example of a combination of emotion analysis results and whether or not to output a customer service support notification. [Figure 6] This is an example of a combination of changes in the emotion analysis results and whether or not to output a customer service support notification. [Figure 7] 10 is a flowchart showing an example of the operation of the customer service support system according to the second embodiment. [Figure 8] FIG. 10 is a block diagram showing an example of the configuration of an analysis system according to a third embodiment. [Figure 9] 10 is a flowchart showing an example of the operation of the customer service support system according to the third embodiment. [Figure 10] 10 is a flowchart showing a modified example of the operation of the customer service support system according to the third embodiment. [Figure 11] FIG. 10 is a block diagram showing an example of the configuration of an analysis system according to a fourth embodiment. [Figure 12] 10 is an example of a screen output of real-time analysis results. [Figure 13] 10 is an example of a screen output of a customer service history result. [Figure 14] FIG. 2 is a block diagram showing an example of hardware constituting each unit of the customer service support system. DETAILED DESCRIPTION OF THE INVENTION

[0014] [First embodiment] FIG. 1 is a functional block diagram showing the configuration of a customer service support system 1 according to a first embodiment of the present invention. In FIG. 1, each block shows the configuration of a functional unit. Therefore, the blocks shown in FIG. 1 may be implemented in a single customer service support device, or may be implemented separately in multiple devices. The customer service support system 1 includes emotion analysis means 10 and notification instruction means 20.

[0015] The emotion analysis means 10 analyzes the emotions of the first store clerk and the customer based on the images of the first store clerk and the customer. Well-known techniques can be applied to the emotion analysis.

[0016] The notification instruction means 20 instructs the output of a customer service assistance notice based on the clerk's emotion analysis results and the customer's emotion analysis results. The notification instruction means may also output the customer service assistance notice. The customer service assistance notice is output to a second clerk who is different from the first clerk currently serving the customer. The second clerk may be the first clerk's superior, senior employee, store manager, or colleague, but is not limited to this, as long as the second clerk is different from the first clerk. The customer service assistance notice may be output to one second clerk or to multiple second clerks. Examples of output methods include, but are not limited to, output to a mobile device used by the second clerk or output to a store terminal that the second clerk can view. Additionally, the notification instruction means 20 may instruct the first clerk to output a notice informing the first clerk that the customer service assistance notice has been output. The notice output to the first clerk may be the same as the notice output to the second clerk, or a different notice for the first clerk.

[0017] The operation of the customer service support system according to the first embodiment will be described with reference to the drawings. Fig. 2 is a flowchart showing an example of the operation of the customer service support system 1.

[0018] Sentiment analysis means 10 analyzes the emotions of the first clerk and the customer based on an image of the first clerk and the customer (step S1). Then, emotion analysis means 10 determines whether to output a customer service assistance notice based on the emotion analysis result of the first clerk and the emotion analysis result of the customer (step S2). If it determines to output a customer service assistance notice based on the emotion analysis result of the first clerk and the emotion analysis result of the customer (YES in step S2), notification instruction means 20 instructs the second clerk to output a customer service assistance notice (step S3), and returns to step S1.

[0019] On the other hand, if it is determined that a customer service support notice should not be output based on the emotion analysis results of the first store clerk and the customer (NO in step S2), the process returns to step S1. The determination of whether to output the customer service assistance notification (step S2) may be made by a determination unit (not shown) or by the notification instruction means 20.

[0020] The customer service support system according to the first embodiment can output a customer service support notification based on the emotion analysis results of the first store clerk and the customer, thereby preventing a decline in customer satisfaction.

[0021] [Second embodiment] In customer service support system 2 of the second embodiment, analysis system 100 corresponds to customer service support system 1 of the first embodiment, but differs from customer service support system 1 in that analysis system 100 further includes information acquisition unit 110. In addition, customer service support system 2 further includes store clerk terminal 200 as an output destination of customer service support notifications. This will be described in detail below with reference to the drawings.

[0022] 3 is a diagram showing an overview of a customer service support system 2 according to a second embodiment of the present invention. The customer service support system 2 includes an analysis system 100 and a store clerk terminal 200, which are communicatively connected via a network 300. The analysis system 100 may be implemented in the store clerk terminal 200, or part of the configuration of the analysis system 100 may be implemented in the store clerk terminal 200.

[0023] There may be multiple store clerk terminals 200. Preferably, each store clerk uses one store clerk terminal 200.

[0024] The analysis system 100 will now be described. FIG. 4 is a block diagram showing an example of the configuration of the analysis system 100. In FIG. 4, each block shows the configuration of a functional unit. Therefore, the blocks shown in FIG. 4 may be implemented in a single device, or may be implemented separately in multiple devices. The analysis system 100 includes an information acquisition unit 110, a sentiment analysis unit 120, and a notification instruction unit 130. However, the configuration of the analysis system 100 is not limited to this.

[0025] The information acquisition unit 110 acquires an image of the first clerk and the customer. The image may be acquired from a camera (not shown). Alternatively, the image may be input as image data. When a camera is used, the camera may be included in the analysis system 100, or may be provided as an external device and communicatively connected to the analysis system 100. Alternatively, the camera may be provided in the clerk terminal 200, and the information acquisition unit 110 may acquire an image captured by the camera of the clerk terminal 200.

[0026] The information acquisition unit 110 transmits the acquired images to the emotion analysis unit 120. The acquired images of the first store clerk and the customer may include the first store clerk and the customer in the same frame, or the image including the first store clerk and the image including the customer may be acquired as separate images.

[0027] The emotion analysis unit 120 analyzes the emotions of the first clerk and the customer based on images of the first clerk and the customer. Well-known techniques can be applied to the emotion analysis. The information acquisition unit 110 may acquire vital data of the first clerk and the customer instead of images of the first clerk and the customer. In this case, the emotion analysis unit 120 performs emotion analysis based on the vital data of the first clerk and the customer acquired by the information acquisition unit 110. Alternatively, the information acquisition unit 110 may acquire vital data of the first clerk and the customer in addition to images of the first clerk and the customer. In this case, the emotion analysis unit 120 performs emotion analysis based on the images of the first clerk and the customer acquired by the information acquisition unit 110 and the vital data of the first clerk and the customer. Examples of vital data include body surface temperature and heart rate, but are not limited to these as long as they are biometric information that can be used for emotion analysis. An example of a method for acquiring various vital data is acquisition by a wearable device (not shown), but vital data may also be acquired by other methods. The data acquired by the information acquisition unit 110 is not limited to the exemplified data, as long as it is data that can be used for emotion analysis.

[0028] The emotion analysis unit 120 determines whether to output a customer service assistance notification based on the emotion analysis results of the first clerk and the customer. The determination of whether to output a customer service assistance notification may be made by a determination unit (not shown) or by the notification instruction unit 130. The conditions used for the determination may be determined uniformly in advance or may be determined for each clerk. When conditions are determined for each clerk, the conditions can be determined taking into account clerk information such as the length of service, experience, and skills of each clerk, making it possible to output a customer service assistance notification tailored to each clerk. When determining the predetermined value taking into account the clerk information, there are several methods available, including a method of acquiring clerk information by identifying the clerk who will be serving customers using biometric authentication, a method of registering clerk identification information linked to the clerk information in advance, and a method of directly entering the clerk information. However, the method is not limited to these as long as the clerk information can be acquired.

[0029] A specific example of determining whether to output a customer service assistance notification will be described. As a first example, the emotion analysis results of the first store clerk and the customer are classified, and whether to output a customer service assistance notification is determined based on the combination of the classifications of the emotion analysis results. Here, one example of classifying the emotion analysis results is to classify them into "positive," "neutral," or "negative." Specifically, positive emotions such as "happiness" and "empathy" can be determined as "positive," negative emotions such as "fear," "confusion," "disgust," and "contempt" can be determined as "negative," and emotions that cannot be classified as "positive" or "negative" can be determined as "neutral." However, the classification of the emotion analysis results is not limited to these. Here, an example will be described in which the emotion analysis results are classified into "positive," "neutral," or "negative."

[0030] FIG. 5 shows an example of the correspondence between the combination of classifications of the emotion analysis results of the first clerk and the customer and whether or not to output a customer service assistance notification. If the classifications of the emotion analysis results of the first clerk and the customer are both "positive," it is determined that a customer service assistance notification should not be output, and the notification instruction unit 130 does not instruct the output of the customer service assistance notification. If one of the classifications of the emotion analysis results of the first clerk and the customer is "positive" and the other is "neutral," it is also determined that a customer service assistance notification should not be output, and the notification instruction unit 130 does not instruct the output of the customer service assistance notification. Furthermore, if the classifications of the emotion analysis results of the first clerk and the customer are both "neutral," it is also determined that a customer service assistance notification should not be output, and the notification instruction unit 130 does not instruct the output of the customer service assistance notification.

[0031] On the other hand, if one of the classifications of the emotion analysis results between the first clerk and the customer is "negative" and the other is "positive" or "neutral," it is determined that a customer service assistance notice should be output, and the notification instructing unit 130 instructs the output of the customer service assistance notice. Furthermore, if the classifications of the emotion analysis results between the first clerk and the customer are both "negative," it is also determined that a customer service assistance notice should be output, and the notification instructing unit 130 instructs the output of the customer service assistance notice.

[0032] Furthermore, when outputting a customer service assistance notification, the details of the customer service assistance notification may be changed depending on the combination of classifications of the emotion analysis results between the first sales clerk and the customer. For example, if the classifications of the emotion analysis results between the first sales clerk and the customer are both "negative," it is possible that the relationship between the first sales clerk and the customer is strained. Therefore, by changing the customer service assistance notification in such a case, it is possible to easily grasp that customer service assistance is highly necessary. Possible changes to the details of the customer service assistance notification include, but are not limited to, changing the output destination or the output format.

[0033] The output destination can be changed to the senior employee of the first store clerk, the first store clerk's boss, the store manager, or a supervisor, etc. In this case, it is possible to notify the second store clerk who can take appropriate action depending on the customer situation.

[0034] Possible changes to the output format include, but are not limited to, changing the notification content, or changing the sound, light, or vibration when the notification is made. Specific changes to the notification content include changing the content of the notification text, changing the color or font of the text, or changing the design of the notification screen, such as the color or mark. Changing the output format is convenient because it makes it easier for the second store clerk who receives the customer service assistance notification to understand the customer service situation.

[0035] As a second example, the sentiment analysis results of the salesperson and the customer are classified according to changes, and whether to output a customer service assistance notification is determined based on the combination of classifications based on the changes in the sentiment analysis results. Figure 6 shows an example of the correspondence between the combination of classifications based on changes in the sentiment analysis results and whether to output a customer service assistance notification. Classification based on changes in the sentiment analysis results can be, for example, "there is a change of more than a specified amount on the positive side," "there is a change of more than a specified amount on the negative side," or "there is no change of more than a specified amount." Methods for determining changes in the sentiment analysis results include, but are not limited to, a method of determining from a history of changes in sentiment over a specified period of time, or a method of determining by comparing the sentiment analysis results at the start of customer service with the sentiment analysis results during customer service. In addition, the details of the customer service assistance notification may be changed based on the amount of change in the sentiment analysis results within a specified period of time.

[0036] If the classification based on changes in the emotion analysis results for both the first clerk and the customer is "a change greater than a predetermined amount on the positive side," it is determined that a customer service assistance notice will not be output, and the notification instructing unit 130 does not instruct the output of the customer service assistance notice. If one of the classifications based on changes in the emotion analysis results for the first clerk and the customer is "a change greater than a predetermined amount on the positive side" and the other is "no change greater than a predetermined amount," it is also determined that a customer service assistance notice will not be output, and the notification instructing unit 130 does not instruct the output of the customer service assistance notice. Furthermore, if the classification based on changes in the emotion analysis results for both the first clerk and the customer is "no change greater than a predetermined amount," it is also determined that a customer service assistance notice will not be output, and the notification instructing unit 130 does not instruct the output of the customer service assistance notice.

[0037] On the other hand, if one of the classifications based on the change in the emotion analysis results between the first clerk and the customer is "a change of more than a predetermined amount on the negative side" and the other is "a change of more than a predetermined amount on the positive side" or "no change of more than a predetermined amount," it is determined that a customer service assistance notice should be output, and the notification instructing unit 130 instructs the output of the customer service assistance notice. Furthermore, if the classifications of the change in the emotion analysis results between the first clerk and the customer are both "a change of more than a predetermined amount on the negative side," it is also determined that a customer service assistance notice should be output, and the notification instructing unit 130 instructs the output of the customer service assistance notice.

[0038] Furthermore, when outputting a customer service assistance notification, the details of the customer service assistance notification may be changed depending on the combination of classifications based on changes in the emotion analysis results. For example, if the classifications based on changes in the emotion analysis results for the first sales clerk and the customer are both "changes to the negative side by more than a predetermined amount," it is possible that the relationship between the first sales clerk and the customer is becoming strained. Therefore, by changing the details of the customer service assistance notification in such a case, it is possible to easily grasp that customer service assistance is highly necessary. Possible changes to the details of the customer service assistance notification include, but are not limited to, changing the output destination or the output format.

[0039] As a third example, if the emotion analysis results between the first clerk and the customer are expressed numerically, it can be determined that a customer service assistance notice should be output when the emotion analysis results between the first clerk and the customer diverge by a predetermined value or more. If the emotion analysis results between the first clerk and the customer diverge by a predetermined value or more, there is a high possibility that the customer service is not of high quality or high satisfaction. Therefore, by outputting a customer service assistance notice when the emotion analysis results between the first clerk and the customer diverge by a predetermined value or more, customer service assistance by the second clerk can be encouraged at the necessary time, thereby improving customer satisfaction.

[0040] The determination as to whether to output the customer service assistance notification is not limited to the above-described example, and the above-described determination examples may be combined.

[0041] If it is determined that a customer service assistance notice should be output, the notification instructing unit 130 instructs the clerk terminal 200 to output the customer service assistance notice. The notification instructing unit 130 may instruct a notifying means (not shown) implemented in the clerk terminal to output the notice. Alternatively, the notification instructing unit 130 may function as the notifying means and output the notice. The customer service assistance notice is output to the clerk terminal 200 used by a second clerk who is different from the first clerk currently serving the customer. Specifically, when clerk A using clerk terminal 200-1 is serving customer C, if it is determined that a customer service assistance notice should be output based on the emotion analysis results between clerk A and customer C, the notification instructing unit 130 instructs the clerk terminal 200-2 used by another clerk B to output the customer service assistance notice. The output format of the customer service assistance notice may be changed depending on the customer service situation. Specifically, the notification content, sound, light, or vibration may be changed depending on the customer service situation. Furthermore, when the customer service assistance notice is displayed on the store clerk terminal 200, the display color, font of characters, and screen design may be changed depending on the customer service situation.

[0042] The output destination of the customer service assistance notification may be set in advance for each employee. The output destination of the customer service assistance notification may be the employee terminal 200 of the first employee's superior or senior employee. Alternatively, the output destination of the customer service assistance notification may not be set in advance, and the customer service assistance notification may be output to the employee terminal 200 of a employee who is not serving the customer at the time the customer service assistance notification is output. Furthermore, the customer service assistance notification may be output to the employee terminals 200 of multiple employees other than the first employee. The output destination may also be changed depending on the customer service situation. One example of changing the output destination depending on the customer service situation is changing the output destination to the store manager when a customer is angry, but the customer service situation and the output destination to which the output destination is changed are not limited to this.

[0043] The customer service assistance notice may also include customer service information related to the customer service provided by the first sales clerk. The customer service information may include information about the customer acquired by the first sales clerk while serving the customer, a product history of the first sales clerk's recommendations, etc. The customer information acquired by the first sales clerk while serving the customer may include the customer's preferences, desired products, concerns, etc., but is not limited to these, as long as it is information about the customer.

[0044] If the customer service assistance notice includes customer service information about the customer service provided by the first salesperson, the second salesperson who will provide customer service assistance can check the customer service information in advance, making it easier to provide customer service assistance. Specifically, by checking the customer service information provided by the first salesperson before providing customer service assistance, the second salesperson can avoid repeatedly asking the customer the same questions or suggesting products as the first salesperson. This can improve the quality of customer service.

[0045] The operation of the customer service support system according to the second embodiment will be described with reference to the drawings. Fig. 7 is a flowchart showing an example of the operation of the customer service support system 2.

[0046] The information acquisition unit 110 acquires an image of the first store clerk and the customer (step S10), and the emotion analysis unit 120 analyzes the emotions of the first store clerk and the customer based on the acquired image (step S11). The emotion analysis unit 120 then determines whether to output a customer service assistance notification based on the emotion analysis results of the first store clerk and the customer (step S12). Step S12 may be performed by a determination unit (not shown) or by the notification instruction unit 130.

[0047] If it is determined that a customer service assistance notification should be output based on the emotion analysis results of the first store clerk and the emotion analysis results of the customer (YES in step S12), the notification instruction unit 130 instructs the second store clerk to output a customer service assistance notification (step S13), and returns to step S10.

[0048] On the other hand, if it is determined that a customer service assistance notice should not be output based on the emotion analysis results of the first store clerk and the customer (NO in step S12), the process returns to step S10.

[0049] The customer service support system according to the second embodiment can output a customer service support notification when customer service support is required because it determines whether to output the customer service support notification based on the emotion analysis result of the first store clerk and the emotion analysis result of the customer, and instructs the output of the customer service support notification when it determines that the customer service support notification should be output.

[0050] Furthermore, in the customer service support system according to the second embodiment, when images of the first store clerk and the customer are used for emotion analysis, emotion analysis can be performed based on facial expressions and gestures, and a customer service support notification can be output.

[0051] Furthermore, in the customer service support system according to the second embodiment, if the details of the customer service support notification (notification content) include customer service information about the customer service provided by the first sales clerk, customer service support by the second sales clerk becomes easier because the customer service information can be checked in advance when providing customer service support.

[0052] [Third embodiment] The customer service support system 3 in the third embodiment differs from the customer service support system 2 in the second embodiment in that it further includes a text analysis unit 140. Hereinafter, the same components as those in the second embodiment will be assigned the same reference numerals and descriptions thereof will be omitted.

[0053] The overview of a customer service support system 3 according to a third embodiment of the present invention is the same as that shown in FIG. 3. However, the analysis system 100 in the customer service support system 3 differs from the customer service support system 2 according to the second embodiment in that it further includes a text analysis unit 140. The sentiment analysis unit 120 and the text analysis unit 140 may be implemented in a single device, or may be implemented separately in multiple devices. The analysis system 100 may be implemented in a store clerk terminal 200, or part of the configuration of the analysis system 100 may be implemented in the store clerk terminal 200. The text analysis unit 140 serves as text analysis means for analyzing a conversation between a first store clerk and a customer.

[0054] An analysis system 100 according to a third embodiment will now be described. FIG. 8 is a block diagram showing an example of the configuration of the analysis system 100 according to the third embodiment. In FIG. 8, each block shows the configuration of a functional unit. Therefore, the blocks shown in FIG. 8 may be implemented in a single device, or may be implemented separately in multiple devices. The analysis system 100 according to the third embodiment includes an information acquisition unit 110, a sentiment analysis unit 120, a notification instruction unit 130, and a text analysis unit 140. However, the configuration of the analysis system 100 is not limited to this.

[0055] The information acquisition unit 110 acquires the voice of the first store clerk and the customer. The voice may be acquired from a microphone (not shown). Alternatively, the voice may be input as voice data from a server (not shown). The microphone may be included in the analysis system 100, or may be provided as an external device and communicatively connected to the analysis system 100 via the network 300. Alternatively, the microphone may be provided in the store clerk terminal 200, and the voice acquired by the microphone of the store clerk terminal 200 may be acquired by the information acquisition unit 110.

[0056] The information acquisition unit 110 transmits the acquired speech to the text analysis unit 140. The text analysis unit 140 performs text analysis of the conversation between the first store clerk and the customer based on the speech between the first store clerk and the customer. Well-known techniques can be applied to the text analysis.

[0057] The text analysis unit 140 determines whether a predetermined keyword is included in the content of the conversation between the first store clerk and the customer. The determination of whether a predetermined keyword is included may be performed by a determination unit (not shown). The predetermined keyword may be, for example, a word that a store clerk uses that makes the customer feel uncomfortable, or a word that a customer is likely to use when they are dissatisfied. The predetermined keyword may be registered in advance, or may be updated as the customer service progresses.

[0058] When updating the predetermined keywords while customer service is being provided, a keyword that frequently appears when it is determined based on the sentiment analysis results that a customer service support notification should be output may be registered as a new predetermined keyword. Also, if a keyword is registered as a predetermined keyword but it is rarely determined based on the sentiment analysis results that a customer service support notification should be output when that keyword is uttered, that keyword may be deleted from the list of predetermined keywords. These registrations and deletions may be performed automatically.

[0059] If it is determined that the conversation between the first clerk and the customer contains a predetermined keyword, the notification instruction unit 130 instructs the clerk terminal 200 to output a customer service assistance notice. The notification instruction unit 130 may instruct a notifying means (not shown) implemented in the clerk terminal to output the notice. Alternatively, the notification instruction unit 130 may function as the notifying means and output the notice. The customer service assistance notice is output to the clerk terminal 200 used by a second clerk who is different from the first clerk currently serving the customer. Specifically, when clerk A using clerk terminal 200-1 is serving customer C, if it is determined that the conversation between clerk A and customer C contains a predetermined keyword, the notification instruction unit 130 instructs the clerk terminal 200-2 used by another clerk B to output a customer service assistance notice.

[0060] In addition, if it is determined that the conversation between the first store clerk and the customer contains a predetermined keyword, but it is determined that a customer service support notification should not be output based on the results of the emotion analysis of the first store clerk and the customer, the system may be configured not to instruct the output of the customer service support notification.

[0061] The output destination of the customer service assistance notification may be set in advance for each employee. Possible output destinations of the customer service assistance notification include the employee terminal 200 of the first employee's superior or a senior employee. Alternatively, the output destination of the customer service assistance notification may not be set in advance, and the customer service assistance notification may be output to the employee terminal 200 of an employee who is not serving a customer at the time the customer service assistance notification is output. Furthermore, the customer service assistance notification may be output to the employee terminals 200 of multiple employees other than the first employee. The output destination may also be changed depending on a detected predetermined keyword. For example, if a keyword that is likely to be uttered when a customer is angry is detected, the output destination may be changed to the store manager, but this is not limited to this.

[0062] The customer service assistance notice may include the content of the conversation between the first salesperson and the customer as customer service information. The content of the conversation may be converted into text and output.

[0063] The text analysis unit 140 may also determine the classification of the conversation content based on the text analysis results. The classification of the conversation content may be determined by a determination unit (not shown). Therefore, the text analysis unit 140 and the determination unit function as a determination means for determining the classification of the conversation content based on the results of the text analysis means. Concretely, the classification of the conversation content may include, but is not limited to, product proposals, precautions, handling instructions, and casual conversation. One example of a method for determining the classification of the conversation content is to pre-register words used during customer service and determine the classification of the conversation content based on the detected words. For example, words frequently used during customer service may be pre-registered in groups according to conversation content, such as product proposals and handling instructions, and the classification of the conversation content may be determined based on the group to which the most frequently detected words belong within a predetermined time period. Alternatively, the classification of the conversation content may be determined by counting the number of times each word belonging to each group is detected within a predetermined time period and determining the group with the most frequent detections. Furthermore, if the number of times a pre-registered word used during customer service is detected within a predetermined time period is less than a predetermined number, the classification of the conversation content may be determined as casual conversation. If the conversation content is determined to be classified as a product proposal, the proposed product may be identified. Furthermore, the identified product may be stored as a proposed product and output as customer service information included in the customer service support notification.

[0064] Furthermore, if the classification of the conversation content is determined to be casual conversation, the output of the customer service assistance notification may not be instructed. When the configuration is such that the output of the customer service assistance notification is not instructed based on the classification of the conversation content, the text analysis unit 140 may be configured not to instruct the output of the customer service assistance notification, or the notification instructing unit 130 may be configured not to instruct the output of the customer service assistance notification. Alternatively, a determination unit (not shown) may be configured not to instruct the output of the customer service assistance notification. Furthermore, if the classification of the conversation content is determined to be casual conversation, the determination conditions for outputting the customer service assistance notification may be changed. This makes it possible to suppress erroneous notifications for parts of the conversation that are not related to products.

[0065] The operation of the customer service support system according to the third embodiment will be described with reference to the drawings. Fig. 9 is a flowchart showing an example of the operation of the customer service support system 3.

[0066] The information acquisition unit 110 acquires an image of the first store clerk and the customer (step S20), and acquires the voices of the first store clerk and the customer (step S21). However, the processing order of steps S20 and S21 may be reversed, or steps S20 and S21 may be performed in parallel.

[0067] The emotion analysis unit 120 analyzes the emotions between the first clerk and the customer based on the acquired image (step S22), and the text analysis unit 140 performs text analysis of the conversation between the first clerk and the customer based on the acquired audio (step S23). However, the order of steps S22 and S23 may be reversed, or they may be performed in parallel. Alternatively, steps S20 and S22 may be performed first, followed by steps S21 and S23, or the reverse order may be used. Alternatively, steps S20 and S22 and steps S21 and S23 may be performed in parallel.

[0068] The emotion analysis unit 120 determines whether to output a customer service assistance notice based on the emotion analysis results of the first clerk and the customer (step S24). Step S24 may be performed by a determination unit (not shown) or by the notification instruction unit 130. If it is determined that a customer service assistance notice should be output based on the emotion analysis results of the first clerk and the customer (YES in step S24), the notification instruction unit 130 instructs the second clerk to output a customer service assistance notice (step S25), and the process returns to step S20.

[0069] On the other hand, if it is determined not to output a customer service assistance notice based on the emotion analysis results of the first clerk and the customer (NO in step S24), the text analysis unit 140 determines whether the conversation between the first clerk and the customer contains a predetermined keyword (step S26). If the conversation between the first clerk and the customer contains a predetermined keyword (YES in step S26), the notification instruction unit 130 instructs the second clerk to output a customer service assistance notice (step S27), and the process returns to step S20.

[0070] On the other hand, if the conversation between the first store clerk and the customer does not contain the predetermined keyword (NO in step S26), the process returns to step S20. However, the order of steps S24 and S26 may be reversed. That is, step S26 may be executed after step S23, and step S24 may be executed if step S26 is NO.

[0071] The customer service assistance system according to the third embodiment can output a customer service assistance notification when customer service assistance is required because it determines whether to output a customer service assistance notification based on the results of text analysis in addition to the results of sentiment analysis, and instructs the output of the customer service assistance notification when it determines that the customer service assistance notification should be output.

[0072] Furthermore, in the customer service assistance system according to the third embodiment, it is possible to determine the classification of the conversation content from the text analysis results, and if the classification of the conversation content is determined to be casual conversation, it is possible not to instruct the output of a customer service assistance notification. This makes it possible to prevent erroneous notifications of customer service assistance notifications. The reason for this is that whether or not to instruct the output of a customer service assistance notification is changed depending on whether the conversation is about a product or casual conversation unrelated to the product.

[0073] [Modification of the third embodiment] A modification of the third embodiment will now be described. In the modification of the third embodiment, the operation of the customer service support system is different. Fig. 10 is a flowchart showing a modification of the operation of the customer service support system 3.

[0074] The operations from step S20 to step S23 are the same as those described in FIG. 9 . After step S23, the emotion analysis unit 120 determines whether to output a customer service assistance notice based on the emotion analysis results of the first clerk and the customer (step S34). Step S34 may be performed by a determination unit (not shown) or the notification instruction unit 130. If it is determined that a customer service assistance notice should be output based on the emotion analysis results of the first clerk and the customer (YES in step S34), the text analysis unit 140 determines whether a predetermined keyword is included in the conversation between the first clerk and the customer (step S35). If the predetermined keyword is included in the conversation between the first clerk and the customer (YES in step S35), the notification instruction unit 130 instructs the second clerk to output a customer service assistance notice (step S36), and the process returns to step S20. However, the order of steps S34 and S35 may be reversed.

[0075] On the other hand, if it is determined not to output a customer service support notification based on the emotion analysis results of the first clerk and the customer (NO in step S34), or if the conversation between the first clerk and the customer does not contain a predetermined keyword (NO in step S35), the process returns to step S20.

[0076] In the customer service assistance system according to the modification of the third embodiment, a customer service assistance notice can be output when the need for customer service assistance is high. This is because the system determines whether to output a customer service assistance notice based on the results of text analysis in addition to the results of sentiment analysis, and instructs the system to output the customer service assistance notice when it determines that the customer service assistance notice should be output.

[0077] More specifically, in the customer service support system according to the modification of the third embodiment, when it is determined that a customer service support notice should be output based on the emotion analysis results of the first store clerk and the customer, and when a predetermined keyword is detected, the customer service support notice is output. When it is determined that a customer service support notice should be output based on the emotion analysis results of the first store clerk and the customer, and when a predetermined keyword is detected, there is often a high need for support. Therefore, the customer service support system according to the modification of the third embodiment can appropriately output a customer service support notice when there is a high need for customer service support.

[0078] [Fourth embodiment] Customer service support system 4 in the fourth embodiment differs from customer service support system 3 in the third embodiment in that it further includes a result output unit. Hereinafter, the same components as those in the third embodiment will be assigned the same reference numerals and descriptions thereof will be omitted.

[0079] The overview of customer service support system 4 according to the fourth embodiment of the present invention is the same as that shown in Fig. 3. However, customer service support system 4 differs from customer service support system 3 according to the third embodiment in that it further includes result output unit 150. Result output unit 150 serves as result output means for outputting emotion analysis results to the first store clerk.

[0080] An analysis system 100 according to a fourth embodiment will now be described. FIG. 11 is a block diagram showing an example of the configuration of the analysis system 100 according to the fourth embodiment. In FIG. 11, each block shows the configuration of a functional unit. Therefore, the blocks shown in FIG. 11 may be implemented in a single device, or may be implemented separately in multiple devices. The analysis system 100 according to the fourth embodiment includes an information acquisition unit 110, a sentiment analysis unit 120, a notification instruction unit 130, a text analysis unit 140, and a result output unit 150. However, the configuration of the analysis system 100 is not limited to this.

[0081] Furthermore, the notification instruction unit 130 of the analysis system 100 may perform the function of the result output unit 150. The analysis system 100 may be implemented in the store clerk terminal 200, or part of the configuration of the analysis system 100 may be implemented in the store clerk terminal 200.

[0082] The following describes the result output unit 150. The result output unit 150 outputs the emotion analysis result from the emotion analysis unit 120 and the text analysis result from the text analysis unit 140.

[0083] The output of the emotion analysis results and text analysis results can be divided into, for example, real-time analysis results that are output in real time during customer service, and customer service results that are output after the customer service has ended, but the output content is not limited to these.

[0084] Examples of real-time analysis results that can be displayed include graphs of emotion analysis results, customer service support comments based on the analysis results, and whether or not specific keywords have been detected. However, it is not necessary to display all of these, and other displays may also be used.

[0085] Examples of the customer service results include the customer service score calculated from the analysis results, whether or not a suggested product was purchased, the number of times a specific keyword was detected, and the number of customer service support notifications. Other information that may be displayed include the date and time, the person in charge, the length of the customer service, customer information, points that can be evaluated, and points that need improvement. It is not necessary to display all of the information described here regarding the customer service results, and other information may also be displayed.

[0086] The emotion analysis results and text analysis results are output to, for example, the clerk terminal 200 used by the first clerk serving the customer. Alternatively, the emotion analysis results and text analysis results may be output to the clerk terminal 200 used by a second clerk different from the first clerk. When the emotion analysis results and text analysis results are output to the clerk terminal 200 used by the second clerk, the second clerk can use the emotion analysis results and text analysis results to evaluate the customer service of the first clerk.

[0087] Furthermore, the result output unit 150 may output the real-time analysis results to the clerk terminal 200 used by the first clerk, and after the customer service is over, output the customer service results to the clerk terminal 200 used by the first clerk and the clerk terminal 200 used by the second clerk. In this case, the first clerk can serve the customer while checking the real-time analysis results, and after the customer service is over, can reflect on his or her own customer service by checking the customer service results. Furthermore, the second clerk can evaluate the customer service of the first clerk based on the customer service results.

[0088] The result output unit 150 outputs an emotion analysis result comparison screen that chronologically displays the emotion analysis results of the first sales clerk and the customer on the same screen. FIG. 12 is an example of a real-time analysis result screen output. The real-time analysis result screen shown in FIG. 12 displays an emotion analysis result graph 410 and a support comment section 440. The emotion analysis result graph 410 is a graph with customer service time on the horizontal axis and emotion analysis results on the vertical axis. The emotion analysis result graph 410 displays the emotion analysis results 420-1 of the first sales clerk and the emotion analysis results 420-2 of the customer on the same screen in chronological order. Displaying the emotion analysis results 420-1 of the first sales clerk and the emotion analysis results 420-2 of the customer on the same screen in chronological order is convenient because it allows the user to recognize at a glance the discrepancy between the emotion analysis results 420-1 of the first sales clerk and the emotion analysis results 420-2 of the customer. Furthermore, the emotion analysis result graph 410 displays a conversation content classification 430. These display contents are merely examples, and the contents of the real-time analysis results that are output on the screen are not limited to these.

[0089] The conversation content classification 430 indicates whether the conversation content is casual conversation or a product proposal. In the case of a product proposal, the proposed product name and image may be displayed. Whether the conversation content is casual conversation or a product proposal may be determined based on the results of text analysis.

[0090] The black circles in the emotion analysis result graph 410 indicate the timing at which a predetermined keyword (shown as an NG word in FIG. 12) was detected.

[0091] The support comment field 440 displays comments to support the store clerk serving the customer based on the analysis results. Examples of displayed content include whether the customer is interested, whether a specific keyword was detected, and whether a customer service support notification was issued, but the content is not limited to these, as long as the comment supports the customer service. In addition, the frame shape and color, font and color of the text, etc. of the support comment may be changed depending on the content. For example, positive content such as "the customer is interested" may be displayed in a warm color such as orange, and negative content such as "the customer has lost interest" may be displayed in a cool color such as blue.

[0092] Additionally, the support comment field 440 may display predictions of emotional changes and suggestions of words and topics that are in line with the flow of the conversation. By predicting emotional changes and suggesting appropriate words and topics, the customer's emotions can be maintained in a good state.

[0093] Figure 13 is an example of the screen output of the customer service results. The customer service results screen shown in Figure 13 displays the date of the customer service 510, a video of the customer service 520, a comparison of the emotion analysis results with the average value 530, an emotion analysis result graph 540, and emotion analysis results for a short period of time 550. Figure 13 displays examples of emotion analysis items such as "Happy," "Attention," "Fear," "Confusion," "Surprise," "Disgust," "Empathy," and "Contempt," but the analysis items are not limited to these.

[0094] The emotion analysis result graph 540 is a graph with customer service time on the horizontal axis and emotion analysis results on the vertical axis, and displays a graph for each emotion analysis item. When emotion analysis results are obtained for multiple emotion analysis items, the emotion with the highest score among the emotion analysis results at a certain time may be determined as the emotion at that time. Also, a single emotion analysis result may be obtained by statistically processing the results of each of the multiple emotion analysis items.

[0095] The short-term emotion analysis results 550 show the proportion of each emotion for each period. The period may be set, for example, by dividing the period into predetermined time intervals, or by classification of conversation content based on the text analysis results. Specific examples of dividing the period into classifications of conversation content include "casual conversation," "proposal of product I," and "proposal of product II."

[0096] Additionally, a customer service score may be calculated and displayed as a customer service result. Examples of customer service scores include a salesperson customer service score, an overall customer service score, and a product customer service score. The salesperson customer service score is calculated from the results of an analysis of the customer's emotions. The overall customer service score is calculated from the results of an analysis of the customer's emotions during the period from the start of the customer service to the end of the customer service. The product customer service score is calculated from the results of an analysis of the customer's emotions during the conversation about the product during the period from the start of the customer service to the end of the customer service.

[0097] The customer service score is calculated using statistical values ​​such as the average value of emotional transitions, the maximum value, and the minimum value of emotional transitions. The customer service score displayed in the customer service results may be the average value, minimum value, maximum value, a comparison of the most recent multiple customer service scores, a comparison of customer service scores for the same customer, etc. The customer service score may also be calculated and displayed for each emotion analysis item.

[0098] The customer service result may also include text data of the conversation that took place during the customer service. Furthermore, the customer service result may be output to the clerk terminals 200 of all the clerks other than the clerk who served the customer and his / her superior. When the customer service result is output to the clerk terminals 200 of all the clerks, the customer service skills can be easily shared.

[0099] The customer service support system according to the fourth embodiment can provide customer service support to store clerks who are serving customers because it can output the analysis results in real time so that the store clerk can check them and provide customer service support information that is useful for customer service.

[0100] It is difficult for one salesperson to grasp the customer service situation of another salesperson who is serving a customer, so it can be difficult for a supervisor or manager to evaluate the customer service of the salesperson, but the customer service support system according to the fourth embodiment makes it possible to appropriately evaluate the customer service of the salesperson. This is because the customer service results can be output to the supervisor of the salesperson who served the customer.

[0101] Furthermore, the customer service support system according to the fourth embodiment is convenient for the first salesperson to review his or her customer service because the emotion analysis results of the first salesperson and the customer are displayed in chronological order on the same screen.

[0102] The present invention can be applied not only to customer service in brick-and-mortar stores but also to online customer service, where it is often difficult to grasp the customer service situation of each store clerk.

[0103] For example, when the fourth embodiment of the present disclosure is applied to online customer service, it becomes easier for the second salesperson to provide customer service support to the first salesperson, and it also becomes easier to appropriately evaluate each salesperson. This is because it is possible to output at least one of a customer service support notice, customer service information, real-time analysis results of the first salesperson, and customer service results of the first salesperson to the second salesperson, whose customer service situation is difficult to grasp.

[0104] In each of the embodiments of the present invention described above, each component of the customer service support system represents a functional block.

[0105] The processing of each component may be realized, for example, by a computer system reading and executing a program stored in a computer-readable storage medium that causes the computer system to execute that processing. A "computer-readable storage medium" includes, for example, portable media such as optical disks, magnetic disks, magneto-optical disks, and non-volatile semiconductor memory, as well as storage devices built into the computer system, such as read-only memory (ROM) and hard disks. A "computer-readable storage medium" also includes media that dynamically store programs for a short period of time, such as communication lines used when transmitting programs over networks such as the Internet or telephone lines, and media that temporarily store programs, such as volatile memory within computer systems serving as servers or clients. The program may also be designed to realize some of the functions described above, or may be capable of realizing the functions described above in combination with programs already stored in the computer system.

[0106] The "computer system" is, for example, a system including a computer 900 having the following configuration. ·CPU(Central Processing Unit)901 ROM902 ·RAM(Random Access Memory)903 Program 904A and storage information 904B loaded into RAM 903 A storage device 905 for storing programs 904A and storage information 904B A drive device 907 that reads and writes from the storage medium 906 A communication interface 908 for connecting to a communication network 909 Input / output interface 910 for inputting and outputting data Bus 911 connecting each component For example, each component of each device in each embodiment is realized by CPU 901 loading program 904A that realizes the function of that component into RAM 903 and executing it. Program 904A that realizes the function of each component of each device is stored in advance in storage device 905 or ROM 902, for example. CPU 901 then reads out program 904A as needed. Storage device 905 is, for example, a hard disk. Program 904A may be supplied to CPU 901 via communication network 909, or may be stored in storage medium 906 in advance, read out by drive device 907, and supplied to CPU 901. Storage medium 906 is, for example, a portable medium such as an optical disk, a magnetic disk, a magneto-optical disk, or a non-volatile semiconductor memory.

[0107] There are various variations in the method of realizing each device. For example, each device may be realized by a possible combination of a separate computer 900 and a program for each component. Also, multiple components included in each device may be realized by a possible combination of a single computer 900 and a program.

[0108] Furthermore, some or all of the components of each device may be realized by other general-purpose or dedicated circuits, computers, etc., or a combination of these. These may be configured by a single chip, or by multiple chips connected via a bus.

[0109] When some or all of the components of each device are realized by multiple computers, circuits, etc., the multiple computers, circuits, etc. may be centrally located or distributed. For example, the computers, circuits, etc. may be realized as a client-server system, a cloud computing system, or the like, in a form in which each is connected via a communication network.

[0110] The present disclosure is not limited to the above-described embodiments, and various modifications are possible. Embodiments obtained by appropriately combining the configurations, operations, and processes disclosed in different embodiments are also included in the technical scope of the present disclosure.

[0111] The present disclosure is not limited to the above-described embodiments, and various aspects that can be understood by a person skilled in the art can be applied to the present disclosure within the scope of the present disclosure.

[0112] [Appendix 1] emotion analysis means for analyzing emotions between a first store clerk and a customer based on an image of the first store clerk and the customer; a notification instruction means for instructing a second store clerk, different from the first store clerk, to output a customer service assistance notification based on the emotion analysis result of the first store clerk and the emotion analysis result of the customer; A customer service support system equipped with: [Appendix 2] the notification instruction means instructs a second store clerk different from the first store clerk to output a customer service assistance notification when there is a discrepancy between the emotion analysis result of the first store clerk and the emotion analysis result of the customer that is equal to or greater than a predetermined value. Attachment 1: A customer service support system. [Appendix 3] further comprising a text analysis means for analyzing a conversation between the first store clerk and the customer; the notification instruction means instructs output of the customer service assistance notification when the text analysis means detects a predetermined keyword. 3. A customer service support system according to claim 1 or 2. [Appendix 4] Further, a determination means is provided for determining a classification of the conversation content based on the result of the text analysis means, not instructing the output of the customer service assistance notification based on the determination result of the determination means; A customer service support system as described in Appendix 3. [Appendix 5] The customer service assistance notice includes customer service information regarding the customer service provided by the first store clerk to the customer. 5. A customer service support system according to any one of appendices 1 to 4. [Appendix 6] further comprising a result output means for outputting a result of the emotion analysis to the first store clerk; 6. A customer service support system according to any one of appendices 1 to 5. [Appendix 7] the result output means outputs an emotion analysis result comparison screen that chronologically displays the emotion analysis result of the first store clerk and the emotion analysis result of the customer on the same screen. A customer service support system as described in Appendix 6. [Appendix 8] It is characterized by being applicable to online customer service. 8. A customer service support system according to any one of appendices 1 to 7. [Appendix 9] an emotion analysis unit that analyzes emotions between a first store clerk and a customer based on an image of the first store clerk and the customer; a notification instruction unit that instructs a second store clerk, different from the first store clerk, to output a customer service assistance notification based on the emotion analysis result of the first store clerk and the emotion analysis result of the customer; A customer service support device comprising: [Appendix 10] Analyzing emotions between a first store clerk and a customer based on an image of the first store clerk and the customer; instructing a second store clerk different from the first store clerk to output a customer service assistance notification based on the emotion analysis result of the first store clerk and the emotion analysis result of the customer; Customer service support methods. [Appendix 11] Analyzing emotions between a first store clerk and a customer based on an image of the first store clerk and the customer; instructing a second store clerk different from the first store clerk to output a customer service assistance notification based on the emotion analysis result of the first store clerk and the emotion analysis result of the customer; A program that makes a computer do something. [Explanation of symbols]

[0113] 10 Sentiment analysis tools 20 Notification instruction means 100 Analysis Systems 110 Information Acquisition Department 120 Emotion Analysis Department 130 Notification instruction section 140 Text Analysis Department 150 Result output section 200 Clerk terminal 300 Network 410 Sentiment analysis result graph 420-1 Emotion analysis results of the first store clerk 420-2 Customer sentiment analysis results 430 Classification of conversation content 440 Support Comments 510 days of service 520 Video of customer service 530 Comparison of sentiment analysis results with the average 540 Sentiment analysis result graph 900 Computers 901 CPU 902 ROM 903 RAM 904A Program 904B Memory Information 905 Storage device 906 Storage medium 907 Drive unit 908 Communication Interface 909 Communication Network 910 Input / Output Interface 911 Bus

Claims

1. emotion analysis means for analyzing emotions between a first store clerk and a customer based on an image of the first store clerk and the customer; a text analysis means for analyzing a conversation between the first store clerk and the customer; a determination means for determining a classification of the conversation content based on the result of the text analysis means; a notification instruction means for instructing a second store clerk different from the first store clerk to output a customer service assistance notification when the emotion analysis result of at least one of the first store clerk and the customer is negative or there is a predetermined or greater change to the negative side based on the emotion analysis result of the first store clerk and the emotion analysis result of the customer, The notification instruction means instructs the output of the customer service support notification when the text analysis means detects a predetermined keyword, and does not instruct the output of the customer service support notification when the classification of the conversation content is determined to be casual conversation based on the judgment result of the judgment means.

2. 2. The customer service support system according to claim 1, wherein the notification instruction means instructs output of a customer service support notification to a second store clerk different from the first store clerk when there is a discrepancy between the emotion analysis result of the first store clerk and the emotion analysis result of the customer that is greater than a predetermined value.

3. The customer service support system according to claim 1 , wherein the customer service support notice includes customer service information relating to the service provided by the first store clerk to the customer.

4. The customer service support system according to claim 1 , further comprising a result output unit that outputs the emotion analysis result to the first store clerk.

5. 5. The customer service support system according to claim 4, wherein the result output means outputs an emotion analysis result comparison screen that chronologically displays the emotion analysis result of the first store clerk and the emotion analysis result of the customer on the same screen.

6. an emotion analysis unit that analyzes emotions between a first store clerk and a customer based on an image of the first store clerk and the customer; a text analysis unit that analyzes a conversation between the first store clerk and the customer; a determination unit that determines a classification of the conversation content based on the result of the text analysis unit; a notification instruction unit that instructs a second store clerk other than the first store clerk to output a customer service assistance notification when the emotion analysis result of at least one of the first store clerk and the customer is negative or there is a predetermined or greater change to the negative side based on the emotion analysis result of the first store clerk and the emotion analysis result of the customer, The notification instruction unit instructs the output of the customer service assistance notification when the text analysis unit detects a predetermined keyword, and does not instruct the output of the customer service assistance notification when the classification of the conversation content is determined to be casual conversation based on the judgment result of the judgment unit.

7. A computer comprising: Analyzing emotions between a first store clerk and a customer based on an image of the first store clerk and the customer; performing a text analysis of the conversation between the first store clerk and the customer by detecting pre-registered words used during the conversation; determining a classification of the conversation content based on the detected words from the results of the text analysis; instructing a second sales clerk different from the first sales clerk to output a customer service assistance notification when the emotion analysis result of at least one of the first sales clerk and the customer is negative or when there is a predetermined or greater change to the negative side based on the emotion analysis result of the first sales clerk and the emotion analysis result of the customer; A customer service support method that instructs the output of the customer service support notification when a predetermined keyword is detected as a result of the text analysis, and does not instruct the output of the customer service support notification when it is determined that the conversation content is classified as casual conversation based on the result of the judgment.

8. Analyzing emotions between a first store clerk and a customer based on an image of the first store clerk and the customer; performing a text analysis of the conversation between the first store clerk and the customer by detecting pre-registered words used during the conversation; determining a classification of the conversation content based on the detected words from the results of the text analysis; instructing a second sales clerk different from the first sales clerk to output a customer service assistance notification when the emotion analysis result of at least one of the first sales clerk and the customer is negative or when there is a predetermined or greater change to the negative side based on the emotion analysis result of the first sales clerk and the emotion analysis result of the customer; A program that causes a computer to instruct the output of the customer service assistance notification if a specified keyword is detected as a result of the text analysis, and not instruct the output of the customer service assistance notification if the conversation content is determined to be classified as casual conversation based on the result of the judgment.

Images