Work evaluation system

Abstract

This work evaluation system includes: a work state collection unit that collects the results of work performed by employees of a company; a company information management unit that stores company information including the corporate ideology of the company; and an evaluation unit that evaluates the work of the employees using the work results and the company information.

Description

Work evaluation system

[0001] The present invention relates to a system for evaluating the performance of a customer service employee.

[0002] Patent Literature 1 describes a customer service attitude evaluation system, which evaluates customer service attitudes by inputting the length of speech of specific keywords and tone of voice and comparing them with standards for good customer service attitudes.

[0003] JP 2017-4224 A

[0004] However, conventional customer service attitude evaluation systems such as those disclosed in Patent Document 1 do not evaluate customers in a way that reflects the management philosophy, including the customer service philosophy, that the store or company being evaluated has independently decided.

[0005] Therefore, an object of the present invention is to evaluate customer service, including whether it satisfies the customer service philosophy and management philosophy that each store or company has independently determined.

[0006] A work evaluation system according to one embodiment of the present invention comprises a work status collection unit that collects work results performed by employees of a company, a company information management unit that stores company information including the company's corporate philosophy, and an evaluation unit that evaluates the work of employees using the work results and company information.

[0007] In this configuration, an evaluation is made using the actual work situation of the employee and the corporate information of the company where the employee works, so that the employee's work can be evaluated not only based on the employee's work situation but also using the corporate philosophy stored as corporate information.

[0008] According to this invention, it is possible to evaluate whether employees are performing their duties in accordance with the corporate philosophy that each company has independently determined.

[0009] FIG. 1 is a functional block diagram showing the basic functional configuration of a customer service evaluation system according to a first embodiment of the present invention. FIG. 2 is a diagram showing an example of the functional and physical configuration of the customer service evaluation system according to the first embodiment of the present invention in more detail. FIG. 3 is a diagram showing an example of the physical configuration of the customer service evaluation system according to the first embodiment of the present invention. FIG. 4 is a diagram showing a schematic configuration of a processing device. FIG. 5 is a diagram showing an example of a store philosophy. FIG. 6(A) is a flowchart showing the processing content of the evaluation unit 13, showing an example of a customer service evaluation according to the first embodiment, and FIG. 6(B) is a flowchart showing a more specific example of the processing for calculating the customer service evaluation in FIG. 6(A). FIG. 7 is a text version of a conversation between an employee with a high customer service evaluation and a customer. FIG. 8 is a text version of a conversation between an employee with a customer whose customer service evaluation is not high (e.g., average). FIG. 9 is a diagram showing an example of a customer service evaluation of an employee with a high customer service evaluation. FIG. 10 is a diagram showing an example of a customer service evaluation of an employee with a low customer service evaluation. FIG. 11 is a flowchart showing the processing content of the evaluation unit 13 and the improvement plan generation unit 14, showing an example of a case where the generation of a customer service evaluation and the generation of an improvement plan are performed in a single process. FIG. 12 is a diagram showing an example of improvements and improvement proposals. FIG. 13 is a flowchart showing the processing details of the feedback unit 16, showing an example of feedback of the improvement proposals. FIG. 14 is a diagram showing an example of interactive feedback of improvements. FIG. 15 is a diagram showing a more specific example of the functional and physical configurations of a customer service evaluation system according to a second embodiment of the present invention. FIG. 16 is a diagram showing an example of the physical configuration of a customer service evaluation system according to the second embodiment of the present invention. FIG. 17(A) is a flowchart showing the processing details of the evaluation unit 13, showing an example of a customer service evaluation according to the second embodiment, and FIG. 17(B) is a flowchart showing a more specific example of the processing for calculating the customer service evaluation in FIG. 17(A). FIG. 18 is a functional block diagram showing the basic functional configuration of a customer service evaluation system according to a third embodiment of the present invention. FIG. 19 is a diagram showing a more specific example of the functional and physical configurations of a customer service evaluation system according to the third embodiment of the present invention.FIG. 20(A) is a flowchart illustrating the processing of the evaluation unit 13S, showing an example of a customer service evaluation according to the third embodiment. FIG. 20(B) is a flowchart illustrating a more specific example of the processing for calculating the customer service evaluation in FIG. 20(A). FIG. 21 is a flowchart illustrating the processing of the evaluation unit 13, showing an example of a customer service evaluation according to the fourth embodiment. FIG. 22 is a functional block diagram illustrating the basic functional configuration of a customer service evaluation system according to a fifth embodiment of the present invention. FIG. 23 is a diagram illustrating a more specific example of the functional and physical configurations of the customer service evaluation system according to the fifth embodiment of the present invention. FIG. 24 is a diagram illustrating a general flow of re-learning the customer service evaluation model in the customer service evaluation system according to the fifth embodiment of the present invention. FIGS. 25(A) and 25(B) are diagrams illustrating an example of a method for evaluation (re-evaluation) by a superior manager. FIG. 26 is a flowchart illustrating an example of a re-learning method for the customer service evaluation model. FIGS. 27(A) and 27(B) are diagrams illustrating a learning method that reflects superior manager evaluation information and a method for executing customer service evaluation. FIG. 28 is a functional block diagram illustrating the basic functional configuration of a customer service evaluation system according to a sixth embodiment of the present invention. FIG. 29 is a diagram showing in more detail an example of the functional and physical configurations of a customer service evaluation system according to a sixth embodiment of the present invention. FIG. 30 is a functional block diagram showing the basic functional configuration of a customer service evaluation system according to a seventh embodiment of the present invention. FIG. 31 is a functional block diagram showing the basic functional configuration of a customer service evaluation system according to an eighth embodiment of the present invention. FIG. 32 is a diagram showing in more detail an example of the functional and physical configurations of a customer service evaluation system according to the eighth embodiment of the present invention. FIG. 33 is a diagram showing an example of the relationship between customer service observation data (audio text data), categories, and philosophies. FIG. 34 is a diagram showing an example of classification of multiple audio text data into categories. FIG. 35 is a diagram showing an example of notification of customer service evaluation results and improvement proposals. FIG. 36(A) is a flowchart showing an example of processing executed by a classifier in a customer service evaluation method according to the eighth embodiment, and FIG. 36(B) is a flowchart showing an example of processing executed by an evaluator in a customer service evaluation method according to the eighth embodiment.FIG. 37 is a flowchart showing a first example of advance association processing in the customer service evaluation method according to the eighth embodiment. FIG. 38 is a flowchart showing a second example of advance association processing in the customer service evaluation method according to the eighth embodiment. FIG. 39 is a functional block diagram showing the basic functional configuration of a business evaluation system according to a ninth embodiment of the present invention. FIG. 40 is a diagram showing in more detail an example of the functional and physical configurations of a business evaluation system according to the ninth embodiment of the present invention. FIG. 41 is a flowchart showing an example of a business result evaluation method according to the ninth embodiment. FIG. 42 is a table showing examples of evaluation targets and application modes that can be implemented in the business evaluation system.

[0010] [First embodiment] A customer service evaluation system according to a first embodiment of the present invention will be described with reference to the drawings. In the first embodiment, a customer service evaluation system will be described as an example of a business evaluation system.

[0011] (Basic Configuration of Customer Service Evaluation System) FIG. 1 is a functional block diagram showing the basic functional configuration of a customer service evaluation system according to a first embodiment of the present invention.

[0012] 1 , customer service evaluation system 10 is functionally configured to include customer service status collection unit 11, store information management unit 12, evaluation unit 13, improvement plan generation unit 14, notification unit 15, and feedback unit 16. Note that customer service evaluation system 10 can omit improvement plan generation unit 14, notification unit 15, and feedback unit 16 up to the point where customer service evaluation is performed, and is therefore configured solely by customer service status collection unit 11, store information management unit 12, and evaluation unit 13.

[0013] The basic process of the customer service evaluation system 10 is to evaluate the customer service of employees, create improvement plans, and provide feedback to employees as follows.

[0014] The customer service status collection unit 11 (work status collection unit) acquires voices relating to customer service between employees and customers in the store (including, for example, conversational voices (voices of conversations during customer service)). The customer service status collection unit 11 outputs the customer service voices (voices relating to customer service, which correspond to work results) to the evaluation unit 13.

[0015] The store information management unit 12 (corporate information management unit) stores store information (corporate information). The store information includes the store's customer service philosophy or the management philosophy related to the store's management. The customer service philosophy does not specify detailed behaviors such as how a store specifically provides customer service (for example, the angle at which an employee bows when greeting a customer as they leave the store), but rather indicates more conceptual customer service goals for the store. Specific examples of customer service philosophy will be described later. The management philosophy represents the basic policy for all corporate activities, such as the purpose of management. Like the customer service philosophy, the management philosophy is not specific, but rather conceptual, such as values ​​and ways of thinking. The store philosophy and management philosophy can be collectively referred to as the corporate philosophy.

[0016] The store information management unit 12 stores and manages store information in text data format (store information text data). In response to a request from the evaluation unit 13, the store information management unit 12 outputs the store information text data to the evaluation unit 13. The store information text data is an example of company information text data.

[0017] The evaluation unit 13 evaluates the customer service performance of the employee using the customer service voice and store information. More specifically, the evaluation unit 13 converts the customer service voice into text data format (speech text data). The evaluation unit 13 receives the speech text data and the store information text data as input and performs a customer service evaluation using a customer service evaluation model that uses a large-scale language model. The speech text data is an example of business result text data.

[0018] In this way, customer service evaluation system 10 performs customer service evaluation (job evaluation) using voice (voice text data) reflecting the customer service situation and store information (store information text data) that represents the store's ideal customer service. Therefore, customer service evaluation system 10 can evaluate customer service not only based on the customer service voice, but also on whether the customer service philosophy and management philosophy decided by the store are met.

[0019] The evaluation unit 13 outputs the customer service evaluation of the employee to the notification unit 15. The notification unit 15 notifies the customer service evaluation using an image or sound.

[0020] The improvement plan generating unit 14 generates an improvement plan for the customer service attitude of the employee using the customer service evaluation.

[0021] In response to an inquiry from an employee or the like, the feedback unit 16 provides feedback on an improvement plan to the inquirer. For example, when the feedback unit 16 receives a request from an employee for areas that need improvement, it generates a response to the areas that need improvement based on the improvement plan. The feedback unit 16 notifies the employee of the generated response by image or sound.

[0022] (Specific Configuration of Customer Service Evaluation System) Fig. 2 is a diagram showing in more detail an example of the functional and physical configuration of the customer service evaluation system according to the first embodiment of the present invention. Fig. 3 is a diagram showing an example of the physical configuration of the customer service evaluation system according to the first embodiment of the present invention.

[0023] 2 and 3 , customer service evaluation system 10 includes data cloud 100, information communication terminal 93, microphone 111, and store communication device 119. Data cloud 100 includes a processing device 91, a recording device 92, and a customer service evaluation DB 99. Information communication terminal 93 includes information communication terminal 93A and information communication terminal 93B.

[0024] The arithmetic processing device 91 functionally comprises a text data generation unit 911, an evaluation unit 13, and an improvement plan generation unit 14. The arithmetic processing device 91, the recording device 92, the information communication terminals 93A and 93B, and the customer service evaluation DB 99 have computer functions and are configured as shown in FIG. 4. The data cloud 100 has the arithmetic processing device 91, the recording device 92, and the customer service evaluation DB 99 as functions, and is configured as a single computer function. Note that the data cloud 100 may be configured as multiple data clouds, each of which is configured as a computer function.

[0025] FIG. 4 is a diagram showing a schematic configuration for implementing computer functions. As shown in FIG. 4, the computer includes a CPU, memory, a storage medium, an input / output interface (IF), and a communication interface. In the case of the arithmetic processing device 91, for example, the text data generation unit 911, the evaluation unit 13, and the improvement plan generation unit 14 are realized by the CPU executing programs related to the customer service evaluation process and the improvement plan generation process stored in a storage medium (e.g., SSD, HDD). The memory is used, for example, when the CPU executes the programs. The input / output interface is a connection interface with input devices such as an external display or keyboard. The communication interface is a connection interface with the network 900.

[0026] The arithmetic processing device 91 is placed in a data cloud 100 separate from the store, as shown in Fig. 3. The arithmetic processing device 91 may be configured as a computer placed in a headquarters building (for example, the headquarters building of a management company that runs the store) or in the store, rather than in the data cloud 100.

[0027] The recording device 92 is hardware configured from a recording medium. The recording medium that configures the recording device 92 has, for example, a relatively large capacity. Store information text data and customer service observation data are recorded in the recording device 92. The customer service observation data is configured to include voice text data (voice text data of customer service voices) during customer service.

[0028] The recording device 92 is located in the data cloud 100, as shown in FIG. 3, for example, similar to the arithmetic processing device 91, but may also be located in a headquarters building or a store.

[0029] The information communication terminal 93A includes an information communication terminal control unit (e.g., a CPU), a display unit (e.g., a liquid crystal display), an operation unit (e.g., a touch panel), a microphone, and a speaker. The information communication terminal 93A is, for example, a smartphone or a tablet terminal.

[0030] The information communication terminal 93A is held by, for example, an employee to be evaluated, an evaluator who performs the evaluation, etc. The information communication terminal 93A is connected to the network 900 wirelessly or the like.

[0031] The information communication terminal 93B includes an information communication terminal control unit (e.g., a CPU), a storage unit (e.g., a memory), a display device 912 (e.g., a liquid crystal display), an operation unit 913 (e.g., a keyboard, a mouse), and an input / output unit 914 (e.g., a speaker, a microphone). The information communication terminal 93B is, for example, a laptop computer or a desktop computer.

[0032] The information communication terminal 93B is used, for example, by an area manager or a head office staff member who performs the evaluation. The information communication terminal 93B is connected to the network 900 wirelessly or the like. Here, the information communication terminal 93B may be a portable terminal such as a smartphone.

[0033] The customer service evaluation DB 99 is configured, for example, by a server or a cloud. For example, the customer service evaluation DB 99 is provided in the data cloud 100 as described above and is connected to the network 900. The customer service evaluation DB 99 stores the store ID (store identification information), employee ID (employee identification information), the store where each employee works, and their working hours. The customer service evaluation DB 99 stores customer service evaluations that have been evaluated so far and new customer service evaluations evaluated this time for each employee (for each employee ID). Furthermore, the customer service evaluation DB 99 stores improvement proposals for each employee.

[0034] 3, the customer service evaluation DB 99 is placed in the data cloud 100 in the same manner as the arithmetic processing device 91. The customer service evaluation DB 99 may be placed in the headquarters building or in a location different from the headquarters building or the store.

[0035] The microphone 111 is attached to the employee to be evaluated, as shown in Fig. 3. The microphone 111 may be installed in the store as long as it can pick up at least the voice of the employee to be evaluated.

[0036] The store communication device 119 is installed in the store where the employee to be evaluated works. The store communication device 119 is configured with a router device or the like. The store communication device 119 is connected to the microphone 111 via wired or wireless communication. The store communication device 119 is connected to the network 900.

[0037] In this configuration, the microphone 111 and the store communication device 119 correspond to the customer service status collection unit 11 in Fig. 1. The recording device 92 and the customer service evaluation DB 99 correspond to the store information management unit 12 in Fig. 1. The information communication terminal 93B corresponds to the notification unit 15 in Fig. 1. The information communication terminal 93A corresponds to the feedback unit 16 in Fig. 1.

[0038] (Example of customer service evaluation processing) A: Pre-processing before customer service evaluation The microphone 111 picks up the voice (customer service voice) of the employee being evaluated and outputs it to the store communication device 119. The store communication device 119 transmits the customer service voice to the text data generation unit 911 of the arithmetic processing device 91 via the network 900. At this time, the store communication device 119 links the voice pickup time, store ID, and employee ID to the customer service voice and transmits it.

[0039] The text data generation unit 911 converts the customer service voice into text data to generate voice text data. The text data generation unit 911 transmits the voice text data to the recording device 92. At this time, the text data generation unit 911 associates the voice collection time, the store ID, and the employee ID with the voice text data and transmits it.

[0040] The recording device 92 records the received voice text data in association with the audio pickup time, store ID, and employee ID.

[0041] The recording device 92 also pre-records store information text data indicating store information.

[0042] Here, the store information includes the target store's customer service philosophy and management philosophy (store philosophy) and store ID. FIG. 5 is a diagram showing an example of a store's customer service philosophy. As shown in FIG. 5, the customer service philosophy is composed of a broad concept (the ideal customer service concept for the store) and a medium concept (customer service goals that include abstract concepts). The medium concept is composed of content that is more specific than the broad concept. Store philosophies often indicate concepts rather than specifying detailed employee behavior, such as employee posture or specific language use.

[0043] As an example, as shown in Figure 5, "personalized service" is listed as a major concept of the customer service philosophy, and the corresponding medium concept is "providing customized service that matches the customer's preferences and needs, creating a sense of exclusivity." Since preferences and needs differ from customer to customer, there are no set phrases for customer service in order to achieve this goal; rather, it is necessary to find out the customer's preferences and think about and implement services that will please that person.

[0044] The recording device 92 records at least the medium concepts shown in Fig. 5 as store information text data. Note that the recording device 92 may also store the major concepts shown in Fig. 5 as store information text data.

[0045] B: Specific example of customer service evaluation Figure 6 (A) is a flowchart showing the processing contents of the evaluation unit 13, showing an example of customer service evaluation in the first embodiment, and Figure 6 (B) is a flowchart showing a more specific example of the processing for calculating the customer service evaluation in Figure 6 (A).

[0046] The evaluation unit 13 of the arithmetic processing device 91 determines the store to be evaluated and the time period to be evaluated based on the working hours of the employee to be evaluated. The working hours and store of the employee to be evaluated are stored, for example, in the customer service evaluation DB 99. The evaluation unit 13 reads the working hours and store from the customer service evaluation DB 99 based on the employee ID of the employee to be evaluated.

[0047] The evaluation unit 13 acquires customer service voice (voice text data) for the time period to be evaluated from the recording device 92 (S11). The evaluation unit 13 acquires customer service philosophy (store information text data) from the recording device 92 (S12).

[0048] The evaluation unit 13 generates a customer service evaluation using a customer service evaluation model (S13). The customer service evaluation model receives customer service speech (speech text data) and a customer service philosophy (store information text data). More specifically, the evaluation unit 13 inputs the speech text data and the store information text data into a natural language-based customer service evaluation model (S131). The natural language-based customer service evaluation model is configured based on, for example, a large-scale language model. The evaluation unit 13 generates a customer service evaluation using the natural language-based customer service evaluation model (S132).

[0049] In addition, if the customer service voice and customer service philosophy are not saved as text data, the text data generation unit 911 can convert the customer service voice and customer service philosophy into text data before performing the customer service evaluation, and then the processing by the evaluation unit 13 described above can be performed.

[0050] Customer service evaluation using a customer service evaluation model that uses a large-scale language model is performed, for example, as follows.

[0051] Figure 7 is a text version of a conversation between a customer and an employee with a high customer service rating. Figure 8 is a text version of a conversation between a customer and an employee with a customer service rating that is not high (for example, average). The contents of Figures 7 and 8 will not be specifically described in the specification, but will be explained with reference to the respective figures.

[0052] A customer service evaluation model using a large-scale language model (hereinafter simply referred to as the customer service evaluation model) separates customer service philosophy (store information text data) and speech text data into the smallest units (tokens), such as phrases and phrases, and calculates the relevance of each token. From this, features related to customer service phrases and phrases are extracted. These features are used to repeatedly predict the next token, producing output in natural language. Then, phrases that are likely or frequently used by employees with high customer service ratings (phrases with high customer service ratings) are obtained. More specifically, the customer service evaluation model learns and stores in advance phrases with high customer service ratings, based on information about general customer service available for each store or on an information network. Note that the learning target is not limited to phrases, but may also be phrases. Furthermore, it is preferable to learn both phrases and phrases.

[0053] The customer service evaluation model extracts words from speech-text data based on customer service speech, and calculates feature values ​​using the extracted words from the speech-text data and words with high customer service evaluations.

[0054] For example, in the conversation (customer service speech) of an employee with a high customer service evaluation as shown in Figure 7, there are many phrases that are likely or frequently used by employees with a high customer service evaluation, based on the phrases included in the store philosophy (store information text data). For example, the conversation includes phrases such as "We are honored that you have chosen our restaurant" and "Thank you very much for enjoying your meal today. We look forward to seeing you again," which are examples of phrases that embody the phrase "We will always provide the highest quality service and give our customers the highest level of satisfaction" included in the store information text data (store philosophy).

[0055] In this case, the customer service evaluation model determines that the customer service evaluation is high.

[0056] On the other hand, in the conversation (customer service speech) of an employee whose customer service evaluation is not considered high, as shown in Figure 8, there are few or no phrases that are likely or likely to be used frequently by an employee when the customer service evaluation is high (phrases with high customer service evaluations) based on the phrases included in the store philosophy (store information text data). For example, the conversation does not include phrases such as "We are honored that you have chosen our restaurant" or "Thank you very much for enjoying your meal today. We look forward to seeing you again," which are examples of phrases that embody the phrase "We will always provide the highest quality service and give our customers the highest level of satisfaction" included in the store information text data (store philosophy). Instead, the conversation simply includes "Thank you for visiting us" or "Thank you for today."

[0057] In this case, the customer service evaluation model determines that the customer service evaluation cannot be said to be high.

[0058] The customer service evaluation generated by the evaluation unit 13 using the customer service evaluation model is composed of the level of the customer service evaluation and the reason (evaluation reason), as shown in Figures 9 and 10. The evaluation unit 13 can also express the customer service evaluation as an evaluation value (numerical value).

[0059] Figure 9 shows an example of a customer service evaluation for an employee with a high customer service evaluation. Figure 10 shows an example of a customer service evaluation for an employee who cannot be said to have a high customer service evaluation. The contents of Figures 9 and 10 will not be specifically described in the specification, but will be explained with reference to each figure.

[0060] As shown in Figure 9, the customer service evaluation of employees with high customer service evaluations uses words that match the store information text data (store philosophy) and words that are generally likely or frequently used by employees with high customer service evaluations (words with high customer service evaluations).

[0061] As shown in Figure 10, the customer service evaluation of an employee with a medium customer service evaluation may include fewer words that match the store information text data (store philosophy) and fewer words that are generally likely or frequently used by employees when the customer service evaluation is high (words with high customer service evaluations).

[0062] The evaluation unit 13 can output the generated evaluation result to, for example, the display device 912 or a speaker in the input / output unit 914. In the case of the display device 912, the display device 912 displays the evaluation result on a screen. In the case of the input / output unit 914 being a speaker, the speaker outputs the evaluation result as sound.

[0063] In this way, customer service evaluation system 10 can evaluate customer service not only based on customer service voice but also based on whether the store's management philosophy is met. This allows customer service evaluation system 10 to evaluate the extent to which employees are achieving conceptual store goals such as the customer service philosophy and management philosophy.

[0064] The evaluation unit 13 outputs the generated evaluation results to the customer service evaluation DB 99. The customer service evaluation DB 99 stores the newly generated customer service evaluations. This updates the customer service evaluation history for each employee.

[0065] C: Generation of Improvement Plan The evaluation unit 13 outputs the generated evaluation result to the improvement plan generation unit 14. Note that if an improvement plan is to be generated later for an already evaluated customer service evaluation, the improvement plan generation unit 14 generates an improvement plan based on an operation input for generating an improvement plan from the operation unit 913.

[0066] FIG. 11 is a flowchart showing the processing contents of the evaluation unit 13 and the improvement plan generation unit 14, showing an example of a case where the generation of the customer service evaluation and the generation of the improvement plan are performed in a series of processes.

[0067] The evaluation unit 13 acquires customer service voice (voice text data) for a time period to be evaluated from the recording device 92 (S11). The evaluation unit 13 acquires the store philosophy (store information text data) from the recording device 92 (S12). The evaluation unit 13 generates a customer service evaluation using a customer service evaluation model that receives the customer service voice (voice text data) and the store philosophy (store information text data) as inputs (S13).

[0068] The improvement plan generating unit 14 generates an improvement plan using an improvement plan generation model that receives the customer service evaluation as an input (S14).

[0069] The improvement plan generation model extracts words and phrases corresponding to improvement points from the text data of customer service evaluations. The improvement plan generation model has previously learned words and phrases corresponding to improvement points and improvement plans corresponding to the improvement points. As with the above-mentioned customer service evaluation model, the model is trained using, for example, information set in advance for each store or general information about customer service that exists on an information network.

[0070] The improvement proposal generation model receives text data of customer service evaluations as input, extracts words and phrases of improvement points contained in the text data of customer service evaluations, and outputs improvement proposals corresponding to these improvement points.

[0071] This allows the improvement plan generating unit 14 to output an appropriate improvement plan for the employee being evaluated based on the customer service evaluation.

[0072] 12 is a diagram showing an example of improvements and improvement proposals. As shown in Fig. 12, if the improvement proposal generating unit 14 can extract "Although the response was friendly, more polite language would have been possible," in the customer service evaluation, it outputs "Remember polite language and customer service terminology and respond accordingly" as an improvement proposal.

[0073] In this way, the customer service evaluation system 10 can generate appropriate improvement plans based on the customer service evaluation for the employee being evaluated.

[0074] The improvement plan generating unit 14 outputs the generated improvement plan to the customer service evaluation DB 99. The customer service evaluation DB 99 stores the improvement plan in association with the employee ID.

[0075] D: Feedback of Improvement Suggestions Fig. 13 is a flowchart showing the processing content of the feedback unit, showing an example of feedback of improvement suggestions. Fig. 14 is a diagram showing an example of interactive feedback of improvements.

[0076] The feedback unit 16 accepts an inquiry from an employee regarding the customer service evaluation (improvement) (S21). Specifically, for example, the employee operates the operation unit or microphone of the information communication terminal 93 that the employee owns to make an inquiry regarding the customer service evaluation (improvement). The information communication terminal 93 acquires the customer service evaluation and improvement plan from the customer service evaluation DB 99 based on the inquiry and the employee ID.

[0077] The information communication terminal control unit of the information communication terminal 93 generates a response based on the improvement plan in response to the inquiry. At this time, the information communication terminal control unit generates the response using an interactive large-scale language model.

[0078] More specifically, the information communication terminal control unit constituting the feedback unit 16 converts the inquiry content into text data and interprets the inquiry content (S22). The information communication terminal control unit extracts key points (e.g., important points) of the response from the improvement plan based on the inquiry content (S23). The information communication terminal control unit generates an interactive response based on the key points of the response (S24). An interactive large-scale language model is used here. The information communication terminal 93 notifies the employee of the response in an interactive format (interactive interface) using a display unit and speaker, as shown in FIG. 14.

[0079] In this case, as shown in FIG. 14, the feedback unit 16 responds in an interactive manner to the employee's relatively abstract inquiry by informing the employee of what more specifically he or she should do.

[0080] With this configuration, the customer service evaluation system 10 can dig deeper into areas for improvement rather than providing employees with feedback based on a single quantitative value. This makes it easier for employees who receive feedback to take specific action for improvement. Furthermore, by providing feedback through an interactive interface, the customer service evaluation system 10 can receive inquiries about areas for improvement regardless of location or time, and can propose areas for improvement based on the employee's understanding and interests. This further promotes specific action for improvement.

[0081] [Second Embodiment] A customer service evaluation system according to a second embodiment of the present invention will be described with reference to the drawings. Fig. 15 is a diagram showing in more detail an example of the functional and physical configurations of the customer service evaluation system according to the second embodiment of the present invention. Fig. 16 is a diagram showing an example of the physical configuration of the customer service evaluation system according to the second embodiment of the present invention.

[0082] 15 and 16 , the customer service evaluation system 10A according to the second embodiment differs from the customer service evaluation system 10 according to the first embodiment in that it includes a camera 112 and that the customer service video captured by the camera 112 is also used for customer service evaluation. Other configurations of the customer service evaluation system 10A are the same as those of the customer service evaluation system 10, and therefore a description of similar parts will be omitted.

[0083] In the customer service evaluation system 10A, a store is provided with a microphone 111 and a camera 112. That is, as a basic configuration of the customer service evaluation system 10A, the customer service status collection unit 11 is provided with the microphone 111 and the camera 112.

[0084] The camera 112 captures a customer service video (a video of an employee's actions). The store communication device 119 transmits the customer service video to the text data generation unit 911 via the network 900.

[0085] The text data generation unit 911 converts the customer service video into text data format to generate video text data. Video text data is data in which the content (features) of the video has been converted into text. For example, the video text data is data in which the posture, attitude, and actions of the employee when serving customers have been converted into text. The text data generation unit 911 transmits the video text data to the recording device 92. The recording device 92 records the received video text data by linking it to the recording time, store ID, and employee ID.

[0086] Figure 17 (A) is a flowchart showing the processing contents of the evaluation unit 13, illustrating an example of customer service evaluation in the second embodiment, and Figure 17 (B) is a flowchart showing a more specific example of the processing for calculating the customer service evaluation in Figure 17 (A).

[0087] The evaluation unit 13 acquires customer service voice (voice text data) and customer service video (video text data) for the evaluation time period from the recording device 92 (S11A). The evaluation unit 13 acquires the store philosophy (store information text data) from the recording device 92 (S12).

[0088] The evaluation unit 13 generates a customer service evaluation using a customer service evaluation model that receives as input customer service voice (voice text data), customer service video (video text data), and store philosophy (store information text data) (S13A). More specifically, the evaluation unit 13 inputs the voice text data, video text data, and store information text data into a natural language-based customer service evaluation model (S131A). The natural language-based customer service evaluation model is configured based on, for example, a large-scale language model. The evaluation unit 13 generates a customer service evaluation using the natural language-based customer service evaluation model (S132A).

[0089] In addition, if the customer service audio, customer service video, and store philosophy are not saved as text data, the text data generation unit 911 can convert the customer service audio, customer service video, and store philosophy into text data before performing the customer service evaluation, and then the processing by the evaluation unit 13 described above can be performed.

[0090] The conversion of a customer service video into text is performed, for example, as follows: From an image showing one frame in the video, the text data generation unit 911 outputs text describing the characteristics of the image of one frame, such as "A place with multiple tables and chairs lined up, with three people seated at the table in the foreground. Food is laid out on the table, and one person is standing next to the table, smiling and talking to the three people." By creating text describing such images, for example, every five seconds, the text data generation unit 911 can convert the situation stored in the video into text data.

[0091] With this configuration and processing, the customer service evaluation system 10A can perform customer service evaluations that also reflect the video of the customer service attitude of the employee.

[0092] [Third Embodiment] A customer service evaluation system according to a third embodiment of the present invention will be described with reference to the drawings. Fig. 18 is a functional block diagram showing the basic functional configuration of the customer service evaluation system according to the third embodiment of the present invention. Fig. 19 is a diagram showing in more detail an example of the functional and physical configuration of the customer service evaluation system according to the third embodiment of the present invention.

[0093] 18 and 19 , the customer service evaluation system 10S according to the third embodiment differs from the customer service evaluation system 10A according to the second embodiment in that it includes an environmental condition sensor 17 and an evaluation unit 13S. The other components of the customer service evaluation system 10S are the same as those of the customer service evaluation system 10A, and a description of similar parts will be omitted.

[0094] The customer service evaluation system 10S includes an environmental condition sensor 17. The environmental condition sensor 17 is installed in, for example, a store and includes a sensor 113 that measures the environmental condition (temperature, humidity, etc.) of the customer service space. That is, the sensor 113 can detect the environmental condition around the customer that the employee is serving as part of their work.

[0095] The sensor 113 is configured to include at least one of a temperature sensor, a humidity sensor, an illuminance sensor, etc. The sensor 113 measures the environmental conditions (temperature, humidity, illuminance, etc.) of the customer service space according to the sensor configuration and outputs sensor data. The store communication device 119 transmits the sensor data to the text data generation unit 911 via the network 900. The text data generation unit 911 converts the sensor data into sensor text data, for example, by listing the temperature, humidity, and illuminance in chronological order, separated by semicolons or the like. The text data generation unit 911 transmits the sensor text data to the recording device 92. The recording device 92 records the received sensor text data by linking it to the sound collection time and the store ID.

[0096] Figure 20 (A) is a flowchart showing the processing contents of the evaluation unit 13S, illustrating an example of customer service evaluation in the third embodiment, and Figure 20 (B) is a flowchart showing a more specific example of the processing for calculating the customer service evaluation in Figure 20 (A).

[0097] The evaluation unit 13S acquires customer service voice (voice text data) and customer service video (video text data) for the time period to be evaluated from the recording device 92 (S111). The evaluation unit 13S acquires the surrounding environmental conditions of the customer service (sensor text data) from the recording device 92 (S112). The evaluation unit 13S acquires the store philosophy (store information text data) from the recording device 92 (S12).

[0098] The evaluation unit 13S generates a customer service evaluation using a customer service evaluation model that inputs customer service voice (voice text data), customer service video (video text data), the surrounding environmental conditions of the customer service (sensor text data), and the store philosophy (store information text data) (S13S). More specifically, the evaluation unit 13S inputs the voice text data, video text data, sensor text data, and store information text data into a natural language-based customer service evaluation model (S131S). The natural language-based customer service evaluation model is configured based on, for example, a large-scale language model. The evaluation unit 13S generates a customer service evaluation using the natural language-based customer service evaluation model (S132S).

[0099] In addition, if the customer service audio, customer service video, surrounding environmental conditions of the customer service, and customer service philosophy are not saved as text data, the text data generation unit 911 can convert the customer service audio, customer service video, surrounding environmental conditions of the customer service, and customer service philosophy into text data before performing the customer service evaluation, and then the processing of the above-mentioned evaluation unit 13S can be performed.

[0100] With this configuration and processing, the customer service evaluation system 10S can perform customer service evaluation by also reflecting the customer service environment, i.e., the temperature, humidity, etc. of the customer service space of the store.

[0101] [Fourth embodiment] A customer service evaluation system according to a fourth embodiment of the present invention will be described with reference to the drawings. Fig. 21 is a flowchart showing the processing of the evaluation unit, illustrating an example of customer service evaluation according to the fourth embodiment.

[0102] The customer service evaluation system according to the fourth embodiment performs statistical customer service evaluation using the results of customer service evaluations conducted at multiple stores. For example, this system is useful when a company operates multiple stores. Although not shown, the processing shown in FIG. 21 may be performed by, for example, a processing device 91 equipped with the evaluation unit 13, or by a separately installed processing device connected to the network 900. For convenience of explanation, this processing device will be referred to below as a statistical evaluation processing device.

[0103] The statistical evaluation processing unit acquires the customer service evaluation of each employee from the customer service evaluation DB 99 (S31).

[0104] The statistical evaluation processing unit generates statistical data of customer service evaluations for each store (S32). The statistical data includes the average level of customer service evaluations, items for which customer service evaluations are high for all employees or for all store employees, and items for which customer service evaluations are low for all employees or for all store employees.

[0105] The statistical evaluation processing unit generates a customer service evaluation and improvement plan for each store (S33). This allows a company that operates multiple stores to evaluate and improve each store by comparing it with other stores.

[0106] Fifth Embodiment A customer service evaluation system according to a fifth embodiment of the present invention will be described with reference to the drawings. The customer service evaluation system according to the fifth embodiment differs from the customer service evaluation system according to the first embodiment in that it reflects the intentions of higher management (such as the owner, store managers, and district leaders).

[0107] Fig. 22 is a functional block diagram showing the basic functional configuration of a customer service evaluation system according to a fifth embodiment of the present invention. Fig. 23 is a diagram showing in more detail an example of the functional and physical configuration of the customer service evaluation system according to the fifth embodiment of the present invention. Fig. 24 is a diagram showing an outline of the flow of relearning the customer service evaluation model in the customer service evaluation system according to the fifth embodiment of the present invention.

[0108] As shown in Fig. 22, the customer service evaluation system 10B according to the fifth embodiment differs from the customer service evaluation system 10 according to the first embodiment in that it includes an evaluation unit 13B and a superior manager evaluation information acquisition unit 18B. The other components of the customer service evaluation system 10B are the same as those of the customer service evaluation system 10, and a description of similar parts will be omitted. Also, as shown in Fig. 23, the customer service evaluation system 10B differs from the customer service evaluation system 10 according to the first embodiment in that it includes a superior manager holding device 98B. The other components of the customer service evaluation system 10B are the same as those of the customer service evaluation system 10, and a description of similar parts will be omitted.

[0109] The managerial superior evaluation information acquisition unit 18B acquires managerial superior evaluation information for the evaluation results evaluated using the customer service evaluation model (FIG. 24: S41). The managerial superior evaluation information includes the results of a reevaluation by the managerial superior of the evaluation results (pre-evaluation results) evaluated using the customer service evaluation model. The managerial superior's evaluation (re-evaluation) is determined based on whether the pre-evaluation results are in line with the managerial superior's intentions, etc.

[0110] The manager device 98B includes a manager evaluation information acquisition unit 18B. The manager device 98B is connected to a network 900. The manager device 98B is configured by, for example, a tablet terminal.

[0111] The manager re-evaluates the pre-evaluation results using the manager device 98B. Figures 25A and 25B are diagrams showing an example of a manager's evaluation (re-evaluation) method.

[0112] When a manager inputs the acquisition of a customer service evaluation to perform an evaluation (re-evaluation), the manager-held device 98B acquires the customer service evaluation of the evaluation target from the customer service evaluation DB 99 via the network 900. The manager-held device 98B displays the acquired customer service evaluation on the screen 50 as shown in Figures 25(A) and 25(B).

[0113] When the manager touches the screen 50 or uses the cursor 51 to select an evaluation item or an evaluation sentence, an evaluation window 52 is displayed.

[0114] 25A shows a case where evaluation is performed on an item-by-item basis. As shown in FIG. 25A, when an item ("customer service evaluation") is selected, an evaluation window 52 is displayed in a position overlapping or adjacent to the item ("customer service evaluation").

[0115] 25(B) shows a case where evaluation is performed on a sentence-by-sentence basis. As shown in FIG. 25(B), when a sentence to be evaluated is selected, an evaluation window 52 is displayed at a position overlapping or adjacent to this sentence.

[0116] The evaluation window 52 is composed of, for example, button icons of "GOOD" and "NOT GOOD." When a manager selects the "GOOD" icon, the manager evaluation information acquisition unit 18B detects this and determines that the manager "agrees" with the evaluation result. On the other hand, when a manager selects the "NOT GOOD" icon, the manager evaluation information acquisition unit 18B detects this and determines that the manager "does not agree" with the evaluation result.

[0117] The management superior's evaluation information acquisition unit 18B determines the management superior's evaluation information based on this judgment result. For example, the management superior's evaluation information acquisition unit 18B includes in the management superior's evaluation information the linking of the pre-evaluation result of a "GOOD" judgment with the "GOOD" judgment. On the other hand, the management superior's evaluation information acquisition unit 18B includes in the management superior's evaluation information the linking of the pre-evaluation result of a "NOT GOOD" judgment with the "NOT GOOD" judgment.

[0118] In other words, the managerial superior evaluation information acquisition unit 18B determines the linking information between the preliminary evaluation result and the evaluation result of the managerial superior as the managerial superior evaluation information based on the evaluation result of the managerial superior. Note that for items or sentences that the managerial superior did not evaluate, the managerial superior evaluation information may be determined as items with no change in weighting for the parameters used in the customer service evaluation, for example.

[0119] The manager superior's evaluation information acquisition unit 18B converts the manager superior's evaluation information into text data and outputs it to the evaluation unit 13B. Note that this conversion into text data can also be performed by the evaluation unit 13B.

[0120] The evaluation unit 13B updates (re-learns) the customer service evaluation model using the managerial superior evaluation information (FIG. 24: S42).

[0121] FIG. 26 is a flowchart showing an example of a method for relearning the customer service evaluation model.

[0122] The evaluation unit 13B acquires text data related to customer service (customer service voice (voice text data) and customer service philosophy (store information text data)), customer service evaluation (preliminary evaluation result), and managerial superior evaluation information (S421).

[0123] The evaluation unit 13B re-learns the customer service evaluation model based on the text data related to customer service, the customer service evaluation (preliminary evaluation result), and the managerial superior evaluation information (S422).

[0124] The evaluation unit 13B performs a customer service evaluation using the re-learned customer service evaluation model. This customer service evaluation is judged by the manager of the customer service evaluation model, etc., to determine whether it properly reflects the evaluation information of the superior manager. The result of this judgment is returned to the evaluation unit 13B. Note that this judgment can also be made by calculation processing, etc., using the difference between the customer service evaluation after re-learning and the pre-evaluation result.

[0125] The evaluation unit 13B checks the re-learning result (S423) and determines that the learning is OK if the managerial superior evaluation information is properly reflected. The evaluation unit 13B determines that the learning is NG if the managerial superior evaluation information is not properly reflected.

[0126] If the learning is successful (S424: YES), the evaluation unit 13B saves the customer service evaluation model after the re-learning (S425). On the other hand, if the learning is unsuccessful (S424: NO), the evaluation unit 13B restores the customer service evaluation model to the one before the start of the re-learning (S426).

[0127] With such a configuration and processing, the customer service evaluation system 10B can perform customer service evaluation that reflects the intentions of a superior manager (for example, a manager or a district leader).

[0128] The following method, for example, can be used to reflect the managerial superior evaluation information in the customer service evaluation model.

[0129] 27(A) and 27(B) are diagrams showing a learning method and a customer service evaluation method that reflect the managerial superior evaluation information.

[0130] FIG. 27A illustrates a method for directly retraining a customer service evaluation model. In the method illustrated in FIG. 27A, a pre-service evaluation model is prepared. The pre-service evaluation model is a customer service evaluation model before reflecting the manager's evaluation information. As an example, the pre-service evaluation model is a trained model that inputs the management philosophy and the like. By feeding back the manager's evaluation of the actual evaluation results, the accuracy of the customer service evaluation in line with the management philosophy can be improved. As another example, the pre-service evaluation model may be a general customer service evaluation model that includes at least one of text information on general knowledge about customer service, common conversation patterns in restaurant service situations, text information on philosophies common to the customer service industry, and publicly known information on good customer service. In this case, costs can be reduced by using a common model, and the customer service evaluation model can be updated to be in line with the management philosophy by feeding back the manager's evaluation. This allows for improved evaluation accuracy while reducing the costs associated with preparing and operating the customer service evaluation model.

[0131] The customer service evaluation system 10B adds the manager's evaluation information to the pre-customer service evaluation model and re-learns the customer service evaluation model. The customer service evaluation system 10B inputs text data related to customer service into the re-learned customer service evaluation model and performs customer service evaluation.

[0132] By directly relearning the customer service evaluation model, the customer service evaluation system 10B can gradually improve the evaluation accuracy of the customer service evaluation model, and can perform customer service evaluations with even higher accuracy.

[0133] 27(B) shows an example of utilizing RAG. RAG stands for Retrieval-Augmented Generation, and is a method that uses, for example, a generative AI trained on publicly available information to individually prepare detailed information such as unique store rules as a knowledge source, and then performs customer service evaluations that take this knowledge source into account.

[0134] RAG outputs the final result by referencing a dedicated database of input text data related to customer service. In other words, based on the input data, RAG searches and extracts information on managerial superiors' evaluations that can be processed by the current customer service evaluation model and information necessary for customer service evaluation in store information, and outputs the final evaluation.

[0135] The RAG provides the extracted information to the current customer service evaluation model, and obtains and outputs a customer service evaluation based on the customer service evaluation model.

[0136] For example, if the current customer service evaluation model can only determine the customer service evaluation for serving wine, when a customer asks, "Do you have any wine recommendations?", RAG will analyze this and obtain, for example, the store's recommended wines and food and wine pairings from the store information.

[0137] The customer service evaluation model evaluates customer service based on whether or not the text data about customer service includes answers about the store's recommended wines and food and wine pairings. The RAG acquires and outputs this customer service evaluation.

[0138] By using RAG in this way, the customer service evaluation system 10B can perform more appropriate customer service evaluation even if the customer service evaluation model has not been sufficiently trained.

[0139] In the above description, the operational input that forms the basis of the manager evaluation information is a two-choice option. However, the operational input is not limited to a two-choice option. For example, the operational input may be a numerical value on a predetermined scale (e.g., 10-scale), or the manager may directly input text data.

[0140] Furthermore, instead of operation input, voice input can also be used. In this case, the manager evaluation information acquisition unit 17B has a function of converting voice into text data.

[0141] Sixth Embodiment A customer service evaluation system according to a sixth embodiment of the present invention will be described with reference to the drawings. The customer service evaluation system according to the sixth embodiment differs from the customer service evaluation system according to the first embodiment in that it reflects the thinking of higher-level managers (such as the owner, store managers, and district leaders). Management philosophies are often expressed in relatively abstract terms, and it may be difficult to obtain a specific customer service evaluation based on the management philosophies alone. However, by inputting the owner's more detailed and specific thoughts, it becomes possible to obtain a highly accurate customer service evaluation.

[0142] From another perspective, the customer service evaluation system according to the sixth embodiment replaces the intentions of the superior manager, which are directly input in the customer service evaluation system according to the fifth embodiment, with thoughts that do not rely on direct input.

[0143] Fig. 28 is a functional block diagram showing the basic functional configuration of a customer service evaluation system according to a sixth embodiment of the present invention. Fig. 29 is a diagram showing in more detail an example of the functional and physical configuration of the customer service evaluation system according to the sixth embodiment of the present invention.

[0144] 28 , the customer service evaluation system 10C according to the sixth embodiment differs from the customer service evaluation system 10 according to the first embodiment in that it includes an evaluation unit 13C and a managerial superior thought information acquisition unit 18C. The other components of the customer service evaluation system 10C are the same as those of the customer service evaluation system 10, and a description of the same parts will be omitted.

[0145] 29 , the customer service evaluation system 10C differs from the customer service evaluation system 10 according to the first embodiment in that it includes a manager information input device 98C. The other components of the customer service evaluation system 10C are the same as those of the customer service evaluation system 10, and a description of the same parts will be omitted.

[0146] The management superior information input device 98C includes a management superior thought information acquisition unit 18C. The management superior information input device 98C is connected to a network 900. The management superior information input device 98C includes at least one of an audio acquisition function, a video acquisition function, and an external information acquisition function.

[0147] When the voice acquisition function is used, the supervisor information input device 98C acquires the supervisor's conversation voice and lecture voice.

[0148] When the image acquisition function is used, the supervisor information input device 98C acquires the supervisor's conversation video and lecture video.

[0149] When using the external information acquisition function, the manager information input device 98C acquires the manager's internet browsing history, viewed articles, information on books read, web articles, posts on SNS and blogs, etc. Information on books read can be acquired, for example, by the manager of the customer service evaluation model, etc. inputting the text of the book title.

[0150] Based on the acquired information, the superior manager's thought information acquisition unit 18C acquires information that can be used for relearning the customer service evaluation model. For example, the superior manager's thought information acquisition unit 18C converts the acquired information into text data and performs a text search using keywords related to "customer service," "management," etc. The superior manager's thought information acquisition unit 18C acquires a predetermined range (such as phrases) containing the words extracted by the text search as superior manager's thought information to be used for training the customer service evaluation model.

[0151] The managerial superior's thought information acquisition unit 18C outputs the acquired managerial superior's thought information to the evaluation unit 13C. The evaluation unit 13C uses the managerial superior's thought information to update (re-learn) the customer service evaluation model in the same way as when using the above-mentioned managerial superior's evaluation information.

[0152] The manager's thinking information directly includes specific content that expresses the management philosophy, such as the points that the manager places importance on in management and the value that the manager wants to provide to customers, and also includes thinking content that is indirectly connected to the management philosophy from books that the manager has read, web pages that the manager has viewed, etc. This makes it possible to apply the thinking behind the management philosophy to the customer service evaluation model with greater accuracy than if only the abstract management philosophy were used.

[0153] If there are multiple types of managerial superior's thinking information, the importance of the information can be set and adjusted for each type. For example, a high importance can be set for information that directly expresses the managerial superior's thinking (especially thoughts related to work) (audio of conversation, video of lecture, web article summarizing the managerial superior's thoughts and comments, SNS, blog). This information can be set to a high importance because it is information that the managerial superior himself output, consciously or unconsciously.

[0154] On the other hand, a low importance level is set for information that is not clearly directly related to the manager's thoughts (especially thoughts related to work) (information on books read, articles viewed, internet browsing history, etc.) This information is sent by a third party other than the manager, and does not necessarily match the manager's thoughts, so a low importance level can be set.

[0155] The importance level setting is not limited to these two types, but can be further divided into more detailed categories, such as by the type of information source, words contained in the information source, and the time of presentation of the information.

[0156] In this way, the customer service evaluation system 10C can appropriately update the customer service evaluation model to match the thoughts (intentions) of the superior manager, without having the superior manager directly evaluate the customer service evaluation.

[0157] Seventh Embodiment A customer service evaluation system according to a seventh embodiment of the present invention will be described with reference to the drawings. The seventh embodiment differs from the customer service evaluation system according to the first embodiment in that it reflects the intentions of employees who receive feedback on their customer service evaluations.

[0158] FIG. 30 is a functional block diagram showing the basic functional configuration of a customer service evaluation system according to the seventh embodiment of the present invention.

[0159] The customer service evaluation system 10D according to the seventh embodiment differs from the customer service evaluation system 10 according to the first embodiment in that it includes an evaluation unit 13D and an employee intention information acquisition unit 19. The other components of the customer service evaluation system 10D are the same as those of the customer service evaluation system 10, and a description of the same parts will be omitted.

[0160] The employee intention information acquisition unit 19 acquires the intentions of employees who receive feedback on their customer service evaluations as employee intention information. The employee intention information is composed of, for example, the employee's customer service goals and things they would particularly like to work hard on. The employee intentions can be acquired by the employee's operational input to an information communication terminal held by the employee (e.g., equivalent to the information communication terminal 93A in FIG. 2), or can be acquired based on target values ​​previously stored in the employee information.

[0161] The employee intention information acquisition unit 19 converts the acquired employee intention information into text data and outputs it to the evaluation unit 13D.

[0162] The evaluation unit 13D performs customer service evaluation using the employee intention information.

[0163] As a result, the customer service evaluation system 10D can perform customer service evaluations that are suited to the wishes of each employee, rather than providing a uniform customer service evaluation for all employees, and can create improvement proposals and provide feedback that are suited to the wishes of each employee.

[0164] In this case, it is preferable that the evaluation unit 13D employs the above-mentioned RAG.

[0165] Since the intentions of each employee are different, if the customer service evaluation model is directly updated (relearned) by referring to the intentions of all employees, there is a possibility that the customer service evaluation model will not converge as desired by the manager.

[0166] On the other hand, by using RAG, it is possible to perform customer service evaluations that are appropriate to the intentions of each employee without undesirably updating (relearning) the customer service evaluation model.

[0167] Eighth Embodiment A customer service evaluation system according to an eighth embodiment of the present invention will be described with reference to the drawings. The customer service evaluation system according to the eighth embodiment differs from the customer service evaluation system according to the first embodiment in that intermediate concepts (categories) are set between customer service observation data (speech and text data, etc.) and philosophies (management philosophy and customer service philosophy).

[0168] Fig. 31 is a functional block diagram showing the basic functional configuration of a customer service evaluation system according to an eighth embodiment of the present invention. Fig. 32 is a diagram showing in more detail an example of the functional and physical configuration of the customer service evaluation system according to the eighth embodiment of the present invention.

[0169] As shown in Figures 31 and 32, the customer service evaluation system 10E according to the eighth embodiment differs from the customer service evaluation system 10 according to the first embodiment in that it includes a classification unit 130, a storage device 92E, and an evaluation unit 13E. The customer service evaluation system 10 differs in that it includes a data cloud 100E, a processing device 91E of the data cloud 100E includes the classification unit 130, and records category text data in a recording device 92E. As will be described later, the evaluation unit 13E differs in that it inputs categories in addition to voice text data and store information text data to generate customer service evaluation results. The other components of the customer service evaluation system 10E are the same as those of the customer service evaluation system 10, and a description of similar parts will be omitted.

[0170] The classification unit 130 classifies the voice text data (customer service observation data) into categories. A category is an intermediate concept between the customer service observation data and the philosophy.

[0171] 33 is a diagram showing an example of the relationship between customer service observation data (speech and text data), categories, and ideals. For example, as shown in FIG. 33, categories are set such as "ability to make suggestions," "ability to understand cooking," "ability to grasp the situation," "sincerity," and "consideration." In this way, categories are set to represent higher-level concepts of utterances and actions taken during customer service.

[0172] A list of categories is created using voice-to-text data from employees who are highly rated as embodying the store manual and philosophy.

[0173] Specifically, categories are set or generated as follows: (A) A senior administrator sets categories by operating input, or (B) Model speech and text data is input into a large-scale language model to generate categories.

[0174] Furthermore, a senior administrator may select and reject words and adjust phrases based on a category list generated by inputting the speech text data of (B) into a large-scale language model.

[0175] Each category is associated with at least one philosophy. The association between categories and philosophies is set in advance. The association between categories and philosophies is established in advance by an operational input from a supervisor. The association between categories and philosophies in advance may be established using a large-scale language model, or the supervisor may adjust the association by viewing the association established by the large-scale language model.

[0176] 32, the categories are recorded in the form of text data in the recording device 92E. Association data indicating which corporate philosophy each category text data is associated with is also recorded together with the category text data.

[0177] The classification unit 130 uses a large-scale language model relating to customer service categories and specific utterances based on the association information recorded in the recording device 92E to classify the speech text data of the employee to be evaluated into categories, each of which is segmented to a predetermined length or to group together content from the same field. Figure 34 is a diagram showing an example of classifying multiple segmented speech text data into categories.

[0178] The classification unit 130 outputs the voice text data and the category (category text data) to the evaluation unit 13E.

[0179] The evaluation unit 13E generates a customer service evaluation result using a customer service evaluation model that receives as input voice text data, category text data, and store information text data including customer service philosophy.

[0180] The improvement plan generator 14 generates an improvement plan based on the customer service evaluation. FIG. 35 is a diagram showing an example of a notification of the customer service evaluation result and the improvement plan. As shown in FIG. 35, the classified categories, customer service evaluation, and improvement plan for the customer service voice text data are notified so that the relationship between them can be understood. Furthermore, in the case of FIG. 35, general advice is also notified. The general advice is generated, for example, based on the improvement plan for each category.

[0181] In this way, the customer service evaluation system 10E can perform customer service evaluation based on voice text data, just like the customer service evaluation system 10. In this case, the customer service evaluation system 10E can accurately associate concepts with categories by using categories that supersede the content of the voice text data, thereby more accurately associating the voice text data with concepts. Therefore, the customer service evaluation system 10E can perform more accurate customer service evaluation.

[0182] Furthermore, customer service evaluation system 10E can output the required evaluation without the need for a manager or the supplier preparing the system to directly associate each voice text data with each philosophy or to confirm the validity of the association, thereby reducing the burden of building a customer service evaluation system.

[0183] (Customer service evaluation method) Figure 36 (A) is a flowchart showing an example of processing performed by the classification unit in the customer service evaluation method related to the eighth embodiment, and Figure 36 (B) is a flowchart showing an example of processing performed by the evaluation unit in the customer service evaluation method related to the eighth embodiment.

[0184] 36A, the classification unit 130 acquires text data (e.g., speech text data) related to customer service (S81). The classification unit 130 classifies the text data related to customer service into categories using a large-scale language model based on general customer service concepts (S82). That is, the text data related to customer service and category data are stored in association with each other.

[0185] 36(B), the evaluation unit 13E acquires text data related to customer service, category text data, and store information text data (S83). The evaluation unit 13E generates a customer service evaluation using a customer service evaluation model that receives as input the text data related to customer service, category text data, and store information text data (S84).

[0186] (Pre-association Method 1) Fig. 37 is a flowchart showing a first example of a pre-association process in the customer service evaluation method according to the eighth embodiment. Here, a method (A) in which a manager sets categories by inputting operations will be described.

[0187] 37, the customer service evaluation system 10E acquires category text data representing a plurality of set categories (S891). The customer service evaluation system 10E acquires store information text data, which is a management philosophy including a customer service philosophy and a corporate philosophy (S892).

[0188] The customer service evaluation system 10E uses a large-scale language model to set the relevance between the category text data and the store information text data (philosophy) (S893).

[0189] The customer service evaluation system 10E determines the associations between multiple categories and multiple philosophies through operational input by the manager or the like (S894). If there is a problem with the associations between categories and philosophies based on the large-scale language model, the associations modified by operational input by the manager or the like are determined as the final associations between categories and philosophies. In this example, the associations are determined through operational input by the manager or the like, but the associations may also be determined automatically after the associations between categories and philosophies are determined using a large-scale language model.

[0190] (Pre-Association Method 2) FIG. 38 is a flowchart showing a second example of the pre-association process in the customer service evaluation method according to the eighth embodiment.

[0191] As shown in FIG. 38 , the customer service evaluation system 10E acquires a group of exemplary speech and text data (S895). The exemplary speech and text data may be, for example, text data listed in a store manual, or actual customer service speech or exemplary role-playing by an employee with a high customer service evaluation, or by the store manager or other superior. The customer service evaluation system 10E uses a large-scale language model based on general customer service concepts to generate categories that are superordinate concepts of the multiple speech and text data that make up the speech and text data group (S896). At this time, the customer service evaluation system 10E may also allow the superior to select from the multiple categories generated by the large-scale language model.

[0192] The customer service evaluation system 10E sets associations between multiple categories and multiple philosophies through operational input by a superior manager (S897). Note that, similar to the pre-association method 1, it is also possible to set associations between multiple categories and multiple philosophies using a large-scale language model.

[0193] Ninth Embodiment A customer service evaluation system according to a ninth embodiment of the present invention will be described with reference to the drawings.

[0194] In the above-described embodiments, a system and a method for evaluating customer service have been specifically described. However, the evaluation system is not limited to customer service evaluation, and can also be applied to evaluation of the results of work performed by a company or the like.

[0195] Fig. 39 is a functional block diagram showing the basic functional configuration of a job evaluation system according to a ninth embodiment of the present invention. Fig. 40 is a diagram showing in more detail an example of the functional and physical configuration of a job evaluation system according to the ninth embodiment of the present invention. Fig. 41 is a flowchart showing an example of a job result evaluation method according to the ninth embodiment.

[0196] As shown in FIG. 39, the business evaluation system 10F according to the ninth embodiment includes a business status collection unit 11F, a company information management unit 12F, an evaluation unit 13F, an improvement plan generation unit 14, a notification unit 15, and a feedback unit 16.

[0197] The business status collection unit 11F is a functional unit that replaces the customer service information collection unit 11 according to the first embodiment, and performs the same (similar) processing as the customer service information collection unit 11. The business status collection unit 11F is configured, for example, as shown in FIG. 40 , by a PC (personal computer) 111F, a microphone 111, etc., located at the business premises. The business results include the status of the work being performed by employees, and are, for example, documents prepared by employees such as individual employee goal setting sheets, annual departmental policy documents, materials prepared for regular meetings, and training materials, as well as employee voices such as internal report speeches and business negotiation voices.

[0198] The task status collection unit 11F transmits the acquired task results to the text data generation unit 911 of the data cloud 100F. The text data generation unit 911 converts the task results into text data to generate task result data, and records the data in the recording unit 92F.

[0199] The company information management unit 12F is a functional unit that replaces the store information management unit 12 according to the first embodiment, and performs the same (similar) processing as the store information management unit 12. The company information management unit 12F acquires company information text data recorded in the recording device 92F. The company information text data is a company philosophy in text data format. The company information management unit 12F outputs the company information text data to the evaluation unit 13F.

[0200] The evaluation unit 13F performs the same (similar) processing as the evaluation unit 13 according to the first embodiment. The evaluation unit 13F acquires business results (business result data) (FIG. 41: S11F). The evaluation unit 13F acquires the company philosophy (company information text data) (FIG. 41: S12F).

[0201] The evaluation unit 13F generates a business evaluation using a large-scale language model that inputs the business results (business result data) and the company philosophy (corporate information text data) (FIG. 41: S13F).

[0202] With this configuration and processing, the job evaluation system 10F can evaluate whether the results of the job performed by an employee are in line with the company's philosophy, not limited to customer service.

[0203] The job evaluation system 10F can be applied to, for example, the job shown in Fig. 42. Fig. 42 is a table showing examples of evaluation targets and application modes that can be implemented by the job evaluation system.

[0204] The job evaluation system can be applied to both "general companies" and "civil servants" as major categories of evaluation targets. It can also be applied to "customer service industry," "healthcare and welfare jobs," "education and training jobs," "personnel and recruitment jobs," "leadership and management jobs," "sales jobs," "creative jobs," "clerical jobs," and "field jobs" as medium categories of evaluation targets. Specific examples of the medium categories of evaluation targets are shown in FIG. 42, for example.

[0205] The mode of application of the evaluation of each evaluation target is set appropriately as shown in FIG. 42 based on the business philosophy, business attributes, etc. of each evaluation target.

[0206] In this way, the job evaluation system 10F can be applied to the evaluation of almost all employees belonging to a company or the like.

[0207] The configurations of the above-described embodiments can be combined as appropriate, and effects according to each combination can be achieved.

[0208] DESCRIPTION OF SYMBOLS 10, 10A, 10B, 10C, 10D, 10E, 10S: Customer service evaluation system 10F: Work evaluation system 11: Customer service status collection unit 11F: Work status collection unit 12: Store information management unit 12F: Company information management unit 13, 13E, 13F, 13S: Evaluation unit 14, 14E: Improvement plan generation unit 15: Notification unit 16: Feedback unit 17: Environmental condition sensor 18B: Manager superior evaluation information acquisition unit 18C: Manager superior thought information acquisition unit 19: Employee intention information acquisition unit 91: Processing unit 92, 91E, 92F: Recording device 93A, 93B: Information communication terminal 99: Customer service evaluation DB 98B: Manager superior storage device 98C: Manager superior information input device 100, 100E, 100F: Data cloud 111: Microphone 111F: PC 112: Camera 113: Sensor 119: Store communication device 130: Classification unit 900: Network 911: Text data generation unit 912: Display device 913: Operation unit 914: Input / output unit

Claims

1. A work evaluation system comprising: a work status collection unit that collects work results performed by employees of a company; a company information management unit that stores company information including the company's corporate philosophy; and an evaluation unit that uses the work results and the company information to evaluate the work of the employees.

2. The work evaluation system according to claim 1, wherein the work results include at least one of the employee's voice and documents prepared by the employee.

3. A business evaluation system as described in claim 1 or claim 2, further comprising a text data generation unit that converts the business results into text data format and generates business result text data, the company information management unit stores the company information as company information text data in text data format, and the evaluation unit performs the business evaluation using the business result text data and the company information text data.

4. The job evaluation system according to claim 3, wherein the evaluation unit performs the job evaluation using a large-scale language model.

5. A work evaluation system as described in claim 3 or claim 4, wherein the work status collection unit further collects video of the employee's actions, the text data generation unit converts the video of the actions into text data format to generate video text data, and the evaluation unit further uses the video text data to perform the work evaluation.

6. A work evaluation system as described in any one of claims 3 to 5, wherein the work status collection unit includes an environmental condition sensor that detects the ambient environmental conditions of customers that the employee serves in the course of his / her work and generates sensor data; the text data generation unit converts the sensor data into text data format to generate sensor text data; and the evaluation unit further uses the sensor text data to perform the work evaluation.

7. A work evaluation system according to any one of claims 3 to 6, further comprising a classification unit that classifies the work result text data into categories, and the evaluation unit performs the work evaluation by adding the categories to an input.

8. The work evaluation system according to claim 7, wherein the categories are set by operation input by a senior manager of the company and are recorded in advance.

9. A job evaluation system according to any one of claims 1 to 8, further comprising a notification unit that notifies the job evaluation.

10. A business evaluation system as described in claim 9, further comprising a managerial superior evaluation information acquisition unit that acquires the results of a reevaluation by the company's senior manager of the notified business evaluation, and the managerial superior evaluation information acquisition unit feeds back the results of the reevaluation by the senior manager to the evaluation unit.

11. The business evaluation system according to claim 10, wherein the notification unit sets and notifies a plurality of notification items, and the managerial superior evaluation information acquisition unit acquires the results of the reevaluation for each of the notification items.

12. A business evaluation system as described in any one of claims 1 to 11, comprising a managerial superior thought information acquisition unit that acquires thought information of the company's senior managers, and the managerial superior thought information acquisition unit provides the senior managers' thought information to the evaluation unit.

13. A work evaluation system as described in any one of claims 1 to 12, further comprising an employee intention information acquisition unit that acquires the employee's intention regarding the work results, and the employee intention information acquisition unit provides the acquired employee intention to the evaluation unit.

14. A work evaluation system according to any one of claims 1 to 13, further comprising an improvement plan generation unit that uses the work evaluation to generate improvement plans for the employee's work.

15. A work evaluation system as described in claim 14, further comprising a feedback unit that feeds back the improvement proposal to the employee, wherein the feedback unit receives an inquiry from the employee regarding the work evaluation, generates a response based on the improvement proposal in response to the inquiry, and notifies the employee of the response.

16. A business evaluation system as described in any one of claims 1 to 15, wherein the business status collection unit is provided for each of a plurality of business locations, the company information management unit stores the company information for each of the plurality of business locations, and the evaluation unit performs the business evaluation for each of the plurality of business locations and generates statistical data of the business evaluation for each of the plurality of business locations.

17. A job evaluation system according to any one of claims 1 to 16, wherein the job is customer service.

Images