Information processing device and control structure estimation method for automated operation scenarios

Abstract

An information processing device according to one embodiment of this invention is equipped with a communication unit, a data acquisition unit, an identical operation determination unit, an iterative processing estimation unit, a branching processing estimation unit, and a structure generation unit. The communication unit communicates with a terminal device on which an application that can be controlled along automated operation scenarios has been installed. The data acquisition unit acquires, as an operation log, processing content of the application according to the automated operation scenarios. The identical operation determination unit determines identical operations included in the acquired operation log. The iterative processing estimation unit estimates iterative processing included in the acquired operation log. The branching processing estimation unit estimates branching processing included in the acquired operation log. The structure generation unit generates control structure information indicating a control structure for the automated operation scenarios on the basis of the estimated identical operations, the estimated iterative processing, and the estimated branching processing.

Description

Information processing device and method for estimating control structure of automatic operation scenario

[0001] One aspect of the present invention relates to an information processing device and a method for estimating a control structure of an automatic operation scenario.

[0002] Robotic Process Automation (RPA) is a software tool that operates computers (terminal devices) in place of humans, and is gaining attention as a method for improving business processes amid the current labor shortage. Most tools of this type function according to a pre-given scenario (automated operation scenario). Technology that supports the creation of automated operation scenarios is also known. There are many automated operation tools, each with its own strengths and weaknesses. To upgrade a tool, it is essential for humans to understand the automated operation scenarios created using the tool's unique expressions.

[0003] Therefore, there are known techniques to help understand automatic operation scenarios (see, for example, Non-Patent Document 1). However, even with this type of technique, it is difficult to understand the control structure (repetition, branching, etc.) of the entire scenario.

[0004] There is also a known technology for estimating the structure of a business process model from the history information left in a business system (see, for example, Patent Document 1). However, this technology cannot be applied to cases where the history information recorded in the business system is unavailable.

[0005] Structured programming comes to mind as a similar technical field, but since structured programming infers functional processing within a specific operation, it does not structurally analyze the relationships between multiple operations, i.e., the (black-boxed) functions contained within an operation.

[0006] Japanese Patent Application Laid-Open No. 2006-338274

[0007] Ryo Uchida, Misa Fukai, Taisuke Wakasugi, Kimio Tsuchikawa, Haruo Oishi: Proposal of Annotation Method for RPA Scenario. 361-364, APNOMS2023.

[0008] When switching between automation tools or horizontally deploying automation scenarios, it is essential that the person in charge understands the scenarios. However, the way automation scenarios are written varies greatly depending on the creator, and knowledge of the tool's creation and expression methods is also required. There is a strong demand for improving the readability of automation scenarios.

[0009] The present invention has been made in light of the above circumstances, and aims to provide a technique that can improve the readability of automatic operation scenarios.

[0010] An information processing device according to one aspect of the present invention includes a communication unit, a data acquisition unit, an identical operation determination unit, a repetitive processing estimation unit, a branch processing estimation unit, and a structure generation unit. The communication unit communicates with a terminal device on which an application controllable according to an automatic operation scenario is installed. The data acquisition unit acquires processing details of the application according to the automatic operation scenario as an operation log. The identical operation determination unit determines identical operations included in the acquired operation log. The repetitive processing estimation unit estimates repetitive processing included in the acquired operation log. The branch processing estimation unit estimates branch processing included in the acquired operation log. The structure generation unit generates control structure information indicating a control structure of the automatic operation scenario based on the estimated identical operations, the estimated repetitive processing, and the estimated branch processing.

[0011] According to one aspect of the present invention, the readability of an automatic operation scenario can be improved.

[0012] FIG. 1 is a functional block diagram showing an example of an information processing device according to an embodiment of the present invention. FIG. 2 is a flowchart showing an example of a processing procedure of the information processing device according to the embodiment. FIG. 3 is a diagram showing an example of an acquired operation log. FIG. 4 is a diagram showing an example of a formatted operation log. FIG. 5 is a diagram for explaining a procedure for extracting a repetitive process. FIG. 6 is a diagram for explaining a procedure for extracting a repetitive process / branching process. FIG. 7 is a diagram for explaining a procedure for extracting a repetitive process / branching process. FIG. 8 is a diagram for explaining flow correction. FIG. 9 is a diagram for explaining a case where a flow that does not exist in the actual operation log is included. FIG. 10 is a diagram for explaining processing of the information processing device 1 according to the embodiment.

[0013] Next, an embodiment of the present invention will be disclosed with reference to the drawings. The following explanation will be given of the case where RPA is applied as an automated operation tool. It is known that a scenario control structure can generally be described using repetition and branching (library). Therefore, in the embodiment, a detailed explanation will be given of generating a scenario control structure using the following steps (1) to (5): (1) Execute a scenario multiple times and obtain operation logs. (2) Format each operation log. (3) Identify identical operations for the operation logs. (4) Overlap identical operations for each operation log, extract repetitive processing and branching processing, and construct an operation flow. (5) Modify the flow from the perspective of visibility and maintainability, and estimate the structure of the scenario.

[0014] 1 is a block diagram showing an example of an information processing device according to an embodiment of the present invention. The information processing device 1 is a computer including a processor 10 such as a CPU (Central Processing Unit), a storage 20, and a memory 30. The information processing device 1 further includes a communication unit 40 and an interface unit 50 connected to an input unit 51 and a display unit 52. The storage 20 stores a program 20a for causing the computer to function as the information processing device 1. The information processing device 1 may be included in a terminal device 2.

[0015] The communication unit 40 communicates with a terminal device 2 (computer) that is a target of RPA operation, for example, via a LAN (Local Area Network). A business application, for example, that can be controlled according to an automatic operation scenario (RPA scenario), is installed on the terminal device 2. The information processing device 1 transmits an automatic operation scenario 301 stored in the memory 30, for example, to the terminal device 2, and causes the RPA to execute at least a part of the operation of the business application.

[0016] The processor 10 includes a data acquisition unit 101, a data formatting unit 102, an identical operation determination unit 103, a branch structure estimation unit 104, a repetitive structure estimation unit 105, and a structure generation unit 106 as processing functions according to the embodiment.

[0017] The data acquisition unit 101 acquires, as an operation log, the processing content of the business application of the terminal device 2 operated by the automatic operation scenario 301. The acquired operation log is stored in memory (operation log 302).

[0018] The data reforming unit 102 performs preprocessing, which is generally known in the data processing field, on the acquired operation log 302, such as data reforming, deleting unnecessary data, and removing blank data.

[0019] The identical operation determination unit 103 determines identical operations included in the acquired operation log 302. The branch structure estimation unit 104 estimates branch processes included in the acquired operation log 302. The repetition structure estimation unit 105 estimates repetition processes included in the acquired operation log 302.

[0020] The structure generation unit 106 analyzes the control structure of the automatic operation scenario 301 based on the identical operation estimated by the identical operation determination unit 103, the repetition process estimated by the repetition structure estimation unit 105, and the branch process estimated by the branch structure estimation unit 104. Then, the structure generation unit 106 generates control structure information indicating the control structure of the automatic operation scenario 301. The generated control structure information is visualized as, for example, a flow diagram and displayed on the display unit 52.

[0021] 2 is a flowchart showing an example of a processing procedure of the information processing device according to the embodiment. <Step S1: Acquisition of operation log> In step S1 of Fig. 2, the information processing device 1 causes the RPA to execute an operation based on the automatic operation scenario 301 multiple times. Then, in each trial, an operation log is acquired at the timing when an operation is performed on the terminal device 2.

[0022] 3 is a diagram showing an example of the acquired operation log. The operation log is information indicating the state and operation content of the terminal device 2 when the operation was performed, and includes, for example, attribute values ​​such as operation time, application information, window information, operation location, and operation position, as well as a captured image of the operation screen. The operation screen is a GUI (Graphical User Interface) screen displayed on the front end of the terminal device 2.

[0023] Generally, an automated operation scenario specifies a series of processing procedures for a specific business, and therefore the processing has a "start" and an "end." The operation log acquired by executing an automated operation scenario records the period between this "start" and "end." In other words, the operation log includes a "start" and an "end." In addition, the automated operation scenario is executed multiple times under different conditions so that the execution procedure includes branching.

[0024] <Step S2: Data Shaping> In step S2, the information processing device 1 sorts each operation (row: each entry in the table) by time and arranges them in chronological order by referring to the timestamps (operation times) of the acquired operation log 302. Note that some information necessary for identifying the same operation may be extracted from the operation log 302.

[0025] <Step S3: Determining Identical Operations> In step S3, the information processing device 1 identifies identical operations for each of the operations recorded in the acquired operation log 302. Then, the information processing device 1 assigns an individual name (operation name) to each operation based on the formatted information.

[0026] Fig. 4 shows an example of a formatted operation log. Fig. 4 shows an example in which operation names (a, b, c, ...) are assigned based on the "window information" and "operation location" information. Here, application information, captured images, and operation locations are dropped.

[0027] Note that, for example, a captured image may also be used to identify the same operation. In the case of operation logs, different logs may be obtained even though the same operation content is recorded due to, for example, differences in the browser used in the front end. In such cases, using a captured image can improve the accuracy of determining the identity of operations that indicate different logs.

[0028] <Step S4: Estimation of Repeated Processing / Branched Processing> In step S4, the information processing device 1 overlaps (degenerates) the estimated identical operations and estimates repeated processing / branched processing. In detail, step S4 includes, for example, the following steps.

[0029] (Step 1) Select and read one flow from the operation log. This will be the initial flow. (Step 2) Select one operation log to compare with the initial flow. This may be the same log as the initial flow. In this case, the extracted process is the repeat / branch process within the initial flow. (Step 3) Extract one operation recorded in the operation log in chronological order and use it as the operation to be compared. (Step 4) Compare the target operation with the operations in the initial flow in chronological order, and if an identical operation does not exist, add a link from the previous operation to the target operation. (Step 5) Compare the target operation with the operations in the initial flow in chronological order, and if an identical operation exists, execute (Step 5-1) or (Step 5-2). (Step 5-1) Determine whether there is a link representing the transition between the previous operation and the target operation, and if there is, do nothing and return to (Step 2). (Step 5-2) If there is no link, add a repeat flow from the previous operation to the target operation. (Step 6) When all operations included in the operation log have been extracted, the process estimation is considered complete. At this time, if the estimation is performed on the same log as the initial flow, the nodes in the initial flow that were not used for comparison are deleted, and the initial flow is updated.

[0030] Note that loop processing and branching processing can be extracted simultaneously or individually. Also, if a branch exists, there is a possibility that there will be differences in the system state and screen contents at the time of operation for each operation log. Therefore, when branching processing is extracted, the differences are extracted from the operation logs before and after the branching occurs and acquired together as information on possible branching conditions.

[0031] 5 is a diagram illustrating a procedure for extracting a repetitive process. For example, assume that a time series of processes indicated by a→b→c→a→b→c→a→b→c→a→b is extracted from the operation log.

[0032] If no identical operations exist, a processing sequence such as that shown in Figure 5(A) is extracted in (Procedure 4). If identical operations exist but no links exist, a processing sequence such as that shown in Figure 5(B) is extracted in (Procedure 5-2). Then, in (Procedure 6), all operations are extracted and the estimation is completed. Here, a processing sequence such as that shown in Figure 5(C) is extracted.

[0033] 6 and 7 are diagrams for explaining the procedure for extracting repetitive processing and branching processing. In Fig. 6 and Fig. 7, log (1) shown in the order a → b → c → a → b → c → a → b → c → a → b → c is set as the initial flow, and the cases where logs (2), (3), and (4) are estimated are shown. In Fig. 6, processing including branches, such as a → b → c → a → f → c → a → b → c → a → f → c in log (2), is estimated.

[0034] Furthermore, as shown in FIG. 7, complex processing as shown in logs (2) and (4) is estimated from the initial flow (1).

[0035] <Step S5: Estimating Scenario Structure> In step S4 of FIG. 2, the information processing device 1 modifies the estimated flow, primarily from the perspectives of visibility and maintainability, and then estimates the scenario structure. The constructed operation flow is an extraction of possible structures for all operation logs. As a result, it is often complex and may have poor visibility if left as is. Therefore, to improve visibility and maintainability, the flow is modified to meet, for example, the following requirements: - If the same type of node continues for a certain length (= l) or more, it is displayed in a degenerate manner. - The number of nesting levels (= m) is adjustable. This is to prevent the hierarchy from becoming excessively deep. - If the length of the node after branching is a certain value (= n) or more, it is decomposed into a separate flow. Note that the natural numbers l, m, and n may be set arbitrarily depending on the purpose.

[0036] FIG. 8 is a diagram for explaining flow modification. Applying the above requirements to the log (13), consecutive nodes are degenerated. This shortens the b → c → b → c section to b → c, improving visibility. Synthesis example 1 and synthesis example 2 each show an example of the results of adjusting the number of layers. For synthesis example 1, m = 2, and for synthesis example 2, m = 1.

[0037] Next, we will explain how to modify the flow. For example, if a flow that does not exist in the actual operation log is included, the constructed flow is decomposed. Operational flows that appear less frequently may be decomposed and represented as separate flows. Furthermore, in combination with the scenario structure, flows that appear more frequently may be extracted from the operational flow and proposed as operations that can be created in a library.

[0038] Fig. 9 is a diagram for explaining a case where a flow that does not exist in the actual operation log is included. In Fig. 9, the constructed operation flow (A) includes, for example, a flow d → a → b → c, but this does not exist in the actual operation log ((1) to (4) in Figs. 6 and 7). Therefore, for the operation flow that is not included in the actual operation log, the operation that caused it is separated and set as a new operation flow (operation flow (B) and (C)).

[0039] 10 is a diagram illustrating the processing of the information processing device 1 according to the embodiment. In FIG. 10, the information processing device 1 acquires operation log data of the terminal device 2 operated according to the automatic operation scenario (step S1), and then performs data shaping (step S2). The information processing device 1 then determines whether the operations are identical (step S3), and then estimates a branching structure and a repetitive structure (step S4). The estimation of the branching structure (step S41) and the estimation of the repetitive structure (step S42) may be performed simultaneously or separately.

[0040] Then, the information processing device 1 generates and outputs control structure information of the automatic operation scenario based on the estimated identical operations, branching structures, and repetition structures (step S5).

[0041] As described above, in this embodiment, an automatic operation scenario is executed multiple times to collect operation logs, and after formatting each operation log, identical operations are identified for the operation logs. Furthermore, identical operations are overlapped for each operation log, and repetitive processing and branching processing are extracted to construct an operation flow, and the flow is then modified from the perspectives of visibility and maintainability, and the control structure of the automatic operation scenario is identified.

[0042] This makes it possible to identify the control structure of an automation scenario regardless of the type of automation tool, improving the readability and visibility of the scenario. This has the added benefit of reducing the burden of updating / changing automation tools, and enabling the discovery of new scenarios that can be applied to other tasks.

[0043] In cases where an automation tool has already been introduced, there are cases where the desired process cannot be automated due to insufficient functionality of the automation tool, and so changing or adding the automation tool is considered. However, because the creation and expression methods for automation scenarios differ depending on the tool, when changing the automation tool, it is necessary to understand the processing content of the existing scenario based on knowledge of the original automation tool, and then create the scenario again using knowledge of the new automation tool.

[0044] Furthermore, the readability of an automated operation scenario varies depending on the creator, as it depends on how the block structure, node naming, etc. are created. For this reason, depending on how the scenario is created, it may be difficult to understand the processing content of the original scenario, even if you have knowledge of automated operation tools.

[0045] If the readability of an automated operation scenario is low, cases may arise where the processing content cannot be understood and the scenario is not applied to other tasks even though it could be applied, which could become an obstacle to promoting business efficiency.

[0046] In contrast, according to the embodiment, the control structure of an existing scenario can be inferred from the front end (or system screen) of the business system using history information of multiple attempts to perform multiple operations related to the existing scenario. Because the control structure of operations can be inferred from the front end, the readability and visibility of the automated operation scenario can be improved regardless of the type of automated operation tool or the creator of the automated operation scenario, and the person in charge can be made to understand the automated operation scenario more easily.

[0047] In addition, this invention can be embodied by modifying the components within the scope of the gist of the invention in the implementation stage. Furthermore, various inventions can be formed by appropriately combining multiple components disclosed in the above embodiments. For example, some components may be omitted from all the components shown in the embodiments. Furthermore, components from different embodiments may be appropriately combined.

[0048] 1...Information processing device 2...Terminal device 10...Processor 20...Storage 20a...Program 30...Memory 40...Communication unit 50...Interface unit 51...Input unit 52...Display unit 101...Data acquisition unit 102...Data formatting unit 103...Identical operation determination unit 104...Branch structure estimation unit 105...Structure estimation unit 106...Structure generation unit 301...Automatic operation scenario 302...Operation log.

Claims

1. An information processing apparatus comprising: a communication unit that communicates with a terminal device installed with an application controllable along an automatic operation scenario; a data acquisition unit that acquires, as an operation log, the processing content of the application according to the automatic operation scenario; an identical operation determination unit that determines identical operations included in the acquired operation log; a repetitive process estimation unit that estimates repetitive processes included in the acquired operation log; a branch process estimation unit that estimates branch processes included in the acquired operation log; and a structure generation unit that generates control structure information indicating the control structure of the automatic operation scenario based on the estimated identical operations, the estimated repetitive processes, and the estimated branch processes.

2. The information processing apparatus according to claim 1, further comprising a data shaping unit that performs preprocessing on the acquired operation log.

3. The operation log in the information processing apparatus according to claim 1 includes an attribute value of an operation of the application and a captured image of an operation screen related to the operation.

4. A method for estimating a control structure of an automatic operation scenario, comprising: a process in which a computer communicates with a terminal device installed with an application controllable along an automatic operation scenario and acquires, as an operation log, the processing content of the application according to the automatic operation scenario; a process in which the computer determines identical operations included in the acquired operation log; a process in which the computer estimates repetitive processes included in the acquired operation log; a process in which the computer estimates branch processes included in the acquired operation log; and a process in which the computer generates control structure information indicating the control structure of the automatic operation scenario based on the estimated identical operations, the estimated repetitive processes, and the estimated branch processes.

Images