Estimation device

Abstract

An estimation device according to an embodiment of the present invention includes: an operation screen log extraction unit that receives an operation log related to a plurality of operation components in an operation screen from a first operation terminal among a plurality of operation terminals; a component positional information extraction unit that extracts, on the basis of the operation log, positional information of the plurality of operation components in the operation screen; a space analysis unit that computes the arrangement density of the operation components on the basis of the positional information; and a similarity evaluation parameter calculation unit that computes, on the basis of the arrangement density, a similarity evaluation parameter for evaluating the similarity degree of operation of work in the plurality of operation terminals.

Description

estimation device

[0001] The embodiment relates to an estimation device.

[0002] In order to achieve effective business improvement in companies, it is important to accurately understand the target business operations. The personnel in charge of business operations perform multiple tasks every day using information terminals such as PCs or tablet devices. The tasks performed via information terminals consist of multiple tasks.

[0003] The introduction of robotics process automation (RPA) is being promoted to improve the efficiency of work involving the operation of information terminals. RPA is provided by configuring the operation of information terminals as a combination of multiple commands called scenarios. Based on the scenario (hereinafter referred to as an RPA scenario), an agent operates the information terminal on behalf of a person. This realizes the automation of information terminal operation.

[0004] RPA is introduced into actual business operations after selecting the business operations for which it will be introduced, creating RPA scenarios, and testing the RPA scenarios. Introducing RPA requires selecting the business operations, and in addition to knowledge of the RPA tool, understanding the business operations is also important. Creating an RPA scenario requires knowledge of the business system as well as knowledge of the RPA tool. Introducing RPA requires experts in RPA, business operations, and business systems, as well as a corresponding amount of work.

[0005] In order to simplify and speed up the introduction of RPA, it is being considered to horizontally deploy RPA scenarios for similar tasks across multiple branches of a company or across multiple companies of the same type.

[0006] International Publication No. 2022 / 176134

[0007] In Patent Document 1, the operation sequence of an operation log is used to determine the similarity of tasks in a business. The operation sequence indicates various tasks that have been performed.

[0008] In Patent Document 1, the similarity between two operation sequences acquired from different operation logs is determined based on the number of types of operations common to the two operation sequences.

[0009] Even in a series of operations that are considered similar in business between automatic operation and manual operation of an information terminal, small differences in procedures or differences in waiting times for operations occur.

[0010] Therefore, when each operation log is divided into multiple operation sequences for each section in which the operations are similar, and the similarity of the operation sequences is grasped between multiple operation logs, it may be preferable to divide the operation sequence into sections within a range of n operations before and after the division point of the operation sequence, and to calculate the similarity for each of these sections.

[0011] However, the number of operations (the value of n) is a value specified by the user of the business system to which RPA is applied, and therefore the optimal value of the parameter "n" must be manually found by the user through trial and error.

[0012] Therefore, the parameter "n" must be set to a fixed value or a rough average value based on experience, which may hinder accurate calculation of the similarity between operation sequences.

[0013] The present invention has been made in view of the above, and aims to provide an improved accuracy in evaluating similar tasks.

[0014] The estimation device of the embodiment includes an operation screen log extraction unit that receives an operation log related to a plurality of operation components in an operation screen from a first operation terminal among a plurality of operation terminals; a component position information extraction unit that extracts position information of the plurality of operation components in the operation screen based on the operation log; a spatial analysis unit that calculates a placement density of the operation components based on the position information; and a similarity evaluation parameter calculation unit that calculates a similarity evaluation parameter for evaluating the similarity of business operations on the plurality of operation terminals based on the placement density.

[0015] According to the embodiment, it is possible to improve the accuracy of evaluation of similar tasks.

[0016] FIG. 1 is a diagram showing an example of the configuration of a system including an estimation device of an embodiment. FIG. 2 is a diagram showing an example of the configuration of an operation screen of a business system. FIG. 3 is a block diagram showing an example of the configuration of an estimation device of an embodiment. FIG. 4 is a diagram showing an example of an operation log of the business system. FIG. 5 is a diagram showing an example of processing related to spatial analysis by the estimation device of an embodiment. FIG. 6 is a flowchart showing an example of a processing method of the estimation device of an embodiment.

[0017] Hereinafter, embodiments of the present invention will be described with reference to the drawings. Hereinafter, elements that are identical or similar to elements already described will be assigned the same or similar reference numerals, and duplicate descriptions will generally be omitted. For example, when there are multiple identical or similar elements, a common reference numeral may be used to describe each element without distinguishing between them, or a subnumber may be used in addition to the common reference numeral to describe each element distinctly.

[0018] (Embodiment) An estimation device according to an embodiment will be described with reference to Figs.

[0019] (a) Configuration FIG. 1 is a block diagram showing an example of the overall configuration of a system including an estimation device according to an embodiment.

[0020] 1, a business system 10 is provided as a system that can be commonly used by multiple users. The business system 10 communicates with multiple operation terminals 20 (20A, 20B). The business system 10 provides the operation terminals 20 with applications (software) that allow the users to perform various tasks. For example, the business system 10 is a server.

[0021] Each of the plurality of operation terminals (also called information terminals) 20 executes various tasks related to business by using the business system 10. Each operation terminal 20 is operated by each of the plurality of users.

[0022] The operation terminal 20 is configured to allow operation of an operation screen (also called a work screen) 100, which is a user interface for executing predetermined tasks and operations. The operation screen 100 is provided by an application of the business system 10.

[0023] FIG. 2 is a schematic diagram showing an example of the operation screen 100. As shown in FIG.

[0024] 2, one or more operation screens 100 are displayed on the display of the operation terminal 20. The operation screen 100 is a window (or a dialog) constructed by software. The operation screen 100 includes a plurality of operation components 101 (101A, 101B, 101C) arranged in the window.

[0025] Each of the multiple operation components 101 includes input components such as text boxes 101A and 101B for inputting characters, execution components such as button 101C for instructing the execution of a process, and selection components such as a list box (not shown). Text such as a membership code or member name is entered into text boxes 101A and 101B. Pressing (clicking) button 101C executes a predetermined process such as a search.

[0026] Each of the operation components 101 is arranged at a predetermined coordinate within the operation screen 100. For example, one operation screen 100 is used to perform a search-related operation (process) and a registration-related operation.

[0027] In this embodiment, one operation terminal 20A of the multiple operation terminals 20 functions as a reference for business comparison (hereinafter referred to as a comparison reference). One or more operation terminals 20B of the multiple operation terminals 20 function as targets for business comparison (hereinafter referred to as comparison targets). Each of the multiple operation terminals 20 generates an operation log LG (LG1, LG2) according to the work performed. The operation log LG includes one or more pieces of information such as operation time, operation screen, operation component, operation content, operation location, and operation position (position of the operation component).

[0028] For example, the operation terminal 20A serving as the reference for business comparison may be automatically operated based on RPA or manually operated by a user, while the operation terminal 20B serving as the target for business comparison may be operated by RPA or manually.

[0029] The operation terminal 20 is a smartphone, a tablet terminal, a personal computer, or the like.

[0030] The processing device 30 communicates with a plurality of operation terminals 20. The processing device 30 receives a plurality of operation logs LG from each of the plurality of operation terminals 20. The processing device 30 receives parameters PM from an estimation device 40 of this embodiment, which will be described later. The processing device 30 has a function of executing various processes based on an operation log LG1 from a comparison reference operation terminal 20A, an operation log LG2 from a comparison target operation terminal 20B, and the parameters PM from the estimation device 40.

[0031] The processing device 30 evaluates the similarity between operations of the same type of business performed on different operation terminals 20 by reflecting the parameters PM from the estimation device 40 in the comparison results of the two operation logs LG1 and LG2. The processing device 30 generates business similarity evaluation data X0 by performing various processes on the supplied operation logs LG and parameters PM. The processing device 30 outputs the business similarity evaluation data X0. Using the business similarity evaluation data X0, an operation log similar to the operation log (business) performed on the comparison reference operation terminal 20A is searched for. Based on the search results for similar operation logs, an RPA scenario and an RPA tool that can be horizontally deployed between stores or companies are provided by the administrator or provider of the business system 10.

[0032] The estimation device (also referred to as a calculation device or a processing device) 40 of this embodiment communicates with the comparison reference operation terminal 20A. The estimation device 40 is configured to estimate operations of similar tasks. The estimation device 40 estimates parameters PM for evaluating the similarity of similar tasks based on the operation log LG1 of the comparison reference operation terminal 20A. The estimation device 40 calculates parameters PM for appropriate comparison and evaluation of the similarity of tasks (operations) based on the estimation results. The estimation device 40 provides the calculated parameters PM to the processing device 30. Hereinafter, the parameters PM estimated and calculated by the estimation device 40 are referred to as similarity evaluation parameters PM. The similarity evaluation parameters PM are numerical values ​​used for weighting to evaluate the co-occurrence or similarity between operations.

[0033] The estimator 40 includes hardware components such as an interface 41 , a processor 42 , and a memory 43 .

[0034] The interface 41 has a function for communicating with the operation terminal 20A of the comparison reference. The interface 41 receives the operation log LG1 of the comparison reference from the operation terminal 20A. The interface 41 sends the similarity evaluation parameter PM generated by estimation to the processing device 30.

[0035] The processor 42 has functions for executing various processes executed within the estimation device 40. The processor 42 calculates and generates a similarity evaluation parameter PM based on estimation using the comparison standard operation log LG1. The processor 42 includes multiple function blocks for executing various processes.

[0036] The memory 43 has a function of storing the operation log LG1, various data used by the processor 42, programs (software), etc. The memory 43 includes a DRAM, an HDD, an SSD, etc. The memory 43 stores, for example, a program 90 for generating the similarity evaluation parameter PM. The program 90 is executed by the processor 42 when generating the similarity evaluation parameter PM. The program 90 includes multiple program codes that indicate predetermined procedures for estimating and generating the similarity evaluation parameter PM.

[0037] The estimation device 40 may have access to a database (not shown) external to the device 40 or internal to the device 40 .

[0038] For example, the estimation device 40 is realized by a personal computer.

[0039] <Estimation Device> FIG. 3 is a block diagram showing an example of the configuration of an estimation device according to this embodiment.

[0040] 3, the estimation device 40 receives a comparison reference operation log LG1 from the comparison reference operation terminal 20A. The estimation device 40 estimates a similarity evaluation parameter PM from the operation log LG1.

[0041] The estimation device 40 includes an operation screen log extraction unit 401, a part position information extraction unit 402, a spatial analysis unit 403, a similarity evaluation parameter calculation unit 404, and a control unit 409. For example, each of the units (functional blocks) 401, 402, 403, 404, and 409 of the estimation device 40 is configured by software or firmware.

[0042] The operation screen log extraction unit 401 has a function of receiving an operation log (comparison reference operation log) LG1 from the comparison reference operation terminal 20A. The operation screen log extraction unit 401 has a function of sorting the input operation log LG1 for each operation screen 100. The operation screen log extraction unit 401 has a function of extracting information of the operation log LG1 (also called operation log information) for each operation screen 100.

[0043] The operation screen log extraction unit 401 can extract, from the operation log LG1, operation log information such as the used operation components 101, the operation contents of the operation components 101, and the positions of the used operation components 101. The operation screen log extraction unit 401 has a function of supplying the operation log information extracted from the operation log LG1 to the component position information extraction unit 402.

[0044] FIG. 4 is a diagram showing an example of the comparison reference operation log LG1.

[0045] 4, the operation log LG1 indicates the operation components 101 used, the details of the operations performed, and the position of the operation components 101 used within the operation screen 100. For example, the position of the operation components 101 is indicated by the Y coordinate of the operation screen 100 (the Y coordinate within the window). A set of the operation components 101 used and the corresponding operations and positions constitutes one entry. Multiple entries are written in the operation log LG1 in chronological order according to the order in which they were performed.

[0046] Based on the contents of the illustrated operation log LG1, the execution of multiple tasks such as the following can be detected.

[0047] When a certain task is performed, the operation element 101A in FIG. 3 is used to input a membership code. The membership code "1234" is input into the operation element 101A by operating the operation element 101A. The operation element 101A is located at the "3" position on the operation screen 100.

[0048] After inputting the membership code using operation element 101A, operation element 101B is used to input the membership name. The membership name "○×" is input into operation element 101B by operating operation element 101B. Operation element 101B is located at the "4" position on operation screen 100.

[0049] After the member code and member name are entered in sequence, operation element 101C is used to search for information about the member. The search button, which is operation element 101C, is pressed by operating operation element 101C. Operation element 101C is located at position "6" on operation screen 100.

[0050] Thereafter, a plurality of tasks are performed by operating a plurality of other operation components.

[0051] In this way, a plurality of tasks and operations for business are executed sequentially using the various operation components 101 in the operation screen 100.

[0052] Regarding the results of such work and operations, the operation screen log extraction unit 401 extracts various operation log information for each operation screen 100, such as information on the operation components 101 used in the operation log LG1, information on the content of the operation performed, and location information of the operation components 101 used.

[0053] The part position information extraction unit 402 has a function of receiving information extracted from the operation log LG1 from the operation screen log extraction unit 401. The part position information extraction unit 402 has a function of extracting position information of the operation components 101 included in the operation log information of the received operation log LG1. The part position information extraction unit 402 has a function of supplying the position information of the operation components 101 to the spatial analysis unit 403.

[0054] A location information database DB1 is constructed from a collection of the location information of the extracted operation components 101.

[0055] The spatial analysis unit 403 has a function of receiving a set of position information of the operational components 101 from the component position information extraction unit 402. The spatial analysis unit 403 has a function of analyzing the received set of position information of the operational components 101. The spatial analysis unit 403 has a function of estimating the placement density of the operational components 101 based on the analysis results. The spatial analysis unit 403 has a function of defining a section having a density equal to or greater than a specified density threshold th as a homogeneous task section, and calculating the number of components placed in the homogeneous task section. In a certain task section, a series of tasks for a certain task are performed. For example, the spatial analysis unit 403 calculates the placement density of the operational components 101 in the task section using kernel density estimation.

[0056] Fig. 5 is a schematic diagram for explaining spatial analysis by the spatial analysis unit 403. Fig. 5 shows an example of spatial analysis of one operation screen 100 illustrated in Fig. 2 based on the operation log LG1 illustrated in Fig. 4.

[0057] As shown in Fig. 5, the positions of the operation objects 101 in the coordinate space for spatial analysis are defined based on the position information of the operation objects 101. The vertical axis of the graph in Fig. 5 corresponds to the Y direction (vertical direction) of the operation screen 100. The horizontal axis of the graph in Fig. 5 corresponds to the arrangement density of the operation objects 101.

[0058] Generally, when a plurality of operation components 101 are arranged vertically on the operation screen 100, the operation components 101 are operated in a vertical order from the top to the bottom of the operation screen 100. Therefore, the arrangement of the operation components 101 can be associated with the order in which operations (tasks) are performed.

[0059] The placement density 500 is calculated based on the operational object 101 used and the position of the operational object 101 .

[0060] The spatial analysis unit 403 recognizes a range (section) where the placement density 500 is equal to or greater than the density threshold th as a task section AR (AR1, AR2). The task section AR is also called an operation queue section or a work section.

[0061] The spatial analysis unit 403 recognizes a section sandwiched between two positions where the placement density 500 is equal to or less than the density threshold th as a single task section AR1 (or task section AR2) where the same type of work and operations are performed. This identifies a task section AR that includes a sequence of operations for a certain task. For example, when the operation screen 100 in FIG. 2 is spatially analyzed, the operation screen 100 is recognized as including a task section AR1 related to "search" and a task section AR2 related to "registration."

[0062] The spatial analysis unit 403 calculates the number of operational components 101 (number of components) "p" within the same type of task interval AR. In the example of Fig. 5, the number of components "p" in task interval AR1 is "3". The number of components "p" in task interval AR2 is "4".

[0063] The spatial analysis unit 403 has a function of supplying the calculated number of parts (p) in the same type of business section AR to the similarity evaluation parameter calculation unit 404 .

[0064] The similarity evaluation parameter calculation unit 404 has a function of receiving the calculated number (p) of operation components 101 in the same type of task section AR from the spatial analysis unit 403. The similarity evaluation parameter calculation unit 404 has a function of calculating a similarity evaluation parameter PM for the operation screen 100 from the received number of operation components 101 in the same type of task section AR. The similarity evaluation parameter calculation unit 404 has a function of outputting a similarity evaluation parameter PM having a value of "n". For example, the similarity evaluation parameter PM corresponds to the number (n) of operations before and after determining the task section AR. For example, the value of "n" indicates the length of the operation sequence section used when calculating the similarity of operations.

[0065] If the number of parts in a certain operation section AR is denoted by "p," then "n(i)," which indicates the value of the similarity evaluation parameter PM in the operation section AR with a certain identification number (i), is obtained from "p / 2." In the example of FIG. 5, the value n(1) of the similarity evaluation parameter PM for operation section AR1 is "3 / 2." The value n(2) of the similarity evaluation parameter PM for operation section AR2 is "2."

[0066] A similarity evaluation parameter database DB2 is constructed from the plurality of similarity evaluation parameters n(i) calculated by the similarity evaluation parameter calculation unit 404.

[0067] The similarity evaluation parameter calculation unit 404 calculates a similarity evaluation parameter "n" recommended for one operation screen 100 from the average value (Σn(i)) of the respective parameter values ​​n(1), n(2), ... of a plurality of similarity evaluation parameters obtained from a plurality of operation sections AR. For example, the value "n" of the similarity evaluation parameter PM based on the average value of a plurality of similarity evaluation parameters is a value defined by the INT function.

[0068] In this way, the similarity evaluation parameter calculation unit 404 can calculate and generate the value "n" of the similarity evaluation parameter PM for a certain operation screen 100 based on the comparison reference operation log LG1 for that operation screen 100.

[0069] The control unit 409 has a function of controlling each of the units 401, 402, 403, and 404. The control unit 409 is configured to cause each of the units 401, 402, 403, and 404 to execute processing in a predetermined procedure. The control unit 409 sets a density threshold th based on, for example, input from a user. The control unit 409 supplies the set density threshold th to the spatial analysis unit 403.

[0070] (b) Operation Example A processing method of the estimation device 40 of the embodiment will be described with reference to Fig. 6. Fig. 6 is a flowchart showing the processing method of the estimation device 40 of the embodiment.

[0071] <Step S1> A comparison reference operation log LG1 is input from the comparison reference operation terminal 20A to the estimating device 40. In the estimating device 40, the operation screen log extracting unit 401 acquires the operation log LG1.

[0072] <Step S2> The operation log LG1 is sorted into units of each operation screen 100 by the operation screen log extraction unit 401. Operation log information including various information such as the operation components 101 used and the positions of the operation components 101 is extracted from the operation log LG1 of each operation screen 100 by the operation screen log extraction unit 401.

[0073] <Step S3> With respect to the operation log LG1 of each sorted operation screen 100, various processes are executed for each operation screen 100 under the control of the control unit 409.

[0074] <Step S4> The position information of each operation component 101 in the operation screen 100 is extracted from the operation log information by the component position information extraction unit 402. For example, the acquired position information is stored in the position information database DB1.

[0075] <Step S5> The control unit 409 determines whether or not processing has been completed for all operation screens 100. If processing has not been completed for all operation screens 100 (NO in S5 of FIG. 6), the processes of steps S3 and S4 are executed.

[0076] <Step S6> When processing for all operation screens 100 is completed (YES in S5 of FIG. 6 ), a set of position information of the operation components 101 is read from the position information database DB1 by the spatial analysis unit 403. The spatial analysis unit 403 calculates the arrangement density of the operation components 101 related to the read position information of the operation components 101 using, for example, kernel density estimation.

[0077] <Step S7> Based on the calculated placement density of the operation components 101, the spatial analysis unit 403 identifies one or more business sections AR having a placement density equal to or greater than the density threshold th, which is a set value.

[0078] <Step S8> Various processes are executed for the one or more identified service sections AR under the control of the control unit 409 in units of service sections AR.

[0079] <Step S9> In a certain operation section AR, an individual similarity evaluation parameter PM is calculated from the number of parts in the operation section AR by the similarity evaluation parameter calculation unit 404. When the number of parts is indicated by "p(i)" and the parameter value of the similarity evaluation parameter PM for the operation section AR is indicated by "n(i)", the parameter value of the similarity evaluation parameter PM for the operation section AR is expressed by the following (Equation 1).

[0080] n(i)=p(i) / 2 (Equation 1) Here, "i" is an integer of 1 or more.

[0081] The similarity evaluation parameter n(i) calculated for each service section AR is stored in the similarity evaluation parameter database DB2 by the similarity evaluation parameter calculation unit 404.

[0082] <Step S10> Whether or not processing for all service sections AR has been completed is determined by the control unit 409. If processing for all service sections AR has not been completed (NO in S10 in FIG. 6), the processes of steps S8 and S9 are executed.

[0083] <Step S11> When processing of all service sections AR is completed (YES in S10 of FIG. 6), the similarity evaluation parameter calculation unit 404 calculates the average (n=Σn(i)) of the similarity evaluation parameters of the service sections AR.

[0084] <Step S12> The calculated similarity evaluation parameter (n) is output as the similarity evaluation parameter PM for a certain operation screen 100 (or all operation screens 100) by the similarity evaluation parameter calculation unit 404. The output similarity evaluation parameter PM having the value of “n” is supplied to, for example, the processing device 30.

[0085] Through the above processing, the estimation device 40 of this embodiment can obtain the similarity evaluation parameter (n) based on the operation log LG1 from the comparison reference operation terminal 20A.

[0086] Thereafter, a similarity comparison analysis of the multiple operation logs LG1 and LG2 is performed by the processing device 30. The results of the similarity comparison analysis are obtained based on the number of operations corresponding to the similarity evaluation parameter PM having the value "n" obtained by the estimation device 40 of this embodiment. This generates task similarity evaluation data X0 with high accuracy. Based on the task similarity evaluation data X0, it is possible to search for operation logs similar to the operation log (task) of an RPA scenario executed by, for example, an RPA tool.

[0087] (c) Summary In this embodiment, the business system 10 includes a tendency (hypothesis, basic design) that "when the operation target of the operation terminal is an operation screen provided by the business system, operation components related to the same type of business are placed in close positions on the operation screen."

[0088] Using this tendency, the estimation device 40 of this embodiment estimates operation components 101 for the same type of task from the arrangement positions of the operated operation components 101 included in the operation log LG1. The estimation device 40 calculates the value (n) of the similarity evaluation parameter PM from the size (arrangement density in the space) of the space (task section AR) where the operation components 101 are arranged.

[0089] Furthermore, the business system 10 includes the tendency that "when the operation target of an operation terminal is a plurality of screens, the configuration of the screens provided by one business system is similar for all the screens."

[0090] Using this tendency, the estimation device 40 of this embodiment can calculate a similarity evaluation parameter PM collectively for the plurality of operation components 101 from the plurality of operation components 101 of each of the plurality of operation screens 100. For example, for the plurality of operation screens 100, the similarity evaluation parameter calculated collectively for the plurality of operation screens 100 can obtain a value that is approximately the same as the similarity evaluation parameter calculated for each operation screen 100.

[0091] In this embodiment, the estimation device 40 acquires the position coordinates of each operation component 101 in the operation screen 100. As described above, since the screen configurations of the operation screens 100 in a certain business system 10 are similar, there is no need to perform processing for acquiring the position coordinates for each operation screen 100.

[0092] The estimation device 40 estimates the placement density of the operation objects 101 from the acquired position coordinates of all the operation objects 101 .

[0093] Based on the tendency that operation components 101 used for the same type of work are placed in close positions, the estimation device 40 recognizes that the range in which the placement density 500 of operation components 101 exceeds a certain threshold th is a similar type of work section (operation sequence section) AR.

[0094] The estimation device 40 calculates a similarity evaluation parameter PM from the number of operational components 101 in each task section AR.

[0095] When multiple business sections AR exist within one operation screen 100, the estimation device 40 provides the average value of multiple values ​​obtained from one operation screen 100 as the value “n” of the similarity evaluation parameter PM for that one operation screen 100.

[0096] This allows the estimation device 40 of this embodiment to improve the accuracy of evaluation of similar tasks. Furthermore, the estimation device 40 of this embodiment can reduce the time and burden required for tuning parameters for evaluation of similar tasks.

[0097] (4) Others: The present invention is not limited to the above-described embodiments, and various modifications can be made in the implementation stage without departing from the spirit of the invention. Furthermore, the embodiments may be implemented in appropriate combinations, in which case the combined effects can be obtained. Furthermore, the above-described embodiments include various inventions, and various inventions can be extracted by combining selected elements from the disclosed elements. For example, if the problem can be solved and the desired effect can be obtained even if some elements are deleted from all elements shown in the embodiments, the configuration from which these elements are deleted can be extracted as an invention.

[0098] DESCRIPTION OF SYMBOLS 10... Business system 100... Operation screen 101, 101A, 101B, 101C... Operation components 20, 20A, 20B... Operation terminal 30... Processing device 40... Estimation device 401... Operation screen log extraction unit 402... Component position information extraction unit 403... Spatial analysis unit 404... Similarity evaluation parameter calculation unit 500... Placement density LG, LG1, LG2... Operation log PM... Similarity evaluation parameter th... Density threshold

Claims

1. An estimation device comprising: an operation screen log extraction unit that receives an operation log regarding a plurality of operation components within an operation screen from a first operation terminal among a plurality of operation terminals; a component position information extraction unit that extracts position information of the plurality of operation components within the operation screen based on the operation log; a space analysis unit that calculates an arrangement density of the operation components based on the position information; and a similarity evaluation parameter calculation unit that calculates a similarity evaluation parameter for evaluating a similarity of operations of a business in the plurality of operation terminals based on the arrangement density.

2. The estimation device according to claim 1, wherein the space analysis unit specifies one or more business sections within the operation screen based on the arrangement density.

3. The estimation device according to claim 2, wherein the similarity evaluation parameter calculation unit obtains the similarity evaluation parameter from an average value of parameter values for each of the business sections.

4. The estimation device according to claim 1, wherein the space analysis unit calculates the arrangement density using kernel density estimation.

Images