A task mining method and device, electronic equipment and storage medium

Abstract

The application discloses a task mining method and device, electronic equipment and storage medium, the method comprises the following steps: obtaining an original operation log data set; wherein the original operation log data set comprises a plurality of log records for representing operation events, and each log record comprises at least a task identifier and a position identifier for marking the position of the operation event; for each log record in the original operation log data set, determine operation events under each position identifier; wherein the operation events are obtained by performing mapping on the log record; based on the operation events under each position identifier and the task identifier, determine key operation events under each task identifier; and generate a flowchart according to the key operation events under each task identifier. The application improves the event granularity expressed by the log record, and the generated flowchart is simple and can represent the specific execution of a task, which is convenient for users to understand and optimize the task.

Description

Technical Field

[0001] This invention relates to the field of task mining technology, and in particular to a task mining method, apparatus, electronic device, and storage medium. Background Technology

[0002] Task mining is a technology that enables organizations to capture detailed steps of tasks performed on users' desktops, whether independently or in collaboration with colleagues. By analyzing recorded user actions, task mining allows organizations to gain insights into how they perform tasks, identify common errors during task execution, and determine tasks that can be automated—all of which can help optimize their business processes.

[0003] In existing technologies, flowcharts generated by existing task mining methods often have many nodes, making it difficult for users to understand and optimize tasks. Summary of the Invention

[0004] This invention provides a task mining method and apparatus to solve the technical problem that existing flowcharts have too many nodes, making it difficult for users to understand and optimize tasks.

[0005] In a first aspect, the present invention provides a task mining method, comprising:

[0006] Obtain the original operation log dataset; wherein, the original operation log dataset includes multiple log records used to characterize operation events, and each log record includes at least a task identifier and a location identifier used to indicate the location where the operation event occurred;

[0007] For each log record in the original operation log dataset, determine the various types of operation events under each location identifier; wherein, the operation event is obtained by mapping the log record;

[0008] Based on the operation events under each location identifier and the task identifier, determine the key operation events under each task identifier;

[0009] A flowchart is generated based on the key operation events under each task identifier.

[0010] The above technical solution improves the granularity of events expressed in log records by using operation events. By combining task identifiers and location identifiers, it identifies key operation events under the task identifier. The generated flowchart is both concise and can represent the specific execution status of a task, making it easier for users to understand and optimize tasks.

[0011] Optionally, the location identifier may include a window title or an application name.

[0012] Optionally, the log recording further includes element tags, and the step of determining the operation events under each location identifier includes:

[0013] For each log record in the original operation log dataset, determine the log record under each of the aforementioned location identifiers;

[0014] For log records under the same location identifier, log records with the same element tag are mapped to the same type of operation event.

[0015] Optionally, the steps for determining the key operational events under each task identifier include:

[0016] Based on the operation events under each location identifier and the task identifier, determine the operation events under each task identifier and their corresponding occurrence counts; wherein, the occurrence count is the number of log records corresponding to each type of operation event;

[0017] Based on the number of occurrences of the operation events, calculate the importance parameters of the operation events under each task identifier;

[0018] Operational events that meet the importance parameter threshold are identified as critical operational events.

[0019] Optionally, the importance parameters of each operation event under each task identifier can be calculated using word frequency and inverse text frequency index algorithms.

[0020] Optionally, the step of generating a flowchart based on the key operation events under each task identifier includes:

[0021] Based on the key operation events under each task identifier, filter the log records in the original operation log dataset;

[0022] A corresponding flowchart is generated using the filtered operation log dataset; wherein the nodes in the flowchart are obtained by merging log records based on location identifiers.

[0023] In a second aspect, the present invention provides a data mining apparatus, comprising:

[0024] The acquisition module is used to acquire the original operation log dataset; wherein, the original operation log dataset includes multiple log records for characterizing operation events, and each log record includes at least a task identifier and a location identifier for indicating the location where the operation event occurred;

[0025] The classification module is used to determine various types of operation events under each location identifier for each log record in the original operation log dataset; wherein, the operation events are obtained by classifying the log records.

[0026] The determination module is used to determine the key operation events under each task identifier based on the operation events under each location identifier and the task identifier;

[0027] The generation module is used to generate flowcharts based on the key operation events under each task identifier.

[0028] Thirdly, the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the task mining method described in the first aspect above.

[0029] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the task mining method described in the first aspect above.

[0030] Compared with the prior art, the beneficial effects of the present invention are:

[0031] 1. The method of the present invention identifies various types of operation events under each location identifier for each log record in the original operation log dataset. This step performs a mapping operation on the log record to obtain the operation event, based on the location identifier, thereby improving the granularity of the events expressed by the log record.

[0032] 2. The method of this invention determines the key operation events under each task identifier based on the operation events and task identifiers under each location identifier; then, a flowchart is generated based on the key operation events under each task identifier. This method considers task identifiers, combining task identifiers and location identifiers to determine the key operation events under each task identifier. The generated flowchart is both concise and effectively represents the specific execution status of a task, facilitating user understanding and optimization of the task. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0034] Figure 1 This is a flowchart illustrating the task mining method according to an embodiment of the present invention;

[0035] Figure 2 This is a schematic diagram of the structure of a task excavation device according to an embodiment of the present invention;

[0036] Figure 3 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0038] In related technologies, task mining technology can automatically discover and extract tasks from collected operation log data. Existing task mining products, after analyzing the input data, can generate flowcharts from the perspective of the operation window or from the perspective of operation events. Once analysts obtain the flowcharts of the users being analyzed, they can perform analyses such as Return on Investment (ROI) analysis, task analysis, time consumption analysis, and process comparison to optimize existing task operations. Both analysts and users being analyzed are collectively referred to as users.

[0039] In practice, operation log data collected from user actions often involves different applications and windows, and covers different tasks. Furthermore, the log entries in this operation log data are collected and recorded by the operating system, such as the Event Recorder in Windows. The Event Recorder meticulously records every action a user takes on the current terminal. This granularity of operation events is often overly detailed. For example, consider a user submitting an expense report. In a report form window, information such as amount, expense type, and claimant's name needs to be filled in. Some users might type this directly, some might use the clipboard, some might enter it all at once, and some might modify it multiple times. The Event Recorder records all these operational details. Therefore, for users to understand and analyze a task, the overly detailed event granularity in the operation log data actually hinders their comprehension and analysis. Consequently, this is reflected in flowcharts with an excessive number of nodes, making it difficult for users to analyze the overall task situation.

[0040] To address this problem, this invention proposes a task mining method. This method identifies various types of operation events under different location identifiers for each log record in the original operation log dataset. The purpose of this step is to classify log records by location identifier, grouping them into the same category to represent the same type of operation event, thus merging log records and reducing the data volume in the original operation log dataset. Next, based on the operation events and task identifiers under each location identifier, the method identifies key operation events under each task identifier; and generates a flowchart based on these key operation events. This step considers task identifiers, identifies key operation events under each task identifier, and finally generates corresponding flowcharts based on these key operation events, further reducing the data volume in the original operation log dataset. The resulting flowcharts are more concise than those in existing technologies. Correspondingly, the flowcharts generated by this invention, by considering key operation events under each task identifier, while concise, can still represent the overall execution of a task, helping users to understand and analyze tasks relatively quickly.

[0041] like Figure 1 As shown, the task mining method provided by this invention includes the following steps:

[0042] Step 100: Obtain the original operation log dataset; wherein the original operation log dataset includes multiple log records used to characterize operation events, and each log record includes at least a task identifier and a location identifier used to indicate the location where the operation event occurred.

[0043] In this embodiment, the original operation log dataset includes multiple log records. These log records characterize user operation events and can have various attributes, such as a unique identifier, task identifier, timestamp, location identifier, element tag, and operation event type for each log record. The task identifier distinguishes the task category to which different log records belong, and the location identifier indicates the location where the operation event occurred; for example, it could be a window title or application name. The element tag is the object directly targeted by the user's operation, such as the amount, reimbursement type, and claimant mentioned above. The operation event type can be mouse click, keyboard input, clipboard input, etc.

[0044] In this embodiment, the log entries can be recorded using the operation event logger in the operating system, such as the operation event logger in Windows. Alternatively, screen recording can be used to obtain screenshots of the user's desktop operations, and then text recognition can be performed on the screenshots to extract relevant information to supplement the original operation log dataset. The original operation log dataset includes multiple log records, which can be records of a single user performing different tasks, or records of multiple users performing the same or different tasks.

[0045] Step 101: For each log record in the original operation log dataset, determine the various types of operation events under each location identifier; wherein, the operation events are obtained by mapping each log record.

[0046] In this embodiment, the operation event is obtained by mapping based on one or more key attributes of a log record. The key attribute is semantic information that can represent the operation event corresponding to that log record. Therefore, log records with the same key attributes can represent the same operation event.

[0047] In this embodiment, the key attributes mentioned above can be element tags and operation event types. Mapping can be performed based solely on element tags, or it can be performed based on both element tags and operation event types.

[0048] Specifically, this means that log records with the same element label and operation event type under a certain location, such as under the window title, are mapped to the same type of operation event. For example, in the scenario of a user performing an expense reimbursement, in the corresponding original operation log dataset, the user performs the operation of entering their name in the expense reimbursement form window title. For instance, the user's name is: A, B, C, and the user enters it using the keyboard. This could involve first typing "Jia," then deleting "Jia," then typing "A," and finally typing "B, C." Since there might be more than 10 operation records in the original operation log dataset for the user entering "A, B, C," using the method in this embodiment, all of them can be mapped to the same type or the same operation event, namely, "entering name."

[0049] In this implementation, the aforementioned key attributes can also be element tags and indicators of whether the content has changed. That is, regardless of the type of event operation the user uses—for example, multiple keystrokes, deletion followed by re-entry, or a single clipboard input (e.g., pasting "A, B, C")—as long as the content changes, it is considered the same operation event. This method considers a scenario in real-world user operations. Specifically, users often copy and paste content across multiple windows. They might select one or more fields in one window, right-click and choose copy, then switch to another window to paste. For the former window, the user's operation doesn't modify any content. If only element tags are considered without considering whether the content changes, the log records in the above example will inevitably be mapped to the same type of operation event, which doesn't reflect the true semantics of the user's operation.

[0050] In this embodiment, the location identifier can be the application name or window title, preferably the window title, or other identifiers that can accurately identify the user's operations in the UI (User Interface).

[0051] In some cases, log records lack element tags. For example, an increasing number of applications have built-in smart fill functionality, where users provide a button to fill in multiple fields simultaneously. Using the same scenario, in a CRM (Customer Relationship Management) application with smart fill, a user might simultaneously fill in the amount, reimbursement type, and claimant's name via a smart fill button. In this scenario, the corresponding log record would lack element tags for the amount, reimbursement type, and claimant's name. Therefore, predefined rules can be used to map the user's button click as the user entering the amount, selecting the reimbursement type, and entering the name.

[0052] The purpose of the above steps is to merge operation events by using location identifiers, such as window titles, so that operation events under the same window title are all displayed on a single node in the flowchart. Secondly, for each log record in the original operation log dataset, a mapping method is used to map multiple log records to a single operation event, thereby reducing the data volume and improving the granularity perceived by the user.

[0053] Step 102: Based on the operation events under each location identifier and the task identifier, determine the key operation events under each task identifier.

[0054] In this embodiment, the aforementioned key operation events are semantic information that represents the task corresponding to the operation event. The key operation events under each task identifier can be determined by the following method:

[0055] Step 1021: Based on the operation events under each location identifier and the task identifier, determine the operation events under each task identifier and the corresponding number of occurrences.

[0056] Specifically, the number of times a certain operation event occurs is represented by the number of log records corresponding to that operation event. In other words, the number of times an operation event occurs is the number of log records that map to it.

[0057] Step 1022: Based on the number of occurrences of the operation events, calculate the importance parameters of the operation events under each task identifier.

[0058] In this embodiment, one feasible approach is as follows: Since the operation events and their corresponding occurrence counts under each task identifier can be obtained in step 1021, the occurrence counts of all operation events under a certain task identifier can be summed as the denominator, and the occurrence counts of each operation event can be used as the numerator to obtain the weight parameters of each operation event under that task identifier. Based on this calculation method, the weight parameters of each operation event under each task identifier can be further obtained. This weight parameter can be defined as a first weight parameter. In this embodiment, the first weight parameter measures the frequency of an operation event under a certain task identifier. Correspondingly, a high frequency of an operation event under a certain task identifier does not necessarily mean that the operation event is a key operation event capable of representing the semantic information of the task. Furthermore, a low frequency of an operation event under a certain task identifier does not necessarily mean that the operation event cannot be a key operation event representing the semantic information of the task. For example, in the scenario of a user submitting an expense report, under a given task identifier, the frequency of actions such as entering the amount, expense type, and name of the person submitting the report is often much higher than the frequency of clicking the submit button. However, clicking the submit button is a key action that represents the semantic information of the task. Therefore, horizontal comparisons across multiple task identifiers are also necessary. After determining the first weight parameter for each action under each task identifier, the second weight parameter for each action in the original operation log dataset can be calculated. For example, a logarithmic function log(x) can be constructed, where x is the total number of action events in the original operation log dataset divided by the number of a particular action in the original operation log dataset. Multiplying the first weight parameter by the second weight parameter yields the importance parameter of the action event under each task identifier.

[0059] Another feasible approach is to use the TF-IDF (Term Frequency Inverse Document Frequency) method. The first weight parameter mentioned above is the TF value, and then the IDF value is constructed by using a logarithmic function log(x), where x is the total number of tasks in the original operation log dataset divided by the number of tasks with a certain operation event.

[0060] Step 1023: Identify the operation events that meet the importance parameter threshold as critical operation events.

[0061] In this embodiment, besides determining the key operation events under each task identifier using the methods described above, another feasible approach is to set a first threshold to filter out operation events that occur too infrequently in the original operation log dataset; this threshold is used to exclude operation events that occur too infrequently overall in the original operation log dataset. Then, a second threshold is used to filter out operation events in each task that do not meet the required frequency; the remaining operation events are the key operation events.

[0062] Step 103: Generate a flowchart based on the key operation events under each task identifier.

[0063] In this embodiment, step 103 may include: filtering log records in the original operation log dataset based on key operation events under each task identifier; generating a corresponding flowchart using the filtered operation log dataset; wherein the nodes in the flowchart are obtained by merging log records based on location identifiers. Specifically, by identifying key operation events to filter log records in the original operation log dataset, the original operation log dataset is simplified. Based on the simplified operation log dataset, a process mining algorithm, such as the InductiveMiner algorithm, can be used to obtain a flowchart. The flowchart generated in this way is more concise and can also represent the overall execution status of a task, helping users to understand and analyze tasks relatively quickly.

[0064] The task mining apparatus provided by the present invention is described below. The task mining apparatus described below and the task mining method described above can be referred to in correspondence.

[0065] like Figure 2 As shown, the task excavation device includes the following modules:

[0066] The acquisition module 200 is used to acquire the original operation log dataset; wherein, the original operation log dataset includes multiple log records for characterizing operation events, and each log record includes at least a task identifier and a location identifier for indicating the location where the operation event occurred;

[0067] The classification module 210 is used to determine various types of operation events under each location identifier for each log record in the original operation log dataset; wherein, the operation events are obtained by classifying the log records.

[0068] The determination module 220 is used to determine the key operation events under each task identifier based on the operation events under each location identifier and the task identifier;

[0069] The generation module 230 is used to generate a flowchart based on the key operation events under each task identifier.

[0070] Optionally, the location identifier may include a window title or an application name.

[0071] Optionally, the log recording further includes element tags, and the step of determining the operation events under each location identifier includes:

[0072] For each log record in the original operation log dataset, determine the log record under each of the aforementioned location identifiers;

[0073] For log records under the same location identifier, log records with at least some of the same element tags are mapped to the same type of operation event.

[0074] Optionally, the steps for determining the key operational events under each task identifier include:

[0075] Based on the operation events under each location identifier and the task identifier, determine the operation events under each task identifier and their corresponding occurrence counts; wherein, the occurrence count is the number of log records corresponding to each type of operation event;

[0076] Based on the number of occurrences of the operation events, calculate the importance parameters of the operation events under each task identifier;

[0077] Operational events that meet the importance parameter threshold are identified as critical operational events.

[0078] Optionally, the importance parameters of each operation event under each task identifier can be calculated using word frequency and inverse text frequency index algorithms.

[0079] Optionally, the step of generating a flowchart based on the key operation events under each task identifier includes:

[0080] Based on the key operation events under each task identifier, filter the log records in the original operation log dataset;

[0081] A corresponding flowchart is generated using the filtered operation log dataset; wherein the nodes in the flowchart are obtained by merging log records based on location identifiers.

[0082] Figure 3 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 3 As shown, the electronic device may include: a processor 310, a communication interface 320, a memory 330, and a communication bus 340, wherein the processor 310, the communication interface 320, and the memory 330 communicate with each other via the communication bus 340. The processor 310 can call logical instructions in the memory 330 to execute a task mining method, which includes:

[0083] Obtain the original operation log dataset; wherein, the original operation log dataset includes multiple log records used to characterize operation events, and each log record includes at least a task identifier and a location identifier used to indicate the location where the operation event occurred;

[0084] For each log record in the original operation log dataset, determine the various types of operation events under each location identifier; wherein, the operation event is obtained by mapping the log record;

[0085] Based on the operation events under each location identifier and the task identifier, determine the key operation events under each task identifier;

[0086] A flowchart is generated based on the key operation events under each task identifier.

[0087] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0088] In another aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the task mining methods provided by the methods described above, the method comprising:

[0089] Obtain the original operation log dataset; wherein, the original operation log dataset includes multiple log records used to characterize operation events, and each log record includes at least a task identifier and a location identifier used to indicate the location where the operation event occurred;

[0090] For each log record in the original operation log dataset, determine the various types of operation events under each location identifier; wherein, the operation event is obtained by mapping the log record;

[0091] Based on the operation events under each location identifier and the task identifier, determine the key operation events under each task identifier;

[0092] A flowchart is generated based on the key operation events under each task identifier.

[0093] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0094] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0095] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A task mining method, characterized in that, include: Obtain the original operation log dataset; wherein, the original operation log dataset includes multiple log records used to characterize operation events, and each log record includes at least a task identifier and a location identifier used to indicate the location where the operation event occurred; For each log record in the original operation log dataset, determine the various types of operation events under each location identifier; wherein, the operation event is obtained by mapping the log record, and the mapping action is to map the log records in the original operation log dataset to the same type or the same operation event; Based on the operation events under each location identifier and the task identifier, the key operation events under each task identifier are determined. The key operation events are semantic information that can represent the task corresponding to the operation event. A flowchart is generated based on the key operation events under each task identifier.

2. The task mining method according to claim 1, characterized in that, The location identifier includes the window title or application name.

3. The task mining method according to claim 1, characterized in that, The log recording also includes element tags, and the steps for determining the operation events under each of the aforementioned location identifiers include: For each log record in the original operation log dataset, determine the log record under each of the aforementioned location identifiers; For log records under the same location identifier, log records with the same element tag are mapped to the same type of operation event.

4. The task mining method according to claim 3, characterized in that, The steps to determine the key operational events under each task identifier include: Based on the operation events under each location identifier and the task identifier, determine the operation events under each task identifier and their corresponding occurrence counts; wherein, the occurrence count is the number of log records corresponding to each type of operation event; Based on the number of occurrences of the operation events, calculate the importance parameters of the operation events under each task identifier; Operational events that meet the importance parameter threshold are identified as critical operational events.

5. The task mining method according to claim 4, characterized in that, The importance parameters of each operation event under each task identifier are calculated using word frequency and inverse text frequency index algorithms.

6. The task mining method according to claim 1, characterized in that, The step of generating a flowchart based on the key operation events under each task identifier includes: Based on the key operation events under each task identifier, filter the log records in the original operation log dataset; A corresponding flowchart is generated using the filtered operation log dataset; wherein the nodes in the flowchart are obtained by merging log records based on location identifiers.

7. A data mining device, characterized in that, include: The acquisition module is used to acquire the original operation log dataset; wherein, the original operation log dataset includes multiple log records for characterizing operation events, and each log record includes at least a task identifier and a location identifier for indicating the location where the operation event occurred; The classification module is used to determine various types of operation events under each location identifier for each log record in the original operation log dataset; wherein, the operation events are obtained by classifying the log records. The determination module is used to determine the key operation events under each task identifier based on the operation events under each location identifier and the task identifier; The generation module is used to generate flowcharts based on the key operation events under each task identifier.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the task mining method as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the task mining method as described in any one of claims 1 to 6.

Images