Work management system and method

Abstract

A work management system according to an embodiment comprises: a collection agent provided in a worker terminal to collect input data to configure an input data string; and a work management server configured to generate at least one work vector by embedding at least one piece of work information received from the worker terminal in an artificial neural network-operable format, generate at least one corpus vector by embedding at least one corpus generated from a plurality of input data strings received from the collection agent in an artificial neural network-operable format, and allocate a corpus to work information on the basis of a similarity between the corpus vector and the work vector.

Description

Work Management System and Method

[0001] This technology relates to data analysis technology, and more specifically, to a work management system and method capable of classifying tasks performed by a worker.

[0002] Work management systems are being developed to monitor employee activities in remote or on-site work environments to measure productivity and improve work efficiency.

[0003] Through a work management system, employee activity information such as work time tracking, website visit history, keyboard input, and screenshots can be collected to analyze work patterns and improve inefficient activities.

[0004] In order for managers to fairly and accurately assess employees' work hours and performance, employees must frequently enter information about the work they have performed into the work management system.

[0005] However, as work information is entered manually, the objectivity of monitoring employee activities cannot be guaranteed, and work concentration may decrease.

[0006] An embodiment of the present technology can provide a work management system and method capable of automatically generating work information, classifying it by task, and assigning it.

[0007] A work management system according to one embodiment of the present technology may include: a collection agent provided in a worker terminal and collecting input data for the worker terminal to form an input data column; and a work management server configured to embed at least one work information received from the worker terminal into a format capable of artificial neural network computation to generate at least one work vector, embed at least one work bundle generated from a plurality of input data columns received from the collection agent into a format capable of artificial neural network computation to generate at least one work bundle vector, and assign the work bundle to the work information based on the similarity between the work bundle vector and the work vector.

[0008] A work management method according to one embodiment of the present invention is a method of operation of a work management system comprising a worker terminal having a collection agent that collects input data and forms an input data column, and a work management server connected to the worker terminal via a communication network, wherein the work management server receives at least one work information from the worker terminal and embeds the work information into a format capable of artificial neural network computation to generate at least one work vector; the work management server embeds at least one work bundle generated from a plurality of input data columns received from the collection agent into a format capable of artificial neural network computation to generate at least one work bundle vector; and the work management server assigns the work bundle to the work information based on the similarity between the work bundle vector and the work vector.

[0009] According to the present technology, work information can be automatically generated by collecting and analyzing information input from a worker terminal. In addition, work activities can be easily managed by automatically assigning the generated work information to related tasks.

[0010] Figure 1 is a configuration diagram of a business management system according to one embodiment.

[0011] FIG. 2 is a configuration diagram of a worker terminal according to one embodiment.

[0012] FIG. 3 is a configuration diagram of a business management server according to one embodiment.

[0013] Figure 4 is a diagram illustrating the concept of generating a cluster vector according to one embodiment.

[0014] FIG. 5 is a diagram illustrating the concept of assigning work bundles to a task according to one embodiment.

[0015] FIG. 6 is a diagram illustrating the concept of assigning tasks to a project according to one embodiment.

[0016] FIG. 7 is a flowchart illustrating a business management method according to one embodiment.

[0017] A work management system according to one embodiment may include a worker terminal and a work management server.

[0018] The above worker terminal may include a collection agent that collects input data for the above worker terminal and forms an input data column.

[0019] The above-described work management server may be configured to generate at least one work vector by embedding at least one work information received from the worker terminal into a format capable of artificial neural network computation, generate at least one work bundle vector by embedding at least one work bundle generated from a plurality of input data columns received from the collection agent into a format capable of artificial neural network computation, and assign the work bundle to the work information based on the similarity between the work bundle vector and the work vector.

[0020] Hereinafter, embodiments of the present technology will be described in more detail with reference to the attached drawings.

[0021] Figure 1 is a configuration diagram of a business management system according to one embodiment.

[0022] Referring to FIG. 1, a work management system (10) according to one embodiment may include a worker terminal (100) and a work management server (200) connected to the worker terminal (100) through a communication network (300).

[0023] The worker terminal (100) may be a computing device equipped with a communication module, an output device, and an input device, such as a desktop computer, a tablet PC, or a smartphone. The input device may include at least one of a keyboard, a mouse, a touchpad, a touch screen, or an electronic pen. The output device may include at least a display device.

[0024] A collection agent (110) may be installed on the worker terminal (100).

[0025] The collection agent (110) is for collecting input data for the worker terminal (100) and may be operated when the worker terminal (100) is booted or when a pre-configured application is operated on the worker terminal (100).

[0026] The collection agent (110) can collect all input data provided to the worker terminal (100) through at least one input device and form an input data column.

[0027] In one embodiment, the input data may include all linguistic data (natural language), non-linguistic data (non-natural language such as function keys), and observation image data input by a worker through various input devices connected to the worker terminal (100). Data other than natural language input, such as function key input and captured image data of the worker terminal (100) display, may be included in the non-linguistic data.

[0028] Linguistic data may be natural language data entered through various input devices, for example, language / number / symbol input data via a keyboard, or natural language input data from a virtual keyboard or extended input method via a mouse.

[0029] Non-verbal data may be any input data excluding natural language data, such as input of function keys via keyboard (space key, return key, Korean / English switching key, Ctrl key, Alt key, Shift key, Fn key, F1~F12 keys, Delete key, Backspace key, etc.) and function click data via mouse (left / right click, single / double click, button down, button up, drawing by dragging after button down, coordinate input, etc.).

[0030] Observation image data may be a display state capture image reflecting non-verbal input data, such as drawing input data for picture editing. Observation image data may be obtained by capturing a whole or partial image of the display as a set event is triggered.

[0031] The collection agent (110) can collect input data and form an input data column. In one embodiment, the input data column may be a plurality of input data and additional information for each of the input data listed according to the collection time.

[0032] Additional information may include input data collection time, additional image data, input target application name, input target website information, identification information of the worker terminal (100) or worker, file title (or web page title), etc. The additional image data may include full or partial image data of the screen captured at the moment the mouse is clicked for function input and thereafter, and captured image data of the target clicked by the mouse, and may be collected separately from the observed image data.

[0033] The collection agent (110) can provide an input data column containing input data and additional information thereon to the business management server (200).

[0034] In another embodiment, when the operating system (OS) of the worker terminal (100) or an application running on the worker terminal (100) has an input data collection function, the input data is collected from the operating system or each application where the input data is provided, and the collection agent (110) can organize the input data column together with additional information and provide it to the work management server (200).

[0035] The work management server (200) can receive and manage work information from the worker terminal (100).

[0036] In one embodiment, the work may include at least one project, and each project may include at least one task. Input through the worker terminal (100) may include an action for performing the task.

[0037] The work information may include project information for each of at least one project and task information for each of at least one task.

[0038] Project information may include a title, keyword, available application name, worker terminal (100) assigned to perform the work, or worker identification information.

[0039] Task information can include a title for each task.

[0040] In one embodiment, task information can be registered without dependency on project information.

[0041] The task management server (200) can generate a task vector by converting task information into a format capable of artificial neural network computation, for example, by embedding it. If the task information includes project information and task information, the task management server (200) can generate a task vector containing a project vector and a task vector by embedding the project information and task information, respectively. Embedding is a process of converting numeric or non-numerical data into numeric data, for example, a vectorization process.

[0042] The work management server (200) can divide the input data included in the input data column transmitted from the worker terminal (100) according to a set standard to form at least one work bundle, and convert the work bundle into a format capable of artificial neural network computation, for example, by embedding it to generate a work bundle vector.

[0043] In one embodiment, the work management server (200) may refer to additional information of the input data column to separate the input data collected during a preset period (e.g., 1 day) by the application name in which the input was made, and may form a work bundle by aggregating the input data having the same or similar file title (or web page title) by application name or by worker terminal (100) or by worker.

[0044] The input data included in the work bundle may include data with different attributes, such as linguistic data, non-linguistic data, and observational image data. The work management server (200) may embed the input data constituting the input data column included in the work bundle and integrate the embedded input data into a single vector having arbitrary dimensions to generate a work bundle vector. Ultimately, the work bundle vector may have a structure in which the results of embedding input data of various formats input to perform a specific task are mixed.

[0045] The process of forming a work bundle may further include a cutting process for cutting input data columns according to set criteria, and a connecting process for connecting input data columns according to set criteria. A detailed explanation of the cutting and connecting processes will be provided later with reference to FIG. 3.

[0046] The process of the work management server (200) generating a work cluster vector may include a substitution process that replaces non-verbal data and observation image data with linguistic data.

[0047] In order to convert non-verbal data into linguistic data, the task management server (200) can encode (natural language) non-verbal data, for example, function input data by mouse click. In order to convert observation image data into linguistic data, the task management server (200) can analyze the observation image data using a multimodal learning method, extract text from the observation image data, or use an image-to-text conversion function.

[0048] The work management server (200) can compare the work bundle vector and the task vector and assign the work bundle vector to the task with high similarity. The work management server (200) can compare the task vector with the project vector and assign the task to the project with high similarity. In one embodiment, various similarity measurement methods capable of determining the correlation between vectors, such as Euclidean distance and Mahalanobis distance, may be used to determine the similarity between vectors.

[0049] Accordingly, input data through the worker terminal (100) can be classified by associated task. Additionally, each task can be classified into associated projects.

[0050] FIG. 2 is a configuration diagram of a worker terminal according to one embodiment.

[0051] Referring to FIG. 2, a worker terminal (100) according to one embodiment may include a collection agent (110), a controller (120), a memory (130), a worker interface (UI, 140), and a communication network interface (150).

[0052] The controller (120) may be a central processing unit that controls the overall operation of the worker terminal (100), and may be configured in a combined form of software and hardware that executes it. In one embodiment, the controller (120) may be configured to execute an operating system (OS) to manage each hardware component and to control the processing of various programs, such as applications, firmware, and middleware, that are executed on the worker terminal (100).

[0053] The memory (130) may include a main memory that stores a program to be executed by the controller (120), data, or results processed by the controller (120), and an auxiliary memory such as a disk that stores the program to be executed and data to be executed by the controller (120).

[0054] The operator interface (UI, 140) may include an input device and an output device. The input device may include at least one of a keyboard, a mouse, a touch device, a drawing input device such as an electronic pen, and a touch pad equipped with a display unit. The output device may include at least one of an image display device such as a monitor, a speaker, a printer, and a video projector.

[0055] The communication network interface (150) can provide an environment in which the worker terminal (100) can communicate with other computing devices through the communication network (300).

[0056] The collection agent (110) may be downloaded and installed on the worker terminal (100) from the work management server (200) or other application providing server, and configured to execute program code written to perform a designated function through hardware. The collection agent (110) may be activated in response to various events, such as when the worker terminal (100) boots up, when a pre-configured application is executed on the worker terminal (100), or when a pre-configured website is accessed via the web browser of the worker terminal (100).

[0057] In one embodiment, the collection agent (110) may include a keyboard data collection unit (111), a mouse data collection unit (113), other data collection units (115), and an input data column configuration unit (117).

[0058] The keyboard data collection unit (111) can collect input data and additional information provided through the keyboard, such as the time of each input data collection, the name of the application to be input, information about the website to be input, file title or web page title, etc. The keyboard data can include all input data through the keyboard, such as language, numbers, symbols, and function keys (space key, return key, Korean / English switching key, Ctrl key, Alt key, Shift key, Fn key, F1~F12 keys, Delete key, Backspace key, etc.).

[0059] Tasks involving the input of linguistic data via a keyboard may include, for example, natural language input through document creation programs, natural language input through programs for creating files other than documents (e.g., email, design, image, video file creation, etc.), and input for naming files / folders through the OS.

[0060] Tasks involving the input of non-verbal data via a keyboard may include all keyboard inputs excluding natural language, such as formatting, calling and executing menus, and inputting control signals.

[0061] The mouse data collection unit (113) can collect input data and additional information provided through the mouse, namely the collection time of each mouse input data, additional image data (full or partial image data of the screen captured at the moment the mouse is clicked and thereafter, captured image data of the target clicked by the mouse), the name of the input target application, information of the input target website, file title or web page title, etc.

[0062] Input data provided through a mouse may include linguistic data consisting of natural language, such as characters, symbols, and numbers entered by operating a virtual keyboard or an extended input method, for example, and non-linguistic data excluding linguistic data.

[0063] Non-verbal data from a mouse can include all mouse inputs, such as format editing, menu calling and execution, control signal input, icon clicking, and drawing input.

[0064] In order to collect non-verbal data via a mouse, the mouse data collection unit (113) can collect all mouse input data, such as left / right clicks, single / double clicks, button down, button up, drawing through dragging after a button down, coordinate input, scroll wheel operation information, absolute coordinate values ​​of the clicked location or the drag start / end location, and coordinate values ​​of the clicked window (including the window size). In addition, additional image data, such as an image of the click target, an image around the location of the click or drag, and images before and after the click or drag, can be collected as additional information.

[0065] In one embodiment, data regarding the modification process accompanying the natural language / non-natural language input process via a keyboard or mouse may be optionally collected.

[0066] Other data collection unit (115) can collect observation image data, which is a display state capture image based on non-verbal input data such as drawing input data, and corresponding additional information, such as the observation image collection time, input target application name, input target website information, file title or web page title, etc. The other data collection unit (115) can collect observation image data when a pre-set event is triggered, such as when a set cycle arrives or when mouse input data is continuously input for a set period of time, but is not limited thereto. The observation image data may be a captured image of the entire screen or a part of the screen, or an image displayed in the window of a running program.

[0067] The input data column configuration unit (117) can configure an input data column by listing the input data and additional information collected from the keyboard data collection unit (111), the mouse data collection unit (113), and other data collection units (115) according to the collection time. The input data column configuration unit (117) can transmit the input data column to the business management server (200) at set times or periods, or whenever a set event occurs.

[0068] The input data column may further include worker terminal (100) identification information and / or worker identification information.

[0069] In this way, the collection agent (110) can collect all verbal and non-verbal input data provided to the worker terminal (100) through various input devices and observation image data reflecting the display state based on the non-verbal input data, list them according to the collection time, and transmit them to the work management server (200).

[0070] FIG. 3 is a configuration diagram of a business management server according to one embodiment.

[0071] Referring to FIG. 3, the business management server (200) may include a controller (210), memory (220), an input / output interface (IO IF, 230), a communication network interface (240), a business information management unit (250), a work information management unit (260), an embedding unit (270), a vector integration unit (280), and a classification unit (290).

[0072] The controller (210) may be configured to control the overall operation of the business management server (200) by implementing a combination of a program required for the business management server (200) to operate and hardware capable of executing the program.

[0073] The memory (220) can store programs and data to be executed by the controller (210), or results processed by the controller (210), or data transmitted from the worker terminal (100) through the communication network (300).

[0074] The input / output interface (IO IF, 230) can provide an input / output environment that allows an administrator or operator to access, manage, and control the business management server (200).

[0075] The communication network interface (240) provides an environment in which the work management server (200) can communicate with other computing devices, such as a worker terminal (100), through the communication network (300).

[0076] The work information management unit (250) can receive and store work information from the worker terminal (100).

[0077] In one embodiment, work information may include project information and task information. Project information may include a title, keyword, available application name, identification information of a worker terminal (100) authorized to perform the work, or worker identification information. Task information may include a title for each task and may be provided without dependency on project information.

[0078] As the worker's work information is assigned to an associated task and the task is assigned to an associated project, the work information management department (250) can update the work information.

[0079] The work information management unit (260) can generate work bundles by receiving input data sequences from the worker terminal (100).

[0080] The work information management unit (260) can refer to additional information of the input data column to separate the input data collected during a pre-set period by the application name in which the input was made, and can form a work bundle by collecting input data with the same or similar file title or web page title by application name, by worker terminal (100) or by worker.

[0081] The work information management unit (260) may include a cutting unit (261) and a connecting unit (263).

[0082] The cutting unit (261) can extract cutting trigger data from an input data column and, accordingly, separate the input data to cut the input data column into a plurality of sub-data columns. In one embodiment, the cutting trigger data may be at least one of the data indicating the following.

[0083] - Termination of linguistic input for a sentence or paragraph

[0084] - Changing input position by mouse (Enter / arrow key input, window click by mouse)

[0085] - Switching between active windows

[0086] - Termination of a program or file

[0087] The connection part (263) can connect identical or similar input data columns or sub-data columns that have a similarity level greater than or equal to a set threshold from the input data columns that are cut or before cutting, or input data columns or sub-data columns that have identical or similar file titles (or web page titles) entered through the same application.

[0088] In one embodiment, the connection unit (263) can connect input data columns or sub-data columns that have a similarity higher than a set threshold even if the worker terminal (100) identification information that provided the input data columns is different. Accordingly, work performed by the same worker through different worker terminals (100) can be integrated into a single work group.

[0089] The embedding unit (270) can convert work information and work bundles into a format that can be processed by an artificial neural network.

[0090] In one embodiment, the embedding unit (270) can generate a project vector, a task vector, and a work bundle vector from each of the project information, task information, and work bundle included in the business information. The embedding unit (270) can generate vectors using a language model such as, for example, a Large Language Model, Word2Vec, Globe, FastText, ELMO, BERT, Doc2Vec, etc.

[0091] To generate a bundle vector, the embedding unit (270) can convert non-verbal data and observation image data into linguistic data by replacing them with linguistic data, thereby converting the non-verbal data and observation image data into the same format as the linguistic data.

[0092] In one embodiment, the embedding unit (270) can convert non-verbal data, such as function input data from a mouse click, into linguistic data by digitizing each of them, that is, by assigning a code to the non-verbal data. In one embodiment, the same digital code can be assigned to the same non-verbal data.

[0093] In one embodiment, the embedding unit (270) can convert additional image data or observed image data into linguistic data by analyzing them using a multimodal learning method, extract text from the additional image data or observed image data, or convert them into linguistic data using an image-to-text conversion function.

[0094] When all input data related to the task, such as non-verbal input data and observation image data, is replaced with linguistic data by the embedding unit (270), the input data sequence can be embedded as linguistic data in its entirety.

[0095] The vector integration unit (280) can integrate the embedded input data and convert them into a single bundle vector having arbitrary dimensions.

[0096] In one embodiment, the embedding unit (270) can generate feature information including keywords and / or summaries from an integrated bundle vector.

[0097] The classification unit (290) can assign a work bundle to an associated task based on the similarity between the work bundle vector and the task vector. Additionally, it can assign a task to an associated project based on the similarity between the task vector and the project vector.

[0098] If project information includes an available application name, among the work bundles assigned to a specific project, work bundles performed by an application that is not registered with an available application name for that project may be excluded from the tasks assigned to that project.

[0099] As a work group is assigned to a related task and the task is assigned to a related project, the work information managed by the work information management department (250) can be updated.

[0100] Accordingly, the worker's work information can be automatically assigned to an associated task and the task can also be automatically assigned to an associated project without the worker having to manually input information about the work performed by the worker into the work management server (200) at any time.

[0101] Figure 4 is a diagram illustrating the concept of generating a cluster vector according to one embodiment.

[0102] Referring to FIG. 4, the application name (APP NAME) of the input data included in work group 1 may be "Chrome" and the title (TITLE) may be "Precision Control Quotation". The worker's input data for this may include natural language data such as "Regarding precision control...".

[0103] The embedding unit (270) of the work management server (200) can embedding work bundle 1 to generate a work bundle vector (0, 25, 0.14, ..., 0.72).

[0104] FIG. 5 is a diagram illustrating the concept of assigning work bundles to a task according to one embodiment.

[0105] Referring to Fig. 5, the title of Task 1 may be "Development of precision control system," the title of Task 2 may be "Construction of abnormal behavior server," and the title of Task 3 may be "Identification of attribute recognition object."

[0106] The embedding unit (270) of the task management server (200) can embedding each task information to generate a task vector (0, 25, 0.14, ..., 0.72) for task 1, a task vector (0.95, 0.27, ..., 0.72) for task 2, and a task vector (0.57, 0.32, ..., 0.73) for task 3.

[0107] The classification unit (290) can measure the similarity between the work bundle vector (0, 25, 0.14, ..., 0.72) and the task vectors (0, 25, 0.14, ..., 0.72), (0.95, 0.27, ..., 0.74), (0.57, 0.32, ..., 0.73) as 0.83, 0.25, and 0.47, respectively, and assign the work bundle to the task with the highest similarity.

[0108] Figure 5 illustrates an example where the 1st work group of Figure 4 is assigned to Task 1.

[0109] FIG. 6 is a diagram illustrating the concept of assigning tasks to a project according to one embodiment.

[0110] Referring to Fig. 6, the title of Project 1 is "Building Company A's System," and the keywords may include "precision control, axis, angle, ...". The title of Project 2 is "Company B's Abnormal Behavior Service," and the keywords may include "abnormal behavior, facial expression, motion, ...".

[0111] The embedding unit (270) of the business management server (200) can embedding each project information to generate project vectors (0, 34, 0.14, ..., 0.72) for project 1 and project vectors (0.95, 0.37, ..., 0.72) for project 2.

[0112] The classification unit (290) can measure the similarity between task vectors (0, 25, 0.14, ..., 0.72), (0.95, 0.27, ..., 0.74), (0.57, 0.32, ..., 0.73) and project vectors (0, 34, 0.14, ..., 0.72) (0.95, 0.37, ..., 0.72) as 0.92 and 0.54, respectively, and assign the task to the project with the highest similarity.

[0113] Figure 6 illustrates an example where Task 1 of Figure 5 is assigned to Project 1.

[0114] FIG. 7 is a flowchart illustrating a business management method according to one embodiment.

[0115] The work management server (200) can receive work information from the worker terminal (100) (S101). The work information may include project information and task information, and the task information may be received without dependency on the project information. In one embodiment, the project information may include a title, keyword, available application name, identification information of the worker terminal (100) allowed to perform the work, or worker identification information, etc. for each project. The task information may include a task title.

[0116] The work management server (200) can receive input data sequences from the worker terminal (100) and form work bundles (S103).

[0117] The work management server (200) can generate a project vector, a task vector, and a work bundle vector by embedding the project information and task information included in the work information and the work bundle configured from the input data column (S105).

[0118] In one embodiment, the work management server (200) can form a work bundle by cutting and connecting the input data using the cutting part (261) and connecting part (263) described above before embedding the input data (S103).

[0119] In one embodiment, the work management server (200) can embed the linguistic data, non-linguistic data, and observation image data included in the input data column, respectively (S105), and then integrate them into a single vector to integrate the work bundle into linguistic data as a whole (S107).

[0120] The work management server (200) can measure the similarity between the work bundle vector and the task vector, and the similarity between the task vector and the project vector (S109).

[0121] The work management server (200) can assign work bundle vectors to the task vector with the highest similarity and assign the task vectors to the project vector with the highest similarity, thereby classifying the work bundles of the worker's work by task and each task by project (S111).

[0122] If the project information includes an available application name, among the work bundles assigned to the project, work bundles performed by an application that is not registered with an available application name for that project may be excluded from the tasks assigned to that project.

[0123] In this way, input data for the worker terminal (100) can be configured into an input data column and transmitted to the work management server (200) to be converted into a work bundle vector. Then, each work bundle is automatically assigned to a task included in the work information, and each task is automatically assigned to a project, so that the worker's activities can be easily collected without manually entering work information.

[0124] As workers' work information is automatically generated and assigned to tasks, workers can be freed from the hassle of manually entering work information, and managers can fairly and accurately track employees' work hours and performance.

[0125] Those skilled in the art to which the present invention pertains will understand that the present invention may be implemented in other specific forms without altering its technical concept or essential features. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. The scope of the present invention is defined by the claims set forth below rather than by the detailed description above, and all modifications or variations derived from the meaning and scope of the claims and equivalents thereof should be interpreted as being included within the scope of the present invention.

[0126] According to the present technology, by collecting and analyzing employee activity information such as work time tracking, website visit history, keyboard input, and screenshots from worker terminals to automatically generate work information, work productivity can be improved and inefficient activities can be improved.

[0127] In addition, automatically generated work information can be automatically assigned to related tasks, making it easier to manage work activities.

Claims

1. A collection agent equipped in a worker terminal and collecting input data for said worker terminal to form an input data column; and A work management server configured to generate at least one work vector by embedding at least one work information received from the worker terminal into a format capable of artificial neural network computation, generate at least one work bundle vector by embedding at least one work bundle generated from a plurality of input data columns received from the collection agent into a format capable of artificial neural network computation, and assign the work bundle to the work information based on the similarity between the work bundle vector and the work vector; A business management system including 2. In Paragraph 1, The above input data column includes the application name and file title or webpage title to which the input data is entered, and The above-mentioned work management server is a work management system that aggregates input data from input data columns having identical or similar file titles or web page titles for each application name, and forms the above-mentioned work bundles for each worker terminal identifier or worker identifier.

3. In Paragraph 1, The above business information includes task information for each of at least one task, and A work management system configured such that the above work management server generates a work vector including a task vector that embedding the above task information, and assigns the work bundle to an associated task based on the similarity between the work bundle vector and the above task vector.

4. In Paragraph 3, The above task information is a work management system including a task title.

5. In Paragraph 3, The above business information includes project information for each of at least one project, and The above-described task management server is a task management system configured to generate a task vector including a project vector that embeds the above-described project information, and to assign the task to an associated project based on the similarity between the task vector and the project vector.

6. In Paragraph 5, The above project information is a work management system including a project title, keywords, available application names, and identification information of allowed worker terminals or workers.

7. In Paragraph 6, The above-mentioned task management server is a task management system configured to exclude from task assignment work bundles performed in applications not registered with the above-mentioned available application names among the work bundles assigned to tasks assigned to the above-mentioned project.

8. In Paragraph 1, The above-mentioned worker terminal includes at least one input device and an output device, and A business management system in which the above-mentioned collection agent collects a plurality of input data, including linguistic data, non-linguistic data, and observational image data input through the above-mentioned at least one input device, in the order of input, and organizes the input data and additional information corresponding to the input data into the input data column.

9. In Paragraph 8, The above additional information is an input task management system that includes the collection time of each of the plurality of input data, the name of the input target application, the information of the input target website, identification information of the worker terminal or worker, and the file title or web page title.

10. A method of operation of a work management system comprising a worker terminal having a collection agent that collects input data and forms an input data column, and a work management server connected to the worker terminal via a communication network, The above-mentioned task management server receives at least one task information from the worker terminal and embeds the task information into a format capable of artificial neural network computation to generate at least one task vector; The above-mentioned task management server generates at least one task bundle by embedding at least one task bundle generated from a plurality of input data columns received from the collection agent into a format capable of artificial neural network computation, thereby generating at least one task bundle vector; and The above-mentioned task management server assigns the work bundle to the task information based on the similarity between the work bundle vector and the task vector; A method of operation of a business management system including 11. In Paragraph 10, The above input data column includes the application name and file title or webpage title to which the input data is entered, and The step of generating the above work bundle is, A method of operation of a business management system comprising the step of the above-mentioned business management server collecting input data of an input data column having the same or similar file title or web page title for each application name for each worker terminal identifier or worker identifier.

12. In Paragraph 10, The above business information includes task information for each of at least one task, and The step of generating the above task vector includes the step of generating a task vector by embedding the above task information, and A method of operation of a work management system comprising the step of assigning the above work bundle to the above work information, the step of assigning the above work bundle to an associated task based on the similarity between the above work bundle vector and the above task vector.

13. In Paragraph 12, The above task information is a method of operation of a business management system including a task title.

14. In Paragraph 12, The above business information includes project information for each of at least one project, and The step of generating the above-mentioned work vector further includes the step of generating a project vector by embedding the above-mentioned project information, and A method of operation of a work management system that further includes the step of assigning the above work bundle to the above work information, and the step of assigning the above task to an associated project based on the similarity between the above task vector and the above project vector.

15. In Paragraph 14, The above project information includes a project title, keywords, available application names, and identification information of an allowed worker terminal or worker. A method of operation of a work management system.

16. In Paragraph 15, A method of operation of a business management system, wherein the step of assigning the above work bundle to the above business information further includes the step of excluding from the task assignment work bundles performed in applications not registered with the above available application names among the work bundles assigned to the tasks assigned to the above project.

17. In Paragraph 10, The above-mentioned worker terminal includes at least one input device and an output device, and A method of operation of a business management system in which the above-described collection agent collects a plurality of input data, including linguistic data, non-linguistic data, and observational image data input through the above-described at least one input device, in the order of input, and the input data and additional information corresponding to the input data are configured as the input data column.

18. In Paragraph 17, The above additional information is a method of operation of an input task management system including the collection time of each of the plurality of input data, the name of the input target application, the information of the input target website, identification information of a worker terminal or worker, and a file title or web page title.

Images