Processing method, processing device, and processing program

Abstract

A processing device (10) has: a collection unit (131) that collects an operation log from each terminal utilized by a plurality of staff members who perform jobs; a proficiency factor extraction unit (132) that extracts feature quantities for each staff member from the operation logs collected by the collection unit (131); and a proficiency level calculation unit (134) that calculates the job proficiency level of each staff member from the feature quantities extracted for each staff member by the proficiency factor extraction unit (132), wherein the proficiency level calculation unit (134) calculates the proficiency level on the basis of the job that was being performed by the staff member when the operation log was collected, the proficiency level being calculated by applying a proportionally heavier weight for a longer time required for the job, or by applying a proportionally heavier weight for a higher mistake rate.

Description

[Technical field]

[0001] The present invention relates to a processing method, a processing device, and a processing program. [Background technology]

[0002] There are many cases where business processes are visualized for the purpose of improving efficiency and automating them. For example, the technology described in Non-Patent Document 1 visualizes business processes by introducing a transaction perspective in order to classify the processes. [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Chihiro Suematsu, Shintaro Sengoku, Yasuo Matsubara, “Empirical Study of Venture Management Support Through Business Visualization (Transaction Measurement and Analysis)”, Japan Society for Management Information, November 2008, 2008 Fall National Research Presentation Conference B2-5., [online], [Retrieved March 16, 2021], Internet<URL:https: / / www.jstage.jst.go.jp / article / jasmin / 2008f / 0 / 2008f_0_38 / _pdf / -char / en> Summary of the Invention [Problem to be solved by the invention]

[0004] As described in Non-Patent Document 1, it is extremely rare for work to be completely identical from organization to organization or from individual to individual, making it difficult to uniquely evaluate work, and evaluation is often conducted for the organization or group as a whole (hereinafter referred to as the "organization").

[0005] However, while it is possible to optimize the workload of the entire organization, for example by reviewing work flows, the total productivity of each individual employee in the organization cannot be improved.

[0006] The present invention has been made in consideration of the above, and has an object to provide a processing method, processing device, and processing program capable of calculating the proficiency level for each person belonging to an organization. [Means for solving the problem]

[0007] In order to solve the above-mentioned problems and achieve the object, the processing method of the present invention includes a step of collecting operation logs of each terminal used by a plurality of personnel performing work, a step of extracting features for each personnel from the collected operation logs, and a step of calculating the work proficiency for each personnel from the features extracted for each personnel, and is characterized in that the calculation step calculates the proficiency based on the work that the personnel was performing when the operation logs were collected, by weighting the work so that the longer the time required for the work, or the higher the error rate, the greater the weighting. Effect of the Invention

[0008] According to the present invention, it is possible to calculate the proficiency level for each person belonging to an organization. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram for explaining an outline of the process according to the embodiment. [Diagram 2] FIG. 2 is a diagram illustrating an example of a configuration of a processing apparatus according to an embodiment. [Diagram 3] FIG. 3 is a diagram showing an example of a Windows desktop screen. [Figure 4] FIG. 4 is a diagram illustrating a user interface (UI) tree. [Diagram 5] FIG. 5 is a diagram for explaining the operation logs collected by the collection unit shown in FIG. 2 and the purpose of collection thereof. [Figure 6] FIG. 6 is a diagram for explaining the function of the operation data collection agent. [Figure 7] FIG. 7 is a diagram illustrating an example of a list of proficiency elements extracted from an operation log. [Figure 8]FIG. 8 is a diagram for explaining the counting pattern of the proficiency elements. [Figure 9] FIG. 9 is a diagram in which personnel are associated with the tasks performed by each personnel. [Figure 10] FIG. 10 is a diagram showing data on the time required or error rate when a task is performed by personnel. [Figure 11] FIG. 11 is a diagram showing a calculation formula for the task difficulty coefficient. [Figure 12] FIG. 12 is a diagram showing a table in which the period of work experience, the years of service, the rate of return, and the correlation coefficients with the proficiency elements correspond to each other. [Figure 13] FIG. 13 is a diagram showing a table in which the period of work experience, the years of service, the rate of return, and the correlation coefficients with the proficiency elements correspond to each other. [Figure 14] FIG. 14 is a diagram in which the value of each proficiency element is associated with the period of work experience of the team. [Figure 15] FIG. 15 is a diagram in which the value of each proficiency element is associated with the work experience period of the team. [Figure 16] FIG. 16 is a diagram in which the value of each skill element is associated with the number of years of work experience. [Figure 17] FIG. 17 is a diagram in which the value of each skill element is associated with the number of years of work experience. [Figure 18] FIG. 18 is a diagram in which the value of each familiarity element is associated with the return rate. [Figure 19] FIG. 19 is a diagram showing correlation coefficients between the indexes and the proficiency elements tabulated using tabulation pattern 1. [Figure 20] FIG. 20 is a diagram showing correlation coefficients between the indexes and the proficiency elements tallied using the tallying pattern 2. In FIG. [Figure 21] FIG. 21 is a flowchart showing a processing procedure of the processing method according to the embodiment. [Figure 22] FIG. 22 is a diagram for explaining an outline of a processing method according to an embodiment. [Figure 23] FIG. 23 is a diagram illustrating an example of a computer in which a processing device is realized by executing a program. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to this embodiment. In addition, in the description of the drawings, the same parts are indicated by the same reference numerals.

[0011] [Embodiment Mode] In this embodiment, a technique for calculating the proficiency level of each of multiple personnel belonging to an organization based on operation logs of terminals used by the personnel will be described.

[0012] FIG. 1 is a diagram for explaining an outline of the process in the embodiment. As shown in FIG. 1, first, in the processing method according to the embodiment, an agent is introduced into a business PC (terminal) used by each employee to collect operation data (operation log) ((1) in FIG. 1). Windows (registered trademark) is used as the OS in the business PC. The OS of the business PC may be an OS other than Windows. Each employee may use a tablet or the like instead of the business PC. In the present embodiment, operation characteristics that are highly correlated with the proficiency of an individual are extracted as proficiency elements (feature amounts) from the operation log ((2) in FIG. 1). In the present embodiment, the proficiency elements of multiple employees are statistically analyzed to calculate the difficulty level of each task, for example, a task difficulty coefficient (weighting value) ((3) in FIG. 1).

[0013] In this embodiment, the proficiency of each personnel is calculated based on proficiency elements that have a high correlation with proficiency, such as speed (time required, shortcuts, copying) and accuracy (correction), while taking into account the difficulty of the task ((4) and (5) in FIG. 1). In other words, in this embodiment, it is possible to quantitatively grasp the proficiency of personnel while also taking into account the difficulty of the task and the accuracy of the task ((6) in FIG. 1). Then, using the proficiency calculated for each personnel, it becomes possible to appropriately allocate personnel according to the proficiency of each personnel and the productivity that can be estimated from the proficiency, thereby improving the productivity of the organization.

[0014] [Processing device] Next, a processing device for calculating the proficiency level of each employee will be described. Fig. 2 is a diagram showing a schematic configuration example of a processing device according to an embodiment. A processing device 10 according to embodiment 1 shown in Fig. 2 includes a communication unit 11, a storage unit 12, and a control unit 13.

[0015] The communication unit 11 is a communication interface that transmits and receives various information to and from other devices connected via a network or the like. The communication unit 11 is realized by a NIC (Network Interface Card) or the like, and performs communication between other devices and the control unit 13 (described later) via electric communication lines such as a LAN (Local Area Network) or the Internet. For example, the communication unit 11 receives operation logs from the work PCs of each employee via the network, and outputs the logs to the control unit 13. The communication unit 11 also outputs information indicating the proficiency of each employee calculated by the control unit 13 to an external device via the network.

[0016] The storage unit 12 is a storage device such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), or an optical disk. The storage unit 12 may be a semiconductor memory in which data can be rewritten, such as a RAM (Random Access Memory), a flash memory, or a NVSRAM (Non Volatile Static Random Access Memory). The storage unit 12 stores an OS (Operating System) and various programs executed by the processing device 10. Furthermore, the storage unit 12 stores various information used in the execution of the programs. The storage unit 12 stores an operation log 121 of the work PC used by each staff member, and difficulty level data 122 indicating the difficulty level of each work.

[0017] The control unit 13 controls the entire processing device 10. The control unit 13 is, for example, an electronic circuit such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit), or an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array). The control unit 13 also has an internal memory for storing programs that define various processing procedures and control data, and executes each process using the internal memory. The control unit 13 also functions as various processing units by the operation of various programs. The control unit 13 has a collection unit 131, a proficiency element extraction unit 132 (extraction unit), a difficulty acquisition unit 133, and a proficiency calculation unit 134 (calculation unit).

[0018] The collection unit 131 collects operation logs of the work PCs of multiple personnel. The collection unit 131 uses an API (Application Programming Interface) provided by Windows to acquire events that occur when the focus (a screen element operated by a personnel) moves, thereby acquiring operation logs for the operation target across systems (applications). Note that the logs themselves are directly acquired by the work PC and the storage unit 12 (see FIG. 2), and the collected logs may be consolidated in another terminal, server, or the storage unit 12 (see FIG. 2).

[0019] The proficiency element extraction unit 132 extracts proficiency elements for each personnel from the operation logs collected by the collection unit 131. There are multiple types of proficiency elements. The proficiency elements include, for example, the number of times the alt key on the keyboard of the work PC is used, the number of times a mouse connected to the work PC is used, and the number of times a paste operation is performed among operations on the work PC, all of which are acquired as operation logs.

[0020] The difficulty level acquisition unit 133 acquires a task difficulty coefficient for each task performed by each personnel. The task difficulty coefficient is used as a weighting value for correcting bias between tasks when the proficiency calculation unit 134 calculates the proficiency of each personnel. The difficulty level acquisition unit 133 calculates the task difficulty coefficient based on a plurality of operation logs corresponding to a plurality of personnel. The difficulty level acquisition unit 133 calculates a task difficulty coefficient that increases as the time required for the task increases, or a task difficulty coefficient that increases as the error rate increases, based on the task that the personnel was performing when the operation log was collected.

[0021] The task difficulty coefficient is a value calculated at a predetermined timing for each task environment of a person. The task environment is a group or a site where a certain task is performed. The difficulty acquisition unit 133, for example, periodically calculates the task difficulty coefficient and stores it in the memory unit 12. Alternatively, the difficulty acquisition unit 133 calculates the task difficulty coefficient when the task environment changes. The difficulty acquisition unit 133 may also adjust the task difficulty coefficient as appropriate within the period in which the task is performed. The difficulty acquisition unit 133 may also change the calculation timing of the task difficulty coefficient depending on the task.

[0022] The proficiency calculation unit 134 calculates the work proficiency for each person from the proficiency elements generated for each person. The proficiency calculation unit 134 calculates the proficiency for each person from the proficiency elements based on the relationship between the proficiency elements and the proficiency. The proficiency calculation unit 134 calculates the proficiency for each person based on the relationship between the proficiency elements and the proficiency that is obtained in advance, in which the more the alt key is used, the higher the proficiency, the less the mouse is used, and the less the paste operation is performed, the higher the proficiency. The proficiency calculation unit 134 selects a type of proficiency element according to the work environment of the person from among the multiple proficiency elements, and calculates the proficiency for each person based on the selected type of proficiency element.

[0023] The proficiency calculation unit 134 uses the task difficulty coefficient as a weighting value for correcting the bias between tasks. The proficiency calculation unit 134 calculates the proficiency by weighting the task difficulty coefficient, which increases as the time required for the task increases, or the task difficulty coefficient, which increases as the error rate increases, based on the task that the person was performing when the operation log was collected.

[0024] [Collection Department] Next, a description will be given of each functional unit of the processing device 10. First, a description will be given of the collection unit 131. Fig. 3 is a diagram showing an example of a desktop screen of Windows. Fig. 4 is a diagram showing an example of a UI tree.

[0025] The collection unit 131 uses an API provided by Windows to acquire events that occur when the focus (the screen element to be operated) moves, thereby realizing acquisition of information on the object to be operated across systems (applications).

[0026] For example, a case where the focus moves to an element of input form F1 (FIGS. 3 and 4(1)) will be described as an example. The collection unit 131 can obtain the process and window title displaying the form for the element in the screen to which the focus has moved, by tracing back the UI tree (FIG. 4(a)) managed by Windows to the desktop. The collection unit 131 may use the process name to identify the system name, or may obtain a window title determined by the system. The window title is displayed according to the system and its status. For example, in the case of IE (Internet Explorer), the window title is different for each URL. The window title can be obtained from the UI tree when the focus moves or a click occurs.

[0027] Here, the collection unit 131 acquires only the focused element (e.g., input form F1) and information on the element's direct ancestors in the UI tree (process, screen name, URL, etc.) as an operation log ((2) in FIG. 4). For example, for input form F1, the collection unit 131 acquires information on the pane (in the case of IE, URL, etc.) which is the direct ancestor of input form F1, the window (process, window title, and its screen name), and the desktop. By using information on the focused element's direct ancestors in the UI tree, the collection unit 131 is able to both output an operation log with a smaller amount of data compared to the entire tree and extract proficiency elements with sufficient detail and granularity for analysis.

[0028] Fig. 5 is a diagram for explaining the operation logs collected by the collection unit 131 shown in Fig. 2 and the purpose of collecting them. As shown in Fig. 5, the collection unit 131 collects operation logs of key inputs in order to quantify the number of times input contents are corrected and shortcuts are used, and to determine the time of no operation. The collection unit 131 collects operation logs of mouse cursor and wheel movement in order to quantify mouse operation contents and to determine the time of no operation. The collection unit 131 collects operation logs of printing in order to estimate work outside the system, such as leaving the desk after printing.

[0029] The collection unit 131 collects operation logs of copy (update clipboard) to quantify copy operations. The collection unit 131 collects operation logs of paste (Ctrl+V or "Paste" in a menu) to quantify paste operations. The collection unit 131 collects operation logs of focus (screen element to be operated) movement (excluding buttons in a context menu, etc.) to infer the content of a task from element information of the focus movement destination, detect task boundaries, and determine no-operation time. The collection unit 131 collects operation logs of execution (Invoke) to detect task boundaries due to information transmission and screen transitions accompanying button clicks.

[0030] The collection unit 131 collects an operation log of a left click in order to detect a left click of the mouse and the target screen element. The collection unit 131 collects a period of no operation for a certain period of time or more as an idle operation log. The collection unit 131 collects an operation log of the transition of a target window in order to detect a break. The collection unit 131 collects an operation log of a screen transition (change in window title) in order to detect a break.

[0031] Fig. 6 is a diagram for explaining the function of the operation data collection agent. The collection unit 131 sets the function of the operation data collection agent as follows. The operation data collection agent automatically starts when the PC is started (signed in) according to startup registration, and automatically stops when sign-out is detected (Fig. 6 (1)). This reduces the burden on the user, prevents the agent start and stop operations themselves from being output as an operation log, and prevents loss of operation logs due to forgetting to start or stop the agent.

[0032] In addition, the operation data collection agent sequentially outputs the operation log to a predetermined file set in advance, and automatically sends the predetermined file as an operation log to a file server (copying it to the set file path) periodically (initial value: every hour) ((2) in FIG. 6). This makes it possible to centralize the conversion process from multiple people's quantification data to proficiency element data and the subsequent analysis process and score calculation, making it easier to aggregate data from multiple people. It also makes it possible to limit the range of data loss due to failures on the business PC side, etc., in units of transmission intervals. Note that instead of outputting to a log file, the business PC may transmit the operation log or proficiency factors to the processing device 10 that judges the proficiency in a streaming manner.

[0033] The operation data collection agent then automatically deletes the operation log files that were successfully sent and any files that have been there for a certain period of time (default: one month) (Figure 6 (2)). This makes it possible to reduce the impact on the storage resources of the work PC and limit the extent of data loss during the data retention period due to a failure on the file server side, etc.

[0034] [Familiarity element extraction section] Next, the proficiency element extraction unit 132 will be described. The proficiency element extraction unit 132 extracts proficiency elements by converting the operation log into data (quantified data) that quantitatively represents the behavior of each person during the work. The elements that are used to characterize the proficiency elements should be designed according to the characteristics of the group and the work. The appropriate elements differ depending on the group, site, and characteristics of the work where a certain work is performed. For example, elements in the UI (input forms, buttons, etc.) may be used to characterize. Since the proficiency elements can analyze the behavior of each person in detail, it is possible to point out habits and difficult tasks in detail and encourage improvement. The proficiency element extraction unit 132 also quantifies behaviors such as shortcuts, copy, paste, and the number of times the Backspace / Delete key is used for each person, and this enables analysis from perspectives other than the required time that is mainly collected as data, such as individual operation characteristics and accuracy.

[0035] Specifically, the extraction degree elements extracted by the proficiency element extraction unit 132 will be described. Fig. 7 is a diagram illustrating an example of a list of proficiency elements extracted from operation logs. The list Dp shown in Fig. 7 illustrates an example of a case where operation data by personnel A is extracted from a plurality of operation logs, and the number of times of various operations is extracted and quantified for each period. Note that the list Dp is intermediate data processed by the proficiency element extraction unit 132, and for example, items are selected and each numerical value is processed according to the work environment of personnel A and the period for which the proficiency is desired to be calculated.

[0036] List Dp has items such as person in charge, start date and time, end date and time, required time, required time before transition, required time after operation, app (application), screen name (URL), screen element path, screen element type, total number of key inputs, number of Back / Del operations (number of Backspace / Delete key operations), number of shortcuts, number of mouse movement detections, number of wheel detections, number of clicks, number of copies, number of pastes, and number of prints.

[0037] The person in charge field is the sign-in account name or designated, and list Dp is an example of the extraction of proficiency elements for personnel A. The start date and end date and time are in UNIX (registered trademark) milliseconds (a timestamp that is not affected by time differences). The required time is calculated from the start date and end date and time, and is measured in milliseconds. The time required before the transition is the time required for the last mouse movement and wheel detection only out of the time required for the previous record in the operation sequence. The time required after the operation is the time required for the last mouse movement and wheel detection only out of the time required for the current record. For screen name (URL), the URL is only added if it is IE and can be obtained.

[0038] Screen element paths use a UI tree structure as an identifier because not all screen elements (forms, etc.) have names. Note that even if you click on the title bar or a blank space on the screen simply to bring the window to the foreground, it will be output as one record, but it is possible to eliminate this during analysis using rules such as the time required, screen element type, and total number of keystrokes.

[0039] The number of shortcuts is actually divided into columns such as Ctrl+, Win+, Alt+, Tab, F2~F12, etc. The number of mouse movement detections is counted as 1 time, for example, when the cursor moves 100 pixels or more in either the horizontal or vertical direction. The number of wheel detections is detected at 0.1 second intervals. The number of mouse movement detections is considered to be 1 time if multiple detections are made within 0.1 seconds. When using an application mouse with smooth scrolling enabled, a large number of wheel operations are detected in a short period of time, so a fixed detection interval is set. The number of clicks can also be distinguished between left, center, and right clicks.

[0040] The start date and time, end date and time, required time, required time before the transition, and required time after the operation (La in Fig. 7) are used to analyze the proficiency of personnel. For example, if the average required time for all personnel is large, this task is deemed to be a bottleneck. Also, if the required time is short compared to the average for personnel, this task is deemed to be a task that personnel A is good at, and if the required time is long, this task is deemed to be a task that personnel A is not good at.

[0041] The application, screen name (URL), screen element path, and screen element type (Lb in Fig. 7) are used to identify a series of tasks, in other words, to specify the location of the tasks. The total number of keystrokes, number of Back / Del, number of mouse movements detected, number of wheel detections, and number of clicks (Lc in Fig. 7) are used to analyze personnel proficiency. For example, in the Lc item in Fig. 7, if the average value for all personnel is high, this task can be considered to be a bottleneck, if the value is small compared to the personnel average, this task can be considered to be a task that Person A is good at, and if the value is small, this task can be considered to be a task that Person A is not good at.

[0042] The number of shortcuts, copies, and pastes (Ld in FIG. 7) are used to analyze personnel proficiency. For example, if the value in the Ld item in FIG. 7 is large, it is possible that efficiency has been improved and that this has contributed to reducing the amount of time required. In addition, the number of prints (Le in FIG. 7) is used to analyze the appropriateness of idle time. The correlation between each of the above items and proficiency will be explained in the explanation of the processing by the proficiency calculation unit 134.

[0043] Then, the proficiency element extraction unit 132 may total the proficiency elements for a certain unit of the operation log. Fig. 8 is a diagram for explaining a tallying pattern of proficiency elements. For each personnel, the proficiency element extraction unit 132 regards operations performed while focusing on a certain application as one group, and totals the operation contents (number of times and time of each operation) for each group to obtain the proficiency element of the personnel. At that time, the proficiency element extraction unit 132 also divides each proficiency element by the workload to perform a totalization that reduces the influence of the number of cases.

[0044] Specifically, the proficiency element extraction unit 132 adds up each feature (proficiency element) and the required time for each block (application (e.g., IE)) ((1) in FIG. 8). Then, the proficiency element extraction unit 132 aggregates the proficiency elements, for example, using aggregation pattern 1 or aggregation pattern 2. As a result, in aggregation pattern 1, proficiency element records are created that are the number of personnel times 2 (number of months: October, November), and in aggregation pattern 2, proficiency element records are created that are the number of personnel times the number of days for which data exists.

[0045] [Difficulty level acquisition part] Next, a description will be given of the task difficulty coefficient and the difficulty acquisition unit 133. Fig. 9 is a diagram in which personnel are associated with tasks performed by each personnel.

[0046] As shown in FIG. 9, multiple personnel may perform multiple types of tasks with different task difficulty levels during a certain period of time. Also, the same personnel may perform the same task multiple times. In such cases, it is not appropriate to compare personnel proficiency levels using a score calculated only based on the task speed (time required). For this reason, in this embodiment, the proficiency score is corrected by weighting each task (multiplying by a task difficulty coefficient) to calculate a proficiency score that more fairly and objectively reflects the task skill level.

[0047] This task difficulty coefficient needs to reflect the average and variance of task speed and accuracy (proportion of Backspace / Del keys). For this reason, the task difficulty coefficient needs to satisfy the following conditions (1) to (3). (1) The task difficulty coefficient is a numerical value within a certain range. The task coefficient is designed to be a coefficient between 0 and 1, for example, in order to be used for weighting each task when calculating the proficiency level. (2) The larger the average time required and the average error rate (the larger the volume of work), the larger the value. (3) The greater the variation in average time required and average error rate among personnel (the greater the difference in skill level at the job, the greater the room for learning), the larger the value.

[0048] FIG. 10 is a diagram showing data on the time required or error rate when a task is performed by a person. The time required to perform a task is the time required for that person to complete the task. The error rate when performing a task is, for example, the ratio of the number of times the backspace / delete key is pressed to the total number of keystrokes made by that person. FIG. 11 is a diagram showing a calculation formula for the task difficulty coefficient. FIG. 10 shows a set (set a) of data on the time required for the task or the error rate for the task, and a set (set b) of the average values ​​of the time required for the task or the error rate data for the task, for tasks that have been performed multiple times, which are applied to formula (1) described below.

[0049] The difficulty level acquisition unit 133 obtains a coefficient that satisfies the condition of the task difficulty level, for example, by the formula (1) in FIG. 11. The difficulty level acquisition unit 133 obtains a coefficient by the formula (1) for each task type in terms of both the time required for the task and the accuracy (error rate). As shown in the formula (1), the coefficient has a larger value as the time required for the task increases. In addition, the coefficient has a larger value as the error rate increases. The difficulty level acquisition unit 133 uses the formula (1) for both the time required and the accuracy. Then, the difficulty level acquisition unit 133 multiplies the coefficient calculated using the formula (1) for the time required by the task by the coefficient calculated using the formula (1) for the error rate, and sets the product as the task difficulty coefficient for the task. Therefore, the task difficulty coefficient becomes larger as the time required for the task increases, or the error rate increases. Note that the difficulty level acquisition unit 133 may normalize the task difficulty coefficient again after calculating the coefficients for all task types.

[0050] The difficulty level acquisition unit 133 periodically calculates the task difficulty coefficient. In addition, the difficulty level acquisition unit 133 calculates the task difficulty coefficient when the task environment changes. The difficulty level acquisition unit 133 outputs the calculated task difficulty coefficient of each task to the mastery calculation unit 134. In addition, the difficulty level acquisition unit 133 stores the calculated task difficulty coefficient of each task in the storage unit 12.

[0051] [Proficiency Calculation Section] Next, the proficiency calculation unit 134 will be described. The proficiency calculation unit 134 calculates the proficiency of each person based on the relationship between the proficiency element and the proficiency of each person. In fact, the operation logs of 19 people performing slip insertion proxy work for a total of 18 business days were obtained, and the relationship between the value of the proficiency element extracted from the operation data of each person and the proficiency was obtained.

[0052] 12 and 13 are diagrams showing tables associating the team's work experience period, years of service, and send-back rate with correlation coefficients between proficiency elements. The send-back rate indicates the percentage of forms that were created but had errors and had to be retaken, relative to the number of forms created. The work experience period, years of service, and send-back rate are candidate indicators of proficiency.

[0053] FIG. 12 shows the relationship between the proficiency elements tabulated using tally pattern 1 and the proficiency level, and FIG. 13 shows the relationship between the proficiency elements tabulated using tally pattern 2 and the proficiency level. FIG. 14 and FIG. 15 are diagrams in which the value of each proficiency element corresponds to the period of work experience of the team. FIG. 16 and FIG. 17 are diagrams in which the value of each proficiency element corresponds to the number of years of work experience. FIG. 18 is a diagram in which the value of each proficiency element corresponds to the return rate. FIG. 14 and FIG. 16 use the proficiency elements tabulated using tally pattern 1. FIG. 15, FIG. 17, and FIG. 18 use the proficiency elements tabulated using tally pattern 2.

[0054] 16 to 18, the proficiency elements shown are the total number of keystrokes (Typing), number of Back / Del operations (Back / Del), number of shortcuts (Short Cut), number of shortcuts using the Ctrl key (SC_Ctrl), number of shortcuts using the Alt key (SC_Alt), number of shortcuts using the Win key (SC_Win), number of shortcuts using the Function key (SC_Func), number of shortcuts using the Enter key (SC_Enter), number of shortcuts using the Tab key (SC_Tab), number of mouse movement detections (Mouse Move), number of wheel detections (Mouse Wheel), number of clicks (Click), number of Invoke execution detections (Invoke), number of copies (Copy), number of pastes (Paste), and number of prints (Print).

[0055] As shown in box W1 of Figure 8 and Figure 14, the longer a team's work experience, the less typing they tend to do. As shown in box W2 of Figures 12, 13 and 14, the longer a team's work experience, the more they tend to use the Alt key. Also, as shown in box W3 of Figures 12, 13 and 14 and box W13 of Figure 15, the longer a team's work experience, the less mouse cursor / wheel movement and clicks they tend to do. Also, as shown in box W4 of Figures 12, 13 and 14, the longer a team's work experience, the less pasting they tend to do.

[0056] In this way, it is thought that there is a certain relationship between the proficiency elements, such as the number of times typing, shortcuts using the Alt key, and the number of times the mouse is used, including moving the mouse cursor / wheel and clicking, and the length of work experience.

[0057] Also, as shown in the frame W21 in Figures 12, 13 and 16, the longer the years of service, the less typing tends to occur. As shown in the frame W22 in Figures 12, 13 and 16 and W31 in Figure 17, the longer the years of service, the more the number of times the Enter key is pressed. As shown in the frame W23 in Figures 12, 13 and 16 and W32 in Figure 17, the longer the years of service, the less the number of times the Tab key is pressed. As shown in the frame W24 in Figures 12, 13 and 16 and W33 in Figure 17, the longer the years of service, the less the number of times the mouse cursor / wheel is moved and clicked. Also, as shown in the frame W25 in Figures 12, 13 and 16, the longer the years of service, the less the number of times the paste is performed.

[0058] In this way, it is thought that there is a certain relationship between the proficiency elements, such as the number of times typing, the number of times pressing the Enter key, the number of times using the Tab key, the number of times moving the mouse cursor / wheel and the number of times clicking, and the number of times pasting, and the number of years of employment.

[0059] Also, as shown in box W41 in Figures 12, 13, and 18, the higher the rejection rate, the lower the number of mouse cursor / wheel movements and clicks. In this way, it is thought that there is a certain relationship between the rejection rate and the number of times the mouse is used, including the number of mouse cursor / wheel movements and clicks, which are proficiency elements.

[0060] 19 and 20 are diagrams showing correlation coefficients between indexes and proficiency elements. Fig. 19 shows correlation coefficients between indexes and proficiency elements tallied using counting pattern 1. Fig. 20 shows correlation coefficients between indexes and proficiency elements tallied using counting pattern 2.

[0061] Unlike Figures 14 to 18, Figures 19 and 20 show proficiency factors that were collected and correlation analysis was performed only on data from IE operations that are mainly used for business. In addition to the frequency of various operations, TitleChanges, UniqTitles, UniqTitlesRaw, and SCRate have been added as proficiency factors. TitleChanges is the frequency at which the window title changes. UniqTitles is the frequency of window titles excluding duplicate titles. UniqTitlesRaw is the number of window titles excluding duplicate titles, and is the value itself rather than the frequency divided by the time required. SCRate is the ratio of the number of shortcut operations to the total number of key operations.

[0062] 19 and 20, the processing speed (manager's subjective opinion), the work experience of the team, the level of proficiency (manager's subjective opinion), the number of processed cases, the number of returned cases, and the return rate are candidate indicators of proficiency. In FIG. 19 and FIG. 20, the correlation coefficient between the indicator and the proficiency element is written at the intersection of each. By looking at the correlation coefficient at the intersection of the indicator and the proficiency element, the relationship between the proficiency element and the indicator can be recognized. Therefore, the proficiency element to be used in calculating the proficiency can be selected by considering the relationship between the indicator and the proficiency element.

[0063] As described above, the proficiency elements are related to the indices of proficiency candidates, such as the period of work experience of the team, years of service, processing speed (manager's subjective), level of proficiency (manager's subjective), number of processed cases, number of returned cases, and return rate, depending on the type of proficiency element. Based on such relationships between the proficiency elements and proficiency, the proficiency calculation unit 134 selects a proficiency element corresponding to the work environment of the personnel, and calculates the proficiency of each personnel. For this reason, first, the proficiency calculation unit 134 selects a proficiency element of a type corresponding to the work environment for each work environment of the personnel.

[0064] Then, the proficiency calculation unit 134 calculates the proficiency for each work environment using a corresponding proficiency calculation model. The proficiency calculation model is learned in advance using a data set of the proficiency previously given to the personnel and the proficiency elements selected according to the work environment as teacher data. Each proficiency calculation model learns a data set of the proficiency elements selected according to each work environment and the proficiency previously given to the personnel. The proficiency calculation unit 134 selects a proficiency calculation model corresponding to the work environment of the personnel to be calculated, inputs the proficiency elements of the personnel to be calculated into the selected proficiency calculation model, and obtains the output proficiency to calculate the proficiency of the personnel.

[0065] The proficiency calculation unit 134 may also calculate the proficiency of each person by using a preset calculation formula for each work environment of the person. In this case, the proficiency calculation unit 134 calculates the proficiency of the person to be calculated by applying the proficiency elements selected according to each work environment from among the proficiency elements of the person to be calculated to the calculation formula.

[0066] The formula for calculating the proficiency level is designed, for example, as follows, based on the relationships in Figures 12, 13, 19, and 20. The formula is designed so that the more time it takes to type, the lower the formula, the higher the formula is, and the more the Alt key is used, the lower the formula is. The formula may be designed so that the more the mouse is used, such as cursor movement, wheel operation, and click operation, the lower the formula is. In addition, a proficiency element that has a positive correlation between the proficiency level and the work experience period, years of service, or return rate, may be applied to the formula as is, and a proficiency element that has a negative correlation with both may be the sum of the reciprocals of the values ​​of the elements. In addition, if only one of them is positive, it may be designed according to the purpose.

[0067] [Processing procedure] Fig. 21 is a flowchart showing the processing procedure of the processing method according to the embodiment. As shown in Fig. 21, in the processing device 10, the collection unit 131 collects operation logs of work PCs of multiple personnel (step S11). Next, the proficiency element extraction unit 132 extracts proficiency elements for each personnel from the operation logs collected by the collection unit 131 (step S12).

[0068] Next, the difficulty level acquisition unit 133 determines whether it is time to calculate a task difficulty level coefficient (step S13). The task difficulty level coefficient is calculated when a predetermined period of time has passed since the previous calculation, when the task environment has changed, etc.

[0069] If it is time to calculate the task difficulty coefficient (step S13: Yes), the difficulty acquisition unit 133 calculates the task difficulty coefficient of each task, for example, using formula (1) in Fig. 11 (step S14). If it is not time to calculate the task difficulty coefficient (step S13: No), the difficulty acquisition unit 133 reads and acquires the task difficulty coefficient used to calculate the proficiency level from the storage unit 12 (step S15).

[0070] The proficiency calculation unit 134 calculates the work proficiency for each person from the proficiency elements generated for each person (step S16). In step S16, the proficiency calculation unit 134 selects a proficiency element of a type corresponding to the work environment of the person from among the multiple proficiency elements, and calculates the proficiency for each person based on the proficiency element of the selected type. The proficiency calculation unit 134 outputs the calculated proficiency of the person (step S17).

[0071] [Effects of the embodiment] In the embodiment, the operation logs of each business PC used by multiple personnel who perform the work are collected, proficiency elements are extracted for each personnel from the collected operation logs, and the work proficiency of each personnel is calculated from the proficiency elements extracted for each personnel. In this way, in the embodiment, the proficiency of each personnel belonging to the organization can be calculated appropriately. Then, based on the proficiency calculated for each personnel, it becomes possible to appropriately allocate personnel according to the proficiency of each personnel and the productivity that can be estimated from the proficiency, thereby improving the productivity of the organization.

[0072] Fig. 22 is a diagram for explaining an outline of a processing method according to an embodiment. As shown in Fig. 22, in this embodiment, in order to obtain proficiency elements, an operation data collection agent is used to collect operation logs of each person. In this case, in this embodiment, the operation target is identified at the granularity of a form unit, and various operation contents such as mouse operation, keyboard operation, printing, copying, pasting, etc. are precisely detected (Fig. 22 (A) and (1)).

[0073] In the embodiment, an algorithm is used to extract proficiency elements that have a relationship with proficiency, and proficiency elements are extracted for each personnel from multiple operation logs of multiple personnel ((2) in FIG. 22). As a result, in the embodiment, the operation content during screen input is quantified and summarized at a detailed granularity of screen element units, making it possible to extract proficiency elements that contribute to proficiency calculation using the operation characteristics specific to a skilled person, in addition to the required time. Therefore, in the embodiment, proficiency can be calculated including an individual's strong and weak screen elements, operating habits, etc.

[0074] In the embodiment, the proficiency level of each personnel is calculated from the proficiency elements selected according to the task based on the relationship between the proficiency elements and the proficiency level, so that the proficiency level of each personnel can be appropriately calculated.

[0075] Furthermore, in the embodiment, the proficiency elements of multiple personnel (e.g., within a department, etc.) are used to take statistics on the proficiency elements for each task, and a task difficulty coefficient for each task is calculated (FIG. 22 (B), (3) to (5)). In this way, by statistically analyzing the proficiency elements of multiple personnel, a task difficulty coefficient for each task is obtained. In the embodiment, this task difficulty coefficient is used to calculate the proficiency level for each personnel while correcting for bias between tasks, making it possible to calculate a proficiency level that better reflects the actual situation of the task.

[0076] [System configuration of the embodiment] Each component of the processing device 10 is a functional concept, and does not necessarily have to be physically configured as shown in the figure. In other words, the specific form of distribution and integration of the functions of the processing device 10 is not limited to that shown in the figure, and all or part of it can be functionally or physically distributed or integrated in any unit according to various loads, usage conditions, etc.

[0077] Further, all or any part of each process performed in the processing device 10 may be realized by a CPU, a GPU (Graphics Processing Unit), and a program analyzed and executed by the CPU and the GPU. Further, each process performed in the processing device 10, the learning device 20, and the signal processing device 100 may be realized as hardware using wired logic.

[0078] Furthermore, among the processes described in the embodiments, all or part of the processes described as being performed automatically can be performed manually. Alternatively, all or part of the processes described as being performed manually can be performed automatically by a known method. In addition, the process procedures, control procedures, specific names, and information including various data and parameters described above and shown in the drawings can be changed as appropriate unless otherwise specified.

[0079] [program] 23 is a diagram showing an example of a computer in which a program is executed to realize the processing device 10. The computer 1000 has, for example, a memory 1010 and a CPU 1020. The computer 1000 also has a hard disk drive interface 1030, a disk drive interface 1040, a serial port interface 1050, a video adapter 1060, and a network interface 1070. These components are connected by a bus 1080.

[0080] The memory 1010 includes a ROM 1011 and a RAM 1012. The ROM 1011 stores a boot program such as a BIOS (Basic Input Output System). The hard disk drive interface 1030 is connected to a hard disk drive 1090. The disk drive interface 1040 is connected to a disk drive 1100. A removable storage medium such as a magnetic disk or an optical disk is inserted into the disk drive 1100. The serial port interface 1050 is connected to a mouse 1110 and a keyboard 1120, for example. The video adapter 1060 is connected to a display 1130, for example.

[0081] The hard disk drive 1090 stores, for example, an OS (Operating System) 1091, an application program 1092, a program module 1093, and program data 1094. That is, a program that defines each process of the processing device 10 is implemented as a program module 1093 in which code executable by the computer 1000 is written. The program module 1093 is stored, for example, in the hard disk drive 1090. For example, a program module 1093 for executing the same process as the functional configuration of the processing device 10 is stored in the hard disk drive 1090. The hard disk drive 1090 may be replaced by an SSD (Solid State Drive).

[0082] Furthermore, setting data used in the processing of the above-described embodiment is stored as program data 1094, for example, in memory 1010 or hard disk drive 1090. Then, CPU 1020 reads out program module 1093 or program data 1094 stored in memory 1010 or hard disk drive 1090 into RAM 1012 as necessary, and executes the program.

[0083] Note that the program module 1093 and the program data 1094 are not limited to being stored in the hard disk drive 1090, but may be stored in, for example, a removable storage medium and read by the CPU 1020 via the disk drive 1100 or the like. Alternatively, the program module 1093 and the program data 1094 may be stored in another computer connected via a network (such as a local area network (LAN) or wide area network (WAN)). The program module 1093 and the program data 1094 may then be read by the CPU 1020 from the other computer via the network interface 1070.

[0084] Although the embodiment of the invention made by the present inventor has been described above, the present invention is not limited by the description and drawings that form a part of the disclosure of the present invention according to the present embodiment. In other words, other embodiments, examples, operation techniques, etc. made by those skilled in the art based on the present embodiment are all included in the scope of the present invention. [Explanation of symbols]

[0085] 10 Processing equipment 11 Communications Department 12 Storage section 13 Control section 121 Operation Log 122 Difficulty Data 131 Collection Department 132 Familiarity element extraction unit 133 Difficulty level acquisition part 134 Proficiency Calculation Unit

Claims

1. A processing method performed by a processing device, A collection step of collecting operation logs of each terminal used by a plurality of personnel performing business operations; An extraction step of extracting feature amounts for each personnel from the collected operation logs; A first calculation step of statistically analyzing proficiency factors of a plurality of personnel and calculating, for each task performed by each personnel, a task difficulty coefficient indicating the difficulty of the task; a second calculation step of calculating a proficiency level of the task for each employee from the feature amount extracted for each employee by using the task difficulty coefficient; Including, The first calculation step multiplies a first coefficient, the value of which increases as the average required time of the task increases and the value of which increases as the variation in the average required time increases depending on the personnel, the first coefficient being calculated based on a numerical value indicating the degree of variation in the required time of the task and a numerical value indicating the average required time, by a second coefficient, the value of which increases as the average error rate of the task increases and the value of which increases as the variation in the average error rate of the task increases depending on the personnel, the second coefficient being calculated based on a numerical value indicating the degree of variation in error rate data in the task and a numerical value indicating the average error rate, The second calculation step uses the task difficulty coefficient as a weighting value, and calculates the proficiency level based on the task that the personnel was performing when the operation log was collected by weighting the task so that the weighting increases the longer the time required for the task, and the weighting increases the higher the error rate, which is the ratio of the number of times the Backspace / Delete keys are used to the total number of keystrokes.

2. The feature amount is of a plurality of types, The feature amount includes the number of times an alt key on a keyboard of the terminal is used, the number of times a mouse connected to the terminal is used, and the number of times a paste operation is performed among operations on the terminal, which are acquired from the operation log; The processing method described in claim 1, characterized in that the second calculation step calculates the proficiency level for each personnel based on the relationship that the more the alt key is used, the higher the proficiency level, the less the mouse is used, and the more the proficiency level is increased, and the less the paste operation is performed, the higher the proficiency level.

3. The processing method according to claim 1 or 2, characterized in that the collection step acquires, as the operation log, an input form displayed on the screen of the terminal and a direct ancestor of the input form in a user interface tree.

4. The feature amount is of a plurality of types, The processing method according to any one of claims 1 to 3, characterized in that the second calculation step selects, from the features, a feature of a type that corresponds to the work environment of the personnel, and calculates the proficiency level for each personnel based on the feature of the selected type.

5. The processing method according to any one of claims 1 to 4, characterized in that the second calculation step calculates the proficiency level for each personnel using a model that has been trained to learn the relationship between features of a type corresponding to the work environment and the proficiency level for each work environment of the personnel.

6. The processing method according to any one of claims 1 to 5, characterized in that the weighting value is a value calculated at a predetermined timing for each work environment of multiple personnel based on multiple operation logs of the multiple personnel.

7. A collection unit that collects operation logs of each terminal used by a plurality of personnel performing work; an extraction unit that extracts a feature amount for each person from the operation log collected by the collection unit; a first calculation unit that statistically analyzes proficiency factors of a plurality of personnel and calculates, for each task performed by each personnel, a task difficulty coefficient that indicates the difficulty of the task; a second calculation unit that calculates a proficiency level of the task for each employee from the feature amount extracted for each employee by the extraction unit using the task difficulty coefficient; having the first calculation unit calculates the task difficulty coefficient by multiplying a first coefficient, the value of which increases as the average required time of the task increases and the value of which increases as the variation in the average required time increases depending on personnel, the first coefficient being calculated based on a numerical value indicating the degree of variation in the required time of the task and a numerical value indicating the average required time, and a second coefficient, the value of which increases as the average error rate of the task increases and the value of which increases as the variation in the average error rate of the task increases depending on personnel, the second coefficient being calculated based on a numerical value indicating the degree of variation in error rate data in the task and a numerical value indicating the average error rate; The second calculation unit uses the task difficulty coefficient as a weighting value, and calculates the proficiency level based on the task that the employee was performing when the operation log was collected by weighting the task so that the weighting increases the longer the time required for the task, and the weighting increases the higher the error rate, which is the ratio of the number of times the Backspace / Delete keys are used to the total number of keystrokes.

8. A collection step of collecting operation logs of each terminal used by a plurality of personnel performing business operations; Extracting features for each personnel from the collected operation logs; A first calculation step of statistically analyzing proficiency factors of a plurality of personnel and calculating a task difficulty coefficient indicating the difficulty of each task performed by each personnel; a second calculation step of calculating a proficiency level of the task for each employee from the feature amount extracted for each employee by using the task difficulty coefficient; Run the following on your computer: The first calculation step multiplies a first coefficient, the value of which increases as the average required time of the task increases and the value of which increases as the variation in the average required time increases depending on the personnel, the first coefficient being calculated based on a numerical value indicating the degree of variation in the required time of the task and a numerical value indicating the average required time, by a second coefficient, the value of which increases as the average error rate of the task increases and the value of which increases as the variation in the average error rate of the task increases depending on the personnel, the second coefficient being calculated based on a numerical value indicating the degree of variation in error rate data in the task and a numerical value indicating the average error rate, The second calculation step is a processing program that uses the task difficulty coefficient as a weighting value and calculates the proficiency level by weighting the task so that the longer the time required for the task is, and the greater the error rate, which is the ratio of the number of times the Backspace / Delete keys are used to the total number of keystrokes, based on the task that the employee was performing when the operation log was collected, the greater the weighting.

Images