Information processing device, information processing method, and program

Abstract

We provide technology that promotes changes in internal awareness and behavior to prevent employee turnover by supporting and integrating these actions into habits. [Solution] The present invention is an information processing device for preventing employee turnover, comprising a storage device 7 for storing a training dataset, and a training device 11 for receiving the training dataset, inputting the training dataset into a training target learning model, obtaining processing results from the training target learning model, training the training target learning model by comparing the processing results with the correct labels included in the training dataset, and storing the trained learning model in the storage device, wherein the training dataset comprises at least one of the following: attendance data and correct label pairs, organizational culture diagnosis data and correct label pairs, and behavioral analysis data and correct label pairs.

Description

Technical Field

[0001] The present invention relates to a technique for promoting a change in awareness and behavior within a company in order to prevent employees from leaving the company.

Background Art

[0002] Conventionally, various techniques for suppressing employee turnover have been proposed.

[0003] For example, Patent Document 1 discloses a human resource risk management system that reads stress check data and work attendance data of a plurality of employees, evaluates human resource risks according to the stress check data and work attendance data, and presents them to managers.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, Patent Document 1 does not disclose any mechanism for accompanying actions and continuously providing feedback for habituation as an approach that does not rely solely on the awareness of managers.

[0006] The present invention has been made in view of such problems, and an object thereof is to provide a technique for promoting a change in awareness and behavior within a company to prevent employees from leaving the company by accompanying actions and habituating them.

Means for Solving the Problems

[0007] ​To solve the above problems, an information processing device according to a first aspect of the present invention is an information processing device that supports the prevention of employee turnover, comprising: a storage device that stores at least a training dataset and proposal data corresponding to risk classification; a training device that receives the training dataset, inputs the training dataset into a training target learning model, obtains processing results from the training target learning model, trains the training target learning model by comparing the correct labels included in the training dataset with the processing results, and stores the trained learning model in the storage device; an inference device that obtains inference target data, performs inference using the trained learning model, and outputs a situation classification; and based on the situation classification, retrieves proposal data corresponding to the situation classification from the storage device. The system includes a device for identifying and outputting proposed data, the training dataset comprising at least one of the following: pairs of attendance data and correct labels, pairs of organizational culture diagnostic data and correct labels, and pairs of behavioral analysis data and correct labels, the trained learning model is for inferring the risk classification related to employee turnover, the attendance data includes the number of desired shift days, the number of confirmed shift days, the first day of work, the number of days worked, employees who have not submitted shifts for a certain period of time or longer and their last day of work, the organizational culture diagnostic data is compiled and analyzed from questionnaire response data, and the proposed data includes comment data for managers, comment data for employees, and video data related to behaviors that should be noted in the target class.

[0008] An information processing method according to a second aspect of the present invention is an information processing method that supports the prevention of employee turnover, wherein a storage device stores at least a training dataset and proposed data corresponding to risk classification, a training device receives the training dataset, inputs the training dataset into a training target learning model, obtains processing results from the training target learning model, trains the training target learning model by comparing the correct labels included in the training dataset with the processing results, stores the trained learning model in the storage device, an inference device obtains data to be inferred, performs inference using the trained learning model and outputs a situation classification, and a identifier retrieves data from the storage device corresponding to the situation classification based on the situation classification. The system includes a device for identifying and outputting proposed data, a training dataset comprising at least one of the following: pairs of attendance data and correct labels, pairs of organizational culture diagnostic data and correct labels, and pairs of behavioral analysis data and correct labels, a trained learning model for inferring risk classification related to employee turnover, attendance data including desired shift days, confirmed shift days, first day of work, days worked, employees who have not submitted shifts for a certain period of time and their last day of work, organizational culture diagnostic data compiled and analyzed from questionnaire response data, and proposed data including comment data for managers, comment data for employees, and video data related to behaviors that should be noted in the target class.

[0009] A third aspect of the present invention is a program executed by an information processing device that supports preventing employee turnover, comprising: a computer, a storage device that stores at least a training dataset and proposed data corresponding to risk classifications; a training device that receives the training dataset, inputs the training dataset into a training target learning model, obtains processing results from the training target learning model, trains the training target learning model by comparing the correct labels included in the training dataset with the processing results, and stores the trained learning model in the storage device; an inference device that obtains inference target data, performs inference using the trained learning model, and outputs a situation classification; and a device that, based on the situation classification, retrieves data from the storage device corresponding to the situation classification. The system functions as a identifying device that identifies and outputs corresponding suggestion data. The training dataset includes at least one of the following: pairs of attendance data and correct labels, pairs of organizational culture diagnostic data and correct labels, and pairs of behavioral analysis data and correct labels. The trained learning model is for inferring the risk classification related to employee turnover. The attendance data includes the number of desired shift days, the number of confirmed shift days, the first day of work, the number of days worked, and employees who have not submitted shifts for a certain period of time and their last day of work. The organizational culture diagnostic data is compiled and analyzed from questionnaire response data. The suggestion data includes comment data for managers, comment data for employees, and video data related to behaviors that should be noted in the target class. [Effects of the Invention]

[0010] According to the present invention, it is possible to provide a technology that promotes changes in internal awareness and behavior to prevent employee turnover by accompanying and habituating these actions. [Brief explanation of the drawing]

[0011] [Figure 1] Figure 1 is a diagram showing the configuration of an information processing apparatus according to an embodiment of the present invention. [Figure 2] Figure 2 is a conceptual diagram showing processing by a training device and the like of an information processing device according to an embodiment of the present invention. [Figure 3]Figure 4 is a conceptual diagram showing the processing performed by the inference device and the like in an information processing device according to an embodiment of the present invention. [Figure 4] Figure 4 is a flowchart showing the processing procedure of the training process by the information processing device according to an embodiment of the present invention. [Figure 5] Figure 5 is a flowchart showing the processing procedure for inference and identification processing by an information processing device according to an embodiment of the present invention. [Figure 6] Figure 6 shows the results of the survey. [Figure 7] Figure 7 shows an example of a suggestion screen on an employee's terminal device. [Figure 8] Figure 8 illustrates the gradual effects of this device. [Modes for carrying out the invention]

[0012] One embodiment of the present invention will be described below with reference to the drawings.

[0013] Figure 1 shows and explains the configuration of an information processing device according to an embodiment of the present invention.

[0014] As shown in the figure, the information processing device 1 comprises a CPU (Central Processing Unit) 2 as an arithmetic processing device, a ROM (Read Only Memory) 3, a RAM (Random Access Memory) 4, an input / output device 5, a communication interface (I / F) 6, and a storage device 7. The storage device 7 may consist of, for example, an HDD (Hard Disk Drive) or an SSD (Solid State Drive). The storage device 7 stores a training dataset 21, inference target data 22, proposed data 23, a training target learning model 24, a trained learning model 25, a learning program 26, an inference / specification program 27, and the like.

[0015] During the training phase, the CPU 2 functions as a training device by reading the learning program 26 in the storage device 7, expanding it in the RAM 4, and executing it. The training dataset 21 is a set of pairs of training data and correct labels. The training data includes attendance data, organizational climate diagnosis data, behavior analysis data, and the like. The training device trains the learning model 24 to be trained based on the training dataset 21 and stores the trained learning model 25 in the storage unit 7.

[0016] Also, during the inference / specification phase, the CPU 2 functions as an inference device and a specification device by reading the inference / specification program 27 in the storage device 7, expanding it in the RAM 4, and executing it. The inference device receives the inference target data 22, executes inference using the trained learning model 25, and outputs the situation ID as the situation classification to the specification device. Here, the situation ID is an identification number that categorizes the situation in which the employee to be considered for retention prevention is placed. The inference target data 22 includes attendance data, organizational climate diagnosis data, behavior analysis data, and the like. The specification device reads out and outputs the proposal data corresponding to the situation ID from the storage device 7 based on the situation ID. The proposal data includes, but is not limited to, comment data for the administrator, comment data for the employee, video data related to the actions that the target employee should pay attention to, and the like. The specified proposal data is displayed on a display device (not shown) via the input / output device 5, or is transmitted to the administrator's terminal device via the communication I / F 6 by email, SNS, or the like.

[0017] FIG. 2 further details the training process by the training device of the information processing apparatus according to the embodiment of the present invention.

[0018] As shown in the figure, in the training stage, the CPU 2 functions as a training device 11 by reading out the learning program 26 of the storage device 7, expanding it in the RAM 4, and executing it. The training device 11 trains the training target learning model 24 with the training data set 21. The training data set 21 is a set of pairs of training data and correct labels, and the training data includes attendance and work data, organizational climate diagnosis data, behavior analysis data, and the like. The correct label may be a situation ID for the training data.

[0019] More specifically, the attendance and work data includes at least some of the following. · Desired shift days (days when part-time employees can work or days they hope to work, submitted to the store manager) · Confirmed shift days (days and time increase / decrease data of the shift confirmed for the desired shift) · Desired shift submission date (how many days before the deadline the desired shift was submitted) · First attendance date of newly hired employees · Number of days since newly hired employees started working · Total hours worked by newly hired employees · Number of days since the last work of newly hired employees · Number of days since the last attendance · Employees who have not submitted a shift for a certain period or longer and their last attendance date · Employees who have not worked for a certain period or longer and their last attendance date · Number of employees · Average tenure of employees at each store and each location · Number of days from employment to departure of employees

[0020] The questionnaire response data means all the response data of the organizational climate questionnaire of all employees (employees, part-time employees, part-timers) registered in the store. However, for the training data, the organizational climate diagnosis data obtained by aggregating and analyzing the questionnaire response data is used. As shown in FIG. 6, this organizational climate diagnosis data may be the quantification of work value, growth motivation, same sense of crisis, common values, work earthquake, acceptance of evaluation, respect for superiors, and gratitude to colleagues in 10 levels from -5 to +5.

[0021] Behavioral analysis data refers to data obtained as a result of analyzing measurement data from cameras, etc., using the analysis device 14, and includes at least some of the following: • The number of times senior employees speak to new employees, and the timing of those conversations. • Employees' time spent in each area • Ratio of eye movement directions among new employees (changes in the number of times they look from their hands to the entire store) • Changes in the area where new employees stay and their range of activity (shifts from a narrow area within the store to a wider area). • Whether or not a morning meeting is held, its start time, and its end time. • The gaze of the participants in the morning meeting • Percentage of participants speaking during the morning meeting • Employee's work details (frequency, duration) • The frequency of specific actions performed by employees (stopping work to ponder / search / think) (In the case of service industries) Number of customers waiting, waiting time, number of people staying, length of stay (In the service industry) Time taken from when a customer calls until an employee notices, number of people, waiting time, number of people staying, length of stay • (In the case of service industries) Time elapsed from the time a customer enters the store until initial service is provided. • (In the case of service industries) The time elapsed from the time an order is placed by the customer until the product is delivered. • (In the service industry) The time taken from the start to the completion of the customer's payment.

[0022] In other words, the CPU 2 also functions as an analysis device 14, and by analyzing the recorded data acquired via the input / output device 5 and the communication I / F 6 using the analysis device 14, it obtains the behavioral analysis data described above.

[0023] The training device 11, for example, performs processing on the training model 24 based on the training dataset 7, and trains the training model 24 based on the error between the processing result and the correct label. The trained training model 25 is then stored in the storage device 7.

[0024] In this way, the training device 11 trains the training target learning model 24 based on a training dataset that includes at least one of the following: pairs of attendance data and correct labels, pairs of organizational culture diagnosis data and correct labels, and pairs of behavioral analysis data and correct labels.

[0025] Figure 3 further illustrates the inference processing by the inference device and the identification processing by the identification device of the information processing apparatus according to an embodiment of the present invention.

[0026] As shown in the figure, the information processing device 1 receives input of inference target data 22 via the input / output device 5 and communication I / F 6 and stores it in the storage device 7. Of the inference target data 22, the behavioral analysis data is obtained as a result of analyzing camera measurement data with the analysis device 14 and is stored in the storage device 7. In the inference and identification stage, the CPU 2 reads the inference and identification program 27 from the storage device 7, loads it into RAM 4, and executes it, thereby functioning as the inference device 12 and identification device 13.

[0027] The inference device 12 receives the data to be inferred 22, performs inference using the trained learning model 25, and outputs the situation ID to the identification device 13. The trained learning model 25 is trained on a training dataset that includes at least one of the following: pairs of attendance data and correct labels, pairs of organizational culture diagnosis data and correct labels, and pairs of organizational culture diagnosis data and correct labels.

[0028] Based on this situation ID, the specific device 13 reads and outputs suggestion data corresponding to the situation ID from the storage device 7. This suggestion data includes comment data for the manager, comment data for the employee, and video data related to the employee's actions that should be monitored. In other words, the storage device 7 stores the situation ID in association with the comment data for the manager, the comment data for the employee, and the video data, making it possible to accurately read the suggestion data corresponding to the situation ID.

[0029] The identified suggestion data is displayed on a display device (not shown) via input / output device 5, or sent to the administrator's terminal device via communication I / F 6 via email or SNS. Examples of how the suggestion data is displayed on the administrator's terminal device are shown in Figures 7(a) and 7(b), for example. The content of the distribution includes comments to be given feedback to the new employee, changes in the new employee since the last time (camera measurement content), specific situations for praise (video), things to be careful of each week, and recommended examples from the training staff (camera measurement content).

[0030] The following describes the processing flow of the training process by the information processing device according to an embodiment of the present invention, with reference to the flowchart in Figure 4.

[0031] The CPU 2 functions as a training device 11 by reading the learning program 26 from the storage device 7, loading it into RAM 4, and executing it. The training data set 21 is a collection of pairs of training data and correct labels, and the training data includes attendance data, organizational culture diagnosis data, behavioral analysis data, etc. The training device 11 acquires the training data set 21 (S1), trains the target learning model 24 based on the training data set 21 (S2), and registers the trained learning model 25 in the storage unit 7 (S3).

[0032] Next, with reference to the flowchart in Figure 5, the processing flow of the inference and identification process by the information processing device according to an embodiment of the present invention will be explained.

[0033] During the inference and identification phase, CPU2 functions as an inference device 12 and an identification device 13 by reading the inference and identification program 27 from the storage device 7, loading it into RAM 4, and executing it. The inference device 12 acquires the data to be inferred 22 (S11), performs inference using the trained learning model 25, and outputs a situation ID to the identification device (S12). Based on this situation ID, the identification device 13 identifies the suggested data corresponding to the situation ID from the storage device 7 (S13) and outputs the read suggested data (S14). This suggested data may include comments for the manager, comment data for the employee, and video data related to the employee's actions that should be monitored. The identified suggested data may be displayed on a display device (not shown) via the input / output device 5, or sent to the manager's terminal device via email or SNS via the communication I / F 6.

[0034] As described above, according to the information processing device, etc., of the embodiment of the present invention, it is possible to detect in advance the timing when new employees are likely to experience a decline in morale, which could lead to them leaving the company, and to distribute information related to appropriate guidance to the receiving employees and part-time employees in charge of training, thereby preventing employees from leaving the company.

[0035] In other words, according to the information processing device, etc., as embodied in the present invention, for example as shown in Figure 7, it is possible to provide appropriate information at the appropriate time and encourage behavioral change, etc., so that the target employee can communicate with the supervisor (manager) in order to help the employee move up through the stages of Maslow's hierarchy of needs, such as "maintaining a livelihood / trust in the company," "personal safety," "fitting into the team / learning the job," and "recognition from others."

[0036] Although one embodiment of the present invention has been described above, it goes without saying that the present invention is not limited thereto and various improvements and modifications are possible without departing from its spirit.

[0037] For example, employee surveys can be conducted in various styles, not just those mentioned above. [Explanation of Symbols]

[0038] 1... Information processing device, 2... CPU, 3... ROM, 4... RAM, 5... Input / output device, 6... Communication interface, 7... Memory device, 11... Training device, 12... Inference device, 13... Identification device, 14... Analysis device.

Claims

1. An information processing device that supports the prevention of employee turnover, A memory device that stores at least a training dataset and proposed data corresponding to risk classification, A training device that receives a training dataset, inputs the training dataset into a training model, obtains processing results from the training model, trains the training model by comparing the correct labels included in the training dataset with the processing results, and stores the trained training model in a memory device. An inference device that acquires data to be inferred, performs inference using the pre-trained learning model, and outputs a situation classification, The system includes a device that identifies and outputs proposed data corresponding to the situation classification from the storage device based on the aforementioned situation classification, The aforementioned training dataset includes at least one of the following: pairs of attendance data and ground truth labels, pairs of organizational culture diagnostic data and ground truth labels, and pairs of behavioral analysis data and ground truth labels; and the trained learning model is for inferring the risk classification related to employee turnover. The aforementioned attendance data includes the number of days requested for shifts, the number of days confirmed for shifts, the first day of work, the number of days worked, and the last day of work for employees who have not submitted shifts for a certain period of time or longer. The aforementioned organizational culture diagnostic data is compiled and analyzed from survey response data. The proposed data includes comment data for administrators, comment data for employees, and video data related to the behaviors that should be monitored in the relevant class. Information processing device.

2. An information processing method that helps prevent employee turnover, The memory device stores at least the training dataset and the proposed data corresponding to the risk classification. death, The training device receives the training dataset, inputs the training dataset into the training target learning model, obtains processing results from the training target learning model, trains the training target learning model by comparing the correct labels included in the training dataset with the processing results, and stores the trained learning model in the storage device. An inference device that acquires data to be inferred, performs inference using the trained learning model, and outputs a situation classification, A specific device that identifies and outputs proposed data corresponding to the situation classification from the storage device based on the situation classification, The aforementioned training dataset includes at least one of the following: pairs of attendance data and ground truth labels, pairs of organizational culture diagnostic data and ground truth labels, and pairs of behavioral analysis data and ground truth labels; and the trained learning model is for inferring the risk classification related to employee turnover. The aforementioned attendance data includes the number of days requested for shifts, the number of days confirmed for shifts, the first day of work, the number of days worked, and the last day of work for employees who have not submitted shifts for a certain period of time or longer. The aforementioned organizational culture diagnostic data is compiled and analyzed from survey response data. The proposed data includes comment data for administrators, comment data for employees, and video data related to the behaviors that should be monitored in the relevant class. Information processing methods.

3. A program executed by an information processing device that helps prevent employee turnover, Computers, A memory device that stores at least a training dataset and proposed data corresponding to risk classification, A training device that receives the training dataset, inputs the training dataset into the training target learning model, obtains processing results from the training target learning model, trains the training target learning model by comparing the correct labels included in the training dataset with the processing results, and stores the trained learning model in the storage device, An inference device that acquires data to be inferred, performs inference using the pre-trained learning model, and outputs a situation classification, Based on the aforementioned situation classification, the device functions as a specific device that identifies and outputs proposed data corresponding to the situation classification from the storage device. The aforementioned training dataset includes at least one of the following: pairs of attendance data and ground truth labels, pairs of organizational culture diagnostic data and ground truth labels, and pairs of behavioral analysis data and ground truth labels; and the trained learning model is for inferring the risk classification related to employee turnover. The aforementioned attendance data includes the number of days requested for shifts, the number of days confirmed for shifts, the first day of work, the number of days worked, and the last day of work for employees who have not submitted shifts for a certain period of time or longer. The aforementioned organizational culture diagnostic data is compiled and analyzed from survey response data. The proposed data includes comment data for administrators, comment data for employees, and video data related to the behaviors that should be monitored in the relevant class. program.

Images