Information processing method, learning model generation method, computer program, and information processing device

Abstract

To provide an information processing method capable of expecting support of creating a process chart, a method for generating a learning model, a computer program and an information processing device.SOLUTION: In an information processing method, an information processing device receives an input of a process, inputs the received process to a process prediction model learned so as to output a candidate for the next process for the input of the process, acquires the candidate for the next process outputted by the process prediction model, and outputs the acquired candidate for the next process. Also, the information processing device may output a plurality of candidates for the next process, receive one process selection input from among the plurality of candidates, input the received process to the process prediction model, acquire the candidate for the next process outputted by the process prediction model, and output the acquired candidate for the next process.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an information processing method for supporting the creation of schedules for construction work, civil engineering work, etc., a method for generating a learning model, a computer program, and an information processing device. [Background technology]

[0002] A schedule is used to grasp the progress of construction or civil engineering work at the site. A schedule is a table that summarizes multiple steps performed in a construction or civil engineering work and the implementation period of each step, and can be created as, for example, a Gantt chart or a bar chart.

[0003] Patent document 1 proposes a method of creating a schedule in which a server creates a schedule including multiple processes, the craftsmen in charge of the processes, and the construction period for each process based on data on registered craftsmen, data on the craftsmen's spare time, and construction content data including multiple processes, the order of the multiple processes, the number of days required for each process, and the materials to be used in each of the multiple processes, and notifies the terminal device of the craftsman in charge of the schedule. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2020-197783 A Summary of the Invention [Problem to be solved by the invention]

[0005] In order to create a schedule, it is necessary to know various information such as the content of work to be done in each process, the order in which multiple processes are performed, and the period required to perform each process, and this requires a certain degree of knowledge and experience, etc. For this reason, it is not easy for an inexperienced person to create a schedule.

[0006] The present invention has been made in consideration of the above circumstances, and its purpose is to provide an information processing method, a method for generating a learning model, a computer program, and an information processing device that can be expected to assist in the creation of a schedule. [Means for solving the problem]

[0007] In one embodiment, an information processing method includes an information processing device that receives a process input, inputs the received process into a process prediction model that has been trained to output next process candidates in response to the process input, obtains next process candidates output by the process prediction model, and outputs the obtained next process candidates. Effect of the Invention

[0008] In one embodiment, it is expected to assist in the creation of a schedule. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a schematic diagram for explaining an overview of an information processing system according to a first embodiment. [Diagram 2] 2 is a block diagram showing a configuration of a server device according to the present embodiment. FIG. [Diagram 3] 2 is a block diagram showing a configuration of a terminal device according to the present embodiment; FIG. [Figure 4] FIG. 2 is a schematic diagram for explaining a configuration of a process prediction model according to the present embodiment. [Diagram 5] FIG. 2 is a schematic diagram for explaining a configuration of a process prediction model according to the present embodiment. [Figure 6] FIG. 11 is a schematic diagram for explaining generation of learning data by a server device. [Figure 7] 13 is a flowchart showing a procedure of a process prediction model generation process performed by the server device according to the present embodiment. [Figure 8] FIG. 2 is a schematic diagram for explaining a configuration of a term forecasting model according to the present embodiment. [Figure 9]13 is a flowchart showing the procedure of a period forecasting model generating process performed by the server device according to the present embodiment. [Figure 10] FIG. 13 is a schematic diagram showing an example of a process selection screen that accepts the selection of an initial process. [Figure 11] FIG. 13 is a schematic diagram showing an example of a period selection screen. [Figure 12] 13 is a schematic diagram showing an example of a process selection screen that accepts selection of a next process. FIG. [Figure 13] FIG. 13 is a schematic diagram showing an example of a display of a process chart. [Figure 14] 10 is a flowchart showing the procedure of a process chart creation process performed by the server device according to the present embodiment. [Figure 15] FIG. 11 is a schematic diagram showing an example of a period input screen according to the second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Specific examples of an information processing system according to an embodiment of the present invention will be described below with reference to the drawings. Note that the present invention is not limited to these examples, but is defined by the claims, and is intended to include all modifications within the meaning and scope of the claims.

[0011] <System configuration> FIG. 1 is a schematic diagram for explaining an overview of an information processing system according to a first embodiment. The information processing system according to this embodiment is a system that supports a user, such as an employee of a construction company, in creating a construction work schedule by using a learning model generated by machine learning, so-called AI (Artificial Intelligence). Note that in this embodiment, a construction work schedule will be described as an example, but the present invention is not limited to this, and the schedule that the information processing system according to this embodiment supports in creating may be various types of schedules, such as civil engineering work or manufacturing processes for industrial products.

[0012] The information processing system according to the present embodiment is configured to include a server device 1 that provides a service of supporting the creation of a schedule using AI, and one or more terminal devices 3 used by users who use this service. The server device 1 is a device managed and operated by a company providing the service, and communicates with one or more terminal devices 3 via a network such as the Internet, and performs various information processing to provide the above-mentioned service to the terminal devices 3. The terminal devices 3 are information processing devices such as personal computers used by users such as employees of a construction company, and may be portable information processing devices such as smartphones or tablet terminal devices.

[0013] In the information processing system according to the present embodiment, a user uses a terminal device 3 to perform an operation for creating a schedule. In this embodiment, a plurality of processes performed in a construction work are set in a chronological order, such that one process is performed before the next process is performed. For example, the terminal device 3 displays a list of process names that can be set, and the user performs an operation to select a first process from the plurality of processes displayed, and the terminal device 3 notifies the server device 1 of the process selected by the user. The server device 1 performs a process of predicting a process to be performed next to the process selected by the user using a learning model generated in advance by machine learning, and notifies the terminal device 3 of the predicted one or more processes as candidates for the next process. The terminal device 3 displays one or more candidates for the next process obtained from the server device 1, and accepts a selection of one candidate from the candidates for the next process from the user. The terminal device 3 notifies the server device 1 of the process selected by the user from the candidates for the next process, and the server device 1 predicts a process to be performed next to this process, and notifies the terminal device 3 of one or more candidates for the next process.

[0014] In this way, in the information processing system according to the present embodiment, the terminal device 3 accepts the selection of a process by the user, and the server device 1 repeats the process of predicting candidates for the next process for the selected process, thereby generating information on a plurality of processes to be performed from the first process to the last process, and the order in which the plurality of processes are to be performed. In the present embodiment, the information on the plurality of processes and the order in which they are performed is treated as an array in which, for example, the process names are arranged in the order in which they are performed, and this array is called a process sequence. However, the information on the plurality of processes and the order in which they are performed may be handled in any format. The server device 1 generates a process sequence, which is information on the plurality of processes and the order in which they are performed, creates a process chart such as a Gantt chart or a bar chart based on the generated process sequence, and transmits the created process chart to the terminal device 3.

[0015] In addition, in the information processing system according to the present embodiment, the server device 1 predicts the period required to perform the process. For example, in order to create a process schedule as a Gantt chart or a bar chart, the period of each process is necessary, and the server device 1 predicts the period of each process selected by the user and creates a process schedule based on the prediction result. Alternatively, the server device 1 may predict candidates for the next process for the process selected by the user, predict the period for each candidate, and notify the terminal device 3 of the predicted period of the process together with the candidate for the next process. Furthermore, the server device 1 may predict multiple candidates for the period of each process, and when the user selects the next process, the user may select the period of this process from the candidates, and the terminal device 3 may transmit the next process selected by the user and its period to the server device 1.

[0016] In the information processing system according to this embodiment, the server device 1 performs a process of predicting candidates for the next process using a learning model that has been generated in advance by machine learning. In this embodiment, the server device 1 performs a process of machine learning to generate a learning model that predicts candidates for the next process. Furthermore, the process of predicting the period of each process may be performed using a learning model, and in this case, the server device 1 performs a process of machine learning to generate a learning model that predicts the period of the process. In this embodiment, the server device 1 performs both the process of generating a learning model by machine learning and the prediction process using the generated learning model, but this is not limited to this, and the generation of the learning model may be performed by a device other than the server device 1.

[0017] <Device configuration> 2 is a block diagram showing the configuration of a server device 1 according to this embodiment. The server device 1 according to this embodiment is configured to include a processing unit 11, a memory unit (storage) 12, and a communication unit (transceiver) 13. Note that, in this embodiment, the processing will be described as being performed by one server device, but the processing may be distributed among a plurality of server devices.

[0018] The processing unit 11 is configured using an arithmetic processing device such as a CPU (Central Processing Unit), an MPU (Micro-Processing Unit), a GPU (Graphics Processing Unit) or a quantum processor, a ROM (Read Only Memory), a RAM (Random Access Memory), etc. The processing unit 11 reads out and executes a server program 12a stored in the storage unit 12, thereby performing various processes such as a process of predicting candidates for the next process of a process selected by a user, a process of predicting the period of each process, and a process of generating a learning model used for prediction.

[0019] The storage unit 12 is configured using a large-capacity storage device such as a hard disk. The storage unit 12 stores various programs executed by the processing unit 11 and various data required for the processing of the processing unit 11. In this embodiment, the storage unit 12 stores a server program 12a executed by the processing unit 11. The storage unit 12 also stores a process prediction model 12b and a period prediction model 12c generated by machine learning processing.

[0020] In this embodiment, the server program (computer program, program product) 12a is provided in a form recorded on a recording medium 99 such as a memory card or an optical disk, and the server device 1 reads the server program 12a from the recording medium 99 and stores it in the storage unit 12. However, the server program 12a may be written in the storage unit 12, for example, during the manufacturing stage of the server device 1. Also, for example, the server program 12a may be distributed by another remote server device or the like, and the server device 1 may acquire the server program 12a through communication. For example, the server program 12a may be read from the recording medium 99 by a writing device and written into the storage unit 12 of the server device 1. The server program 12a may be provided in a form distributed via a network, or may be provided in a form recorded on the recording medium 99.

[0021] The process prediction model 12b is a learning model that has been machine-learned in advance to receive a process input and output candidates for processes to be performed next after the process input. In this embodiment, the process prediction model 12b employs a learning model such as an RNN (Recurrent Neural Network), Transformer, or GPT (Generative Pre-Training)-2 that handles time-series information.

[0022] The period prediction model 12c is a learning model that has been machine-learned in advance to receive an input of a process and output a period required to perform the process. In the present embodiment, the period prediction model 12c may be a learning model that handles time-series information, such as RNN, Transformer, or GPT-2, similar to the process prediction model 12b. However, the period prediction model 12c may be a learning model such as a deep neural network (DNN), a support vector machine (SVM), or a random forest.

[0023] The information stored in the memory unit 12 as the process prediction model 12b and the period prediction model 12c may include configuration information indicating the configuration of these learning models, and information such as the values ​​of internal parameters of the learning models determined by machine learning processing.

[0024] The communication unit 13 communicates with various devices via a network N including a mobile phone communication network, a wireless LAN (Local Area Network), the Internet, etc. In this embodiment, the communication unit 13 communicates with one or more terminal devices 3 via the network N. The communication unit 13 transmits data provided from the processing unit 11 to other devices, and provides data received from other devices to the processing unit 11.

[0025] The storage unit 12 may be an external storage device connected to the server device 1. The server device 1 may be a multi-computer including a plurality of computers, or may be a virtual machine virtually constructed by software. The server device 1 is not limited to the above configuration, and may include, for example, a reading unit that reads information stored in a portable storage medium, an input unit that accepts operation input, or a display unit that displays images.

[0026] In addition, in the server device 1 according to this embodiment, the processing unit 11 reads and executes the server program 12a stored in the storage unit 12, whereby an input information acquisition unit 11a, a next process prediction unit 11b, a period prediction unit 11c, a prediction result output unit 11d, a process schedule creation unit 11e, a learning processing unit 11f, and the like are realized as software functional units in the processing unit 11. Note that in this figure, functional units related to the process of creating a process schedule and the process of generating a learning model are illustrated as the functional units of the processing unit 11, and functional units related to other processes are omitted from the illustration.

[0027] The input information acquisition unit 11a performs a process of acquiring information such as a process and a period input by a user via the terminal device 3. For example, the terminal device 3 displays a list of candidates for the next process predicted by the server device 1, accepts a selection input of one process from among these candidates from the user, and transmits information about the accepted process to the server device 1. The input information acquisition unit 11a communicates with the terminal device 3 via the communication unit 13 and acquires the process input by the user by receiving information about the process input by the user. Also, for example, the terminal device 3 displays a list of candidates for the period during which the process predicted by the server device 1 is to be performed, accepts a selection of one period from among these candidates from the user, and transmits information about the accepted period to the server device 1. The input information acquisition unit 11a acquires information about the period input by the user via the terminal device 3.

[0028] The next process prediction unit 11b performs a process to predict the process to be performed next to the process input by the user. In this embodiment, the next process prediction unit 11b predicts candidates for the next process using a process prediction model 12b that has been previously subjected to machine learning. The next process prediction unit 11b inputs the process acquired by the input information acquisition unit 11a to the process prediction model 12b, and acquires process candidates output by the process prediction model 12b in response to this input. The process prediction model 12b outputs the degree to which each process should be a candidate for the next process (probability, likelihood, confidence, etc.) for all processes that can be included in the process schedule, for example. The next process prediction unit 11b selects a predetermined number of processes in descending order of the degree to which each process should be a candidate for the next process, or selects a process whose degree to which each process should be a candidate for the next process exceeds a threshold, and sets the selected process as a candidate for the next process.

[0029] The period prediction unit 11c performs a process of predicting a period (e.g., number of days) for a process input by a user or a process predicted by the next process prediction unit 11b as a candidate for the next process. In this embodiment, the period prediction unit 11c predicts the period of the process using a period prediction model 12c that has been machine-learned in advance. The period prediction unit 11c inputs a process to be predicted (or a process sequence including the process to be predicted) to the period prediction model 12c, and obtains a prediction result of the period output by the period prediction model 12c. The period prediction model 12c may be configured to output a period (number of days) predicted to be required for the implementation of the process as a numerical value, or may be configured to output probabilities (likelihood, confidence, etc.) for multiple periods, such as X% probability that the period is one day, Y% probability that the period is two days, etc.

[0030] The prediction result output unit 11d performs processing to output the prediction result of the next process candidate by the next process prediction unit 11b and / or the prediction result of the process period by the period prediction unit 11c. The prediction result output unit 11d transmits (outputs) information on these prediction results to the terminal device 3, thereby causing the terminal device 3 to display (output) the prediction results.

[0031] After inputting, for example, the first process to the last process, the schedule creation unit 11e performs processing to create a schedule showing the inputted multiple processes in the order of execution. However, the schedule is not limited to the first process to the last process, but may be created from the first process to an intermediate process, from an intermediate process to the last process, or from an intermediate process to an intermediate process. The schedule creation unit 11e creates a schedule for construction work, for example, using a Gantt chart or a bar chart. The schedule creation unit 11e transmits (outputs) the created schedule to a terminal device, and causes the terminal device 3 to display (output) the schedule.

[0032] The learning processing unit 11f performs a process of generating the process prediction model 12b and the period prediction model 12c by machine learning. In the information processing system according to the present embodiment, information on the process of construction work carried out in the past is collected in advance, and the learning processing unit 11f of the server device 1 performs a machine learning process based on the collected information. For example, the learning processing unit 11f reads out the collected information and performs appropriate information processing (pre-processing) to generate data of a process sequence in which a plurality of processes carried out in the past construction work are arranged in the order of execution, and uses this data as learning data (so-called teacher data) for performing machine learning to generate the process prediction model 12b. For example, the learning processing unit 11f reads out the collected information and performs appropriate pre-processing to generate data in which the process carried out in the past construction work is associated with the period spent on this process, and uses this data as learning data for performing machine learning to generate the period prediction model 12c. For example, when a neural network learning model is used as the process prediction model 12b and the period prediction model 12c, the learning processing unit 11f can perform machine learning of the learning model by a method such as gradient descent, stochastic gradient descent, or error backpropagation. Since the machine learning process is an existing technique, detailed description thereof will be omitted. The learning processing unit 11f stores information such as internal parameters of the process prediction model 12b and the period prediction model 12c generated by the machine learning in the storage unit 12.

[0033] 3 is a block diagram showing the configuration of the terminal device 3 according to this embodiment. The terminal device 3 according to this embodiment is configured to include a processing unit 31, a memory unit (storage) 32, a communication unit (transceiver) 33, a display unit (display) 34, and an operation unit 35. The terminal device 3 is a device used by a user such as an employee of a construction company, and may be configured using an information processing device such as a personal computer, a smartphone, or a tablet terminal device.

[0034] The processing unit 31 is configured using an arithmetic processing device such as a CPU or an MPU, a ROM, etc. The processing unit 31 reads out and executes a program 32a stored in the storage unit 32, thereby performing various processes such as a process of accepting input of information such as a process and a period by a user, and a process of presenting information obtained from the server device 1 to the user.

[0035] The storage unit 32 is configured using a non-volatile memory element such as a flash memory or a storage device such as a hard disk. The storage unit 32 stores various programs executed by the processing unit 31 and various data required for the processing of the processing unit 31. In this embodiment, the storage unit 32 stores the program 32a executed by the processing unit 31. In this embodiment, the program 32a is distributed by a remote server device or the like, and the terminal device 3 acquires the program 32a through communication and stores it in the storage unit 32. However, the program 32a may be written in the storage unit 32, for example, during the manufacturing stage of the terminal device 3. For example, the program 32a may be read by the terminal device 3 from the program 32a recorded in a recording medium 98 such as a memory card or an optical disk and stored in the storage unit 32. For example, the program 32a may be read by a writing device from the program 32a recorded in the recording medium 98 and written in the storage unit 32 of the terminal device 3. The program 32a may be provided in a form of distribution via a network, or in a form of being recorded in the recording medium 98.

[0036] The communication unit 33 communicates with various devices via a network N including a mobile phone communication network, a wireless LAN, the Internet, etc. In this embodiment, the communication unit 33 communicates with the server device 1 via the network N. The communication unit 33 transmits data provided from the processing unit 31 to other devices, and provides data received from other devices to the processing unit 31.

[0037] The display unit 34 is configured using a liquid crystal display or the like, and displays various images, characters, and the like based on the processing of the processing unit 31. The operation unit 35 accepts user operations and notifies the processing unit 31 of the accepted operations. For example, the operation unit 35 accepts user operations through input devices such as mechanical buttons or a touch panel provided on the surface of the display unit 34. Furthermore, for example, the operation unit 35 may be input devices such as a mouse and a keyboard, and these input devices may be configured to be detachable from the terminal device 3.

[0038] In addition, in the terminal device 3 according to this embodiment, the processing unit 31 reads out and executes the program 32a stored in the storage unit 32, whereby the selection receiving unit 31a, the display processing unit 31b, and the like are realized as software functional units in the processing unit 31. Note that the program 32a may be a program dedicated to the information processing system according to this embodiment, or may be a general-purpose program such as an Internet browser or a web browser.

[0039] The selection receiving unit 31a performs a process of receiving a selection input of information such as a process and / or a period by a user. The selection receiving unit 31a receives the selection input of the process or period by receiving, for example, a selection of one item from among selection items such as a process or a period displayed in a list on the display unit 34 based on a user's operation on the operation unit 35. The selection receiving unit 31a transmits the received information on the process or period to the server device 1.

[0040] The display processing unit 31b performs processing to display various characters, images, and the like on the display unit 34 based on information transmitted from the server device 1. The display processing unit 31b displays, for example, a reception screen that receives input of a process, a period, and the like by a user. The display processing unit 31b also displays, for example, a process chart created by the server device 1.

[0041] <Process prediction model> 4 and 5 are schematic diagrams for explaining an example of the configuration of the process prediction model 12b according to the present embodiment. The process prediction model 12b according to the present embodiment is a learning model that receives an input of a process and outputs candidates for a process to be performed after the input process, as shown in FIG. 4, for example. For example, when there are N types of processes in total that can be handled in the information processing system according to the present embodiment, the process prediction model 12b outputs N numerical values. The N numerical values ​​output by the process prediction model 12b correspond to processes 1 to N, respectively, and the larger the numerical value, the more likely the corresponding process is to be the next process. A predetermined number of candidates for the next process can be obtained by selecting a predetermined number of numerical values ​​from the largest one among the multiple numerical values ​​output by the process prediction model 12b.

[0042] The input to the process prediction model 12b may be one process, or a process sequence in which multiple processes are arranged in the order of execution. When a process sequence is input to the process prediction model 12b, the process prediction model 12b outputs, for example, a candidate for a process to be performed next after the last process included in the input process sequence. However, the process prediction model 12b may also output, for example, a candidate for the next process for each process included in the process sequence, in which case the process prediction model 12b outputs M×N numerical values ​​for an input of a process sequence including M processes.

[0043] When a process sequence is input to the process prediction model 12b, for example, a configuration may be adopted in which a plurality of processes included in the process sequence are input to the process prediction model 12b all at once, or a configuration may be adopted in which a plurality of processes are input to the process prediction model 12b in sequence. FIG. 5 shows an example of inputting a plurality of processes one by one in sequence to the process prediction model 12b. When a first process is input to the process prediction model 12b, the process prediction model 12b outputs N numerical values ​​corresponding to processes 1 to N for a candidate for the second process. Next, when a second process is input to the process prediction model 12b, the process prediction model 12b outputs N numerical values ​​corresponding to processes 1 to N for a candidate for the third process. This is repeated in sequence, and for example, when a process sequence including M processes is used as input information, when the last M-th process is input to the process prediction model 12b, the process prediction model 12b outputs N numerical values ​​corresponding to processes 1 to N for a candidate for the M+1-th process. As a result, M×N numerical values ​​can be output for an input of a process sequence including M processes from the first to the Mth.

[0044] In this embodiment, a learning model such as RNN, Transformer, or GPT-2 that handles time-series information is adopted as the process prediction model 12b. The RNN learning model is a learning model in which a plurality of pieces of time-series information are input in sequence, and a state at the current time is output according to the state at the previous time and the input at the current time, and it is possible to predict the state at the current time taking into account the past input and state. When the RNN learning model is adopted as the process prediction model 12b, for example, a candidate for the second process is obtained by inputting the first process to the process prediction model 12b, and a candidate for the third process is obtained by selecting the second process from the candidates and inputting it to the process prediction model 12b, and by repeating this process, it is possible to predict the processes from the first process to the last process in order. Note that in this embodiment, the processes handled by the process prediction model 12b include an end process indicating the end (end) of all processes, and the process prediction model 12b may output this end process as a candidate for the next process.

[0045] The learning models of Transformer and GPT-2 are learning models that learn correlations of time-series information using a mechanism called attention. When Transformer or GPT-2 is adopted as the process prediction model 12b, for example, a process sequence in which one or more confirmed processes are arranged in the order of execution is input to the process prediction model 12b, and the process prediction model 12b outputs a candidate for the next process following the input process sequence. For example, a process sequence including only the first process is input to the process prediction model 12b to obtain a candidate for the second process, a process sequence in which the first process and the second process are arranged is input to the process prediction model 12b to obtain a candidate for the third process, a process sequence in which the third process is selected from the candidates and the first to third processes are arranged in this order is input to the process prediction model 12b to obtain a candidate for the fourth process, and so on. By repeating this process, it is possible to predict the processes from the first process to the last process in order.

[0046] In the information processing system according to the present embodiment, information on the processes carried out in construction works carried out in the past (i.e., information on past schedules) is stored in advance, and the server device 1 uses the stored information to generate a schedule prediction model 12b through machine learning processing. The information on the past schedules is converted into information in a table format, such as CSV (Comma Separated Value) or TSV (Tab Separated Values), and stored in a database or the like. The server device 1 acquires the stored table format information, creates information on a sequence of processes based on the acquired information, and uses the created information on the sequence of processes as learning data for machine learning.

[0047] 6 is a schematic diagram for explaining the generation of learning data by the server device 1, and shows an example of information obtained by converting information on a past schedule into a table format. In this example, information on "process name," "start date," "end date," "actual work days," "start line," and "end line" is stored in association with each other for each process carried out in a construction project. This example is for processes carried out in one construction project, and this construction project includes nine types of processes, from process A to process I.

[0048] "Process name" is the name given to the process, and is set to a string of characters such as foundation work, piping work, exterior work, and interior work. In the illustrated example, the process names are abbreviated to Process A to Process I. "Start date" is the date the process started, and "End date" is the date the process ended. "Number of actual working days" is the number of days required to carry out the process, and is set to a number equal to or less than the number of days from the start date to the end date.

[0049] In this embodiment, the schedule is a table with the horizontal axis representing time and multiple rows in the vertical direction, and the order and number of days of multiple processes carried out in a construction project are indicated by entering a start point indicating one process and symbols such as an arrow in each row or across multiple rows. A "start row" is a row in which a start point indicating the timing at which the process began is entered, and an "end row" is a row in which the end point of the arrow extending from the start point is entered. The start point of the next process is entered at the end point of the arrow, and the order of the processes can be determined by tracing the multiple arrows from the first process, and the number of days required to carry out a process can be determined from the horizontal length of the arrow.

[0050] In the illustrated example, process A is the first process. The server device 1 according to this embodiment acquires, for example, the process with the earliest "start date" in the table-formatted information as the first process. Next, the server device 1 searches this information for the second process following the first process A. At this time, the server device 1 determines that the process whose start date matches the end date of process A and whose start row matches the end row of process A is the process performed after process A. In this example, process B is performed after process A. Similarly, by searching for the processes performed after process B in order, it is found that they were performed in the following order: process A, process B, process C, process D, and process E.

[0051] The end date of process E is 2021 / 10 / 22, and the same start date is set for two processes, process F and process G. However, since the end row of process E is 2, the start row of process F is 4, and the start row of process G is 2, it can be seen that the next process after process E is process G. Note that in this example, process G has 0 working days and is not an actually performed process, but a dummy process inserted to create a schedule. For this reason, process G may be excluded from the process sequence that the server device 1 uses as learning data.

[0052] Similarly, by searching the processes performed after process G in order, it is found that the processes were performed in the order of process A, process B, process C, process D, process E, process G, process F, process I, and process H. Based on this result, the server device 1 deletes process G, which is a dummy process, and adds a final process indicating the end of the process to create a process sequence of [process A, process B, process C, process D, process E, process F, process I, process H, final process]. Note that a series of processes may include, for example, a case where the process branches into multiple processes midway, or a case where multiple processes merge into one. In the above example, when process C branches into processes D and J, the server device 1 can create two process sequences: the process sequence of [process A, process B, process C, process D, process E, process F, process I, process H, final process] and the process sequence of [process A, process B, process C, process J, ..., final process].

[0053] The server device 1 creates a process sequence based on the information of the accumulated process chart, and generates a process prediction model 12b by performing machine learning using the information of the created process sequence as learning data. As shown in FIG. 4, the process prediction model 12b according to this embodiment is configured to accept the input of a process and output a candidate for the next process of the input process. The server device 1 uses the process sequence created as learning data, creates data in which one process included in the process sequence is used as input information and the next process of the process is used as output information (correct answer value), and can use this data as so-called teacher data for supervised machine learning processing. For example, the server device 1 can create three data, data of process A and process B, data of process B and process C, and data of process C and the final process, from a process sequence of [process A, process B, process C, final process], and use them for machine learning.

[0054] In addition, when the input to the process prediction model 12b is a process sequence, for example, from a process sequence of [process A, process B, process C, final process], the server device 1 can create three pieces of data for use in machine learning: data for the input process sequence being [process A] and the correct value being process B, data for [process A, process B] and process C, and data for [process A, process B, process C] and the final process.

[0055] The server device 1 performs supervised machine learning processing using the data, which is generated by the above procedure and associates a process with the process that follows this process, to determine internal parameters of the process prediction model 12b and generate the process prediction model 12b. The server device 1 performs machine learning of the learning model using a method such as gradient descent, stochastic gradient descent, or backpropagation. The server device 1 stores the process prediction model 12b generated by this machine learning processing in the storage unit 12, and can perform subsequent process prediction processing using the process prediction model 12b stored in the storage unit 12.

[0056] FIG. 7 is a flowchart showing the procedure of the process prediction model generation process performed by the server device 1 according to the present embodiment. The learning processing unit 11f of the processing unit 11 of the server device 1 according to the present embodiment reads out information of a past process schedule stored in the storage unit 12, for example (step S1). The learning processing unit 11f generates information of a process sequence in which a plurality of processes included in the process schedule are arranged in the order of execution, based on information such as the "start date", "end date", "start row", and "end row" of each process included in the read process schedule (step S2). The learning processing unit 11f judges whether the above process has been completed for all the stored process schedule information (step S3). If the process has not been completed for all the information (S3: NO), the learning processing unit 11f returns the process to step S1, reads out the next information, and repeats the same process.

[0057] When the processing is completed for all information (S3: YES), the learning processing unit 11f further generates data that associates one process with the process following this process from the generated information of the process sequence (step S4). The learning processing unit 11f performs machine learning processing using the process-next process data generated in step S4 as teacher data (step S5) to generate a process prediction model 12b. The learning processing unit 11f stores information such as internal parameters of the generated process prediction model 12b in the storage unit 12 (step S6) and ends the processing.

[0058] <Period forecast model> FIG. 8 is a schematic diagram for explaining the configuration of the period prediction model 12c according to the present embodiment. The process prediction model 12b according to the present embodiment is a learning model that receives, for example, a process input and outputs the number of days required to perform the input process. For example, in the information processing system according to the present embodiment, when the maximum number of days required to perform one process is N days, the period prediction model 12c outputs N numerical values. The N numerical values ​​output by the period prediction model 12c correspond to the number of days to perform the process being 1 day, 2 days, ..., N days, respectively, and the larger the numerical value, the more likely the corresponding number of days is the number of days required to perform the input process. Note that, in the present embodiment, a configuration that outputs a numerical value for each number of days (a so-called classification model configuration) is adopted as the period prediction model 12c, but this is not limited thereto, and a configuration that outputs the number of days required to perform the process as one numerical value (a so-called regression model configuration) may be adopted.

[0059] The input to the period prediction model 12c may be a single process, as in the case of the input to the process prediction model 12b, or may be a process sequence in which multiple processes are arranged in the order of execution. When a process sequence is input to the period prediction model 12c, the period prediction model 12c outputs, for example, the number of days required to execute the last process included in the input process sequence. However, the period prediction model 12c may also output, for example, the number of days required to execute each process included in the process sequence, in which case the period prediction model 12c outputs M×N numerical values ​​for the input of a process sequence including M processes.

[0060] In this embodiment, as for the period prediction model 12c, similarly to the process prediction model 12b, a learning model such as RNN, Transformer, or GPT-2 may be adopted; however, this is not limited thereto, and a learning model such as DNN, SVM, or random forest may also be adopted.

[0061] In order to generate the period prediction model 12c by machine learning processing, the server device 1 generates learning data based on information on a process schedule stored in advance. The information on the process schedule used at this time may be the same as the information (see FIG. 6) used when generating the learning data for the process prediction model 12b. For example, the server device 1 generates data in which the process name is associated with the number of actual working days from the information shown in FIG. 6, and uses this data as learning data. Alternatively, the server device 1 may generate a process sequence using a procedure similar to that used to create the learning data for the process prediction model 12b, and use data in which the process sequence is associated with the number of actual working days for each process included in the process sequence as learning data.

[0062] The server device 1 creates learning data based on the accumulated information on the process chart, and generates the period prediction model 12c by performing machine learning using the created learning data. As shown in FIG. 8, the period prediction model 12c according to the present embodiment is configured to accept the input of a process and output the number of days required to perform the input process. The server device 1 uses data in which the process and the number of days are associated with each other, created as learning data, and performs supervised machine learning processing with one process included in this data as input information and the number of days of this process as output information (correct answer value), thereby determining internal parameters of the period prediction model 12c and generating the period prediction model 12c. The server device 1 performs machine learning of the learning model by a method such as gradient descent, stochastic gradient descent, or backpropagation. The server device 1 stores the period prediction model 12c generated by this machine learning processing in the storage unit 12, and can perform subsequent period prediction processing using the period prediction model 12c stored in the storage unit 12.

[0063] 9 is a flowchart showing the procedure of the period forecasting model generation process performed by the server device 1 according to the present embodiment. The learning processing unit 11f of the processing unit 11 of the server device 1 according to the present embodiment reads out information of a past schedule stored in the storage unit 12, for example (step S21). The learning processing unit 11f generates data in which a process included in the read schedule is associated with the number of actual working days of the process (step S22). The learning processing unit 11f judges whether the generation of the above data has been completed for all the stored schedule information (step S23). If the generation of data has not been completed for all the information (S23: NO), the learning processing unit 11f returns the process to step S21, reads out the next information, and repeats the same process.

[0064] When data generation for all information is completed (S23: YES), the learning processing unit 11f performs machine learning processing using the generated process-number-of-days data as teacher data (step S24) to generate a period prediction model 12c. The learning processing unit 11f stores information such as internal parameters of the generated period prediction model 12c in the storage unit 12 (step S25) and ends the process.

[0065] <Process chart creation process> The server device 1 according to the present embodiment performs a process of creating a process schedule using the created process prediction model 12b and period prediction model 12c. At this time, the server device 1 performs processes such as displaying a screen for accepting selection of a process and period, accepting the selection of the process and period by the user, displaying information such as the next process and period predicted by the learning model, and displaying the created process schedule by exchanging information with the terminal device 3 used by the user.

[0066] FIG. 10 is a schematic diagram showing an example of a process selection screen for accepting the selection of the first process. In the process schedule creation process, the server device 1 first displays a process selection screen for accepting the selection of the first process from the user on the display unit 34 of the terminal device 3. The process selection screen shown in FIG. 10 is a process schedule in which a dialog box for process selection is superimposed on a blank process schedule. In the process selection dialog box, for example, a title character string of "Process Selection" is displayed at the top, and a message prompting selection such as "Please select the first process" is displayed below it. The server device 1 causes the terminal device 3 to display a list of a plurality of buttons with selectable process names attached as labels arranged vertically below this message in the process selection dialog box. In the illustrated example, five processes, Process A to Process E, are displayed as options, and the user can select one of the processes by performing a click operation, touch operation, or the like on any one of the options using the operation unit 35 of the terminal device 3.

[0067] In this embodiment, the process to be presented to the user as an option for the first process is predetermined. The server device 1 transmits information such as the name of one or more processes that can be the predetermined first process to the terminal device 3, and the terminal device 3 that receives this information displays the options for the predetermined first process on the process selection screen. The terminal device 3 accepts a process selection operation by the user on the process selection screen, and transmits information related to the process whose selection has been accepted to the server device 1.

[0068] The server device 1, which has received information about a process selected by a user from the terminal device 3, predicts the period required to perform the selected process using the period prediction model 12c. The server device 1 inputs the selected process (or a process sequence including the selected process as the last process) to the period prediction model 12c, and obtains period information output by the period prediction model 12c. In this embodiment, the period prediction model 12c outputs period information for this process (or the last process in the process sequence) in response to the input of the process (or process sequence), and the period information is N numbers corresponding to 1 to N days, and each number indicates the probability of the corresponding period. The server device 1 obtains, for example, five periods corresponding to the top five numbers from the N numbers, and transmits the obtained five periods and the numbers output by the period prediction model 12c for each period to the terminal device 3.

[0069] The terminal device 3, which has received information on the period required to perform the process from the server device 1, displays a period selection screen on the display unit 34 to accept the selection of a period from the user. FIG. 11 is a schematic diagram showing an example of the period selection screen. The illustrated period selection screen is displayed by superimposing a dialogue box for period selection on a process chart in which no processes are entered. In the period selection dialogue box, for example, a title character string "Period Selection" is displayed at the top, and a message urging a selection such as "Please select the period for the selected process A" is displayed below that.

[0070] Based on the information regarding the period received from the server device 1, the terminal device 3 displays a list of multiple buttons labeled with selectable periods in a vertical direction below a message prompting the user to select, and displays the predicted probability for each period to the right of each button. In the illustrated example, five options from 2 to 6 days are displayed in order from the shortest period to the largest, but this is not limiting, and the options may be displayed in order from the highest probability to the largest. The user can select one of the periods by using the operation unit 35 of the terminal device 3 to perform a click operation or touch operation on one of the options. The terminal device 3 accepts the user's selection of the period on the period selection screen, and transmits information regarding the period for which the selection was accepted to the server device 1.

[0071] The server device 1, which has received information on the period selected by the user from the terminal device 3, stores the process selected on the process selection screen in FIG. 10 in association with the period selected on the period selection screen in FIG. 11. Next, the server device 1 performs a process of predicting the next process of the process selected on the process selection screen using the process prediction model 12b. The server device 1 inputs the selected process (or a process sequence including the selected process as the last process) to the process prediction model 12b, and obtains information on the candidate of the next process output by the process prediction model 12b. In this embodiment, the process prediction model 12b outputs information on the candidate of the process to be performed next to this process (or the last process of the process sequence) in response to the input of the process (or the process sequence), and the output is N numerical values ​​corresponding to the total number of processes N, and each numerical value indicates the probability of the corresponding process. The server device 1 obtains, for example, five processes corresponding to the top five numerical values ​​from the N numerical values, and sets them as candidates of the next process. The server device 1 transmits the five candidates of the next process and the numerical values ​​output by the process prediction model 12b for each candidate process to the terminal device 3.

[0072] The terminal device 3, which has received information about candidates for the next process from the server device 1, displays a process selection screen on the display unit 34 for accepting the selection of the next process from the user. FIG. 12 is a schematic diagram showing an example of a process selection screen for accepting the selection of the next process. The process selection screen for the next process shown in FIG. 12 has substantially the same configuration as the process selection screen for the first process shown in FIG. 10, but the message displayed below the title character string of "Process Selection" has been changed to a message suitable for selecting the next process, for example, "Please select the process following process A." In addition, the process selection screen for the next process displays the probability that each process will be the next process in association with multiple process names displayed as a list of options.

[0073] The terminal device 3 displays a list of process names selected as candidates for the next process as options on a process selection screen based on the information on the candidates for the next process received from the server device 1, and displays the probability of each process based on the information on the numerical values ​​output by the process prediction model 12b included in the received information. The terminal device 3 accepts a selection operation of the next process by the user on the process selection screen, and transmits information on the process whose selection has been accepted to the server device 1.

[0074] The server device 1, which has received information on the next process selected by the user from the terminal device 3, predicts the period required to perform this process using the period prediction model 12c and transmits the prediction result to the terminal device 3. The terminal device 3 displays the period selection screen shown in Fig. 11 based on the period prediction result transmitted from the server device 1, and receives and transmits the period selection from the user to the server device 1. After that, the selection of the process and the period are repeated until the user selects an end process indicating the end of the entire process as the next process.

[0075] When the user selects an end process, the server device 1 performs processing to create a schedule based on information about the processes and periods selected up to that point. For example, for multiple selected processes, the server device 1 indicates each process with an arrow of a length corresponding to the period, and connects multiple arrows corresponding to the multiple processes in the order selected by the user. The server device 1 creates a schedule by drawing the connected arrows over a table with the horizontal axis representing dates. The server device 1 transmits information about the created schedule to the terminal device 3, and the terminal device 3 displays the schedule created by the server device 1 on the display unit 34.

[0076] FIG. 13 is a schematic diagram showing an example of a display of a schedule. The schedule shown in the figure is a schedule for a certain construction work from Monday, December 1st to Sunday, December 14th, 2021. This schedule has work categories (types or classifications of work, etc.) such as "temporary", "exterior", and "interior", and a three-line area is provided for each category to describe the process. In this example, the user has finished inputting the process for the "temporary" work category. The user selects the processes to be performed in the order of process A, process B, process C, process D, and process E, and selects the duration as 3 days for process A, 3 days for process B, 2 days for process C, 1 day for process D, and 4 days for process E.

[0077] The server device 1 accepts the above-mentioned selection of processes and periods for each work division from the user, creates a schedule in which multiple processes are linked by arrows in the order of execution, and finally creates a schedule in which multiple processes are linked for each of the multiple work divisions. Note that in this example, the server device 1 creates a schedule divided into work divisions such as "temporary," "exterior," and "interior," but these work divisions are merely examples and are not limited to these. The server device 1 may also create a schedule divided into building sections (places where work is performed) such as "south side of the building" and "north side of the building" instead of work divisions, and may further create a schedule divided by a combination of work divisions and building sections.

[0078] In this embodiment, information necessary for the terminal device 3 to display the process schedule, such as dates, task categories, and their arrangement, is transmitted from the server device 1 to the terminal device 3 together with information for displaying arrows indicating a series of processes. However, the information necessary for displaying the process schedule may be stored in advance in the terminal device 3. Also, the server device 1 may predict the processes and periods, and the terminal device 3 may perform the process of creating the process schedule.

[0079] 14 is a flow chart showing the procedure of the process schedule creation process performed by the server device 1 according to the present embodiment. The input information acquisition unit 11a of the processing unit 11 of the server device 1 according to the present embodiment causes the terminal device 3 to display an initial process selection screen (step S41). The input information acquisition unit 11a receives information transmitted from the terminal device 3 at the communication unit 13, thereby accepting the selection of an initial process by the user via the terminal device 3 (step S42).

[0080] Next, the period prediction unit 11c of the processing unit 11 performs a process of predicting the period of the process selected by the user (step S43). At this time, the period prediction unit 11c inputs the process selected by the user (or a process sequence in which a plurality of processes selected by the user are arranged in chronological order) to the period prediction model 12c, and obtains the period prediction result output by the period prediction model 12c, thereby predicting the period. The prediction result output unit 11d of the processing unit 11 transmits the period prediction result by the period prediction unit 11c to the terminal device 3, thereby causing the display unit 34 of the terminal device 3 to display a period selection screen (step S44). The input information acquisition unit 11a accepts the selection of the period by the user via the terminal device 3 (step S45).

[0081] Next, the next process prediction unit 11b of the processing unit 11 performs a process of predicting the next process of the process selected by the user (step S46). At this time, the next process prediction unit 11b inputs the process selected by the user in step S42 or S48 (or a process sequence in which a plurality of processes selected by the user are arranged in chronological order) to the process prediction model 12b, and obtains candidates for the next process output by the process prediction model 12b, thereby predicting the next process. The prediction result output unit 11d transmits the prediction result of the next process by the next process prediction unit 11b to the terminal device 3, thereby displaying a process selection screen for the next process on the display unit 34 of the terminal device 3 (step S47). The input information acquisition unit 11a accepts the selection of the next process by the user via the terminal device 3 (step S48). The input information acquisition unit 11a determines whether the process selected by the user is an end process indicating the end of all processes (step S49). If the selected step is not the end step (S49: NO), the processing unit 11 returns the process to step S43, and repeats the processes of steps S43 to S49.

[0082] If the selected process is an end process (S49: YES), the process schedule creation unit 11e of the processing unit 11 creates a process schedule as shown in Fig. 13 based on the process and period information received so far (step S50). The process schedule creation unit 11e transmits the information of the created process schedule to the terminal device 3, thereby causing the terminal device 3 to display the process schedule (step S51), and ends the process.

[0083] <Summary> In the information processing system according to the present embodiment configured as above, the server device 1 accepts process input by the user via the terminal device 3, predicts candidates for the next process using the process prediction model 12b, and transmits (outputs) the predicted candidates for the next process to the terminal device 3, thereby displaying (outputting) them on the display unit 34 of the terminal device 3. This allows the user to easily grasp the process to be performed next to the input process, so the information processing system according to the present embodiment is expected to support the user in creating a process schedule.

[0084] In the information processing system according to the present embodiment, the server device transmits a plurality of candidates for the next process to the terminal device 3 based on the output of the process prediction model 12b, accepts a selection input of one process from the plurality of candidates via the terminal device 3, and further predicts a candidate for the next process of the selected process. The output of the process prediction model 12b includes an end process indicating the end of all processes, and the server device 1 repeatedly predicts the candidate for the next process and accepts the selection of the next process until the user selects this end process. By repeatedly predicting the candidate for the next process and accepting the selection of the next process, the server device 1 can obtain information on a series of processes performed in chronological order from the user. Since the user can select the next process from limited and highly reliable candidates, it is expected that the selection of the next process will be made easier than when selecting the next process from a large number of processes.

[0085] Furthermore, in the information processing system according to this embodiment, the server device 1 predicts the period required for performing a process using the period prediction model 12c, and transmits (outputs) the predicted period to the terminal device 3, thereby displaying (outputting) the period of the process on the display unit 34 of the terminal device 3. In response to the input of a process or a sequence of processes, the period prediction model 12c outputs the period required for performing the input process or the final process of the sequence of processes. This allows the user to easily grasp how long each process of a construction work or the like will take, so the information processing system according to this embodiment is expected to assist the user in creating a schedule.

[0086] Furthermore, in the information processing system according to this embodiment, the server device 1 predicts the period required for carrying out a process and transmits multiple candidates to the terminal device 3, and the terminal device 3 displays multiple candidates for the period. This allows the user to determine and input an appropriate period from the multiple candidates, so the information processing system according to this embodiment is expected to support the user in creating a process schedule.

[0087] Furthermore, in the information processing system according to this embodiment, the server device 1 accepts input of multiple processes and input of the duration of each process from the user via the terminal device 3, creates a schedule showing the multiple processes in the order of execution based on the accepted processes and durations, transmits (outputs) the created schedule to the terminal device 3, and causes the display unit 34 of the terminal device 3 to display (output) the schedule. The server device 1 may also accept input of processes and durations from the user for each of multiple work sections or building blocks, and create a schedule showing the order of execution of multiple processes for each of multiple work sections or building blocks. As a result, the information processing system according to this embodiment can automatically create and output a schedule based on information entered by the user, and is expected to assist users in creating schedules.

[0088] In the information processing system according to the present embodiment, the server device 1 acquires, as learning data, information on a sequence of processes, including information on a plurality of processes and the order in which these processes are performed, and performs machine learning using the acquired learning data to generate a process prediction model 12b that outputs candidates for the next process for a process input. This allows the server device 1 to use the generated process prediction model 12b to predict and output candidates for the next process for a process selected by the user, which is expected to support the user in creating a process schedule.

[0089] Furthermore, in the information processing system according to this embodiment, the server device 1 acquires data in which a plurality of processes are associated with the start and end times of each process, generates learning data for a process sequence in which a plurality of processes are arranged in the order of execution based on this data, and performs machine learning of the process prediction model 12b. This allows the server device 1 to generate learning data based on information such as a schedule of construction work or the like carried out in the past, and perform machine learning of the process prediction model 12b using the generated learning data.

[0090] In the information processing system according to the present embodiment, the server device 1 acquires learning data including a plurality of processes and the implementation period of each process, and performs machine learning using the acquired learning data to generate a period prediction model 12c that outputs a period required for the implementation of each process when a process is input. Alternatively, the server device 1 acquires learning data including a plurality of processes, information regarding the implementation order of the plurality of processes, and the implementation period of each process, and performs machine learning using the acquired learning data to generate a period prediction model 12c that outputs a period required for the implementation of each process when a process is input. As a result, the server device 1 can predict and output a period required for the implementation of a process selected by the user using the generated period prediction model 12c, which is expected to support the creation of a process schedule by the user.

[0091] In this embodiment, the creation of a construction work schedule has been described as an example, but the schedule handled by the information processing system is not limited to construction work, and may be, for example, a schedule for a civil engineering work or a product manufacturing process, or any other schedule. Also, the screen displays shown in Figures 10 to 12 in this embodiment are merely examples and are not limiting, and the screen displays may be displayed in any manner.

[0092] <Embodiment 2> In the information processing system according to the second embodiment, the server device 1 does not predict a period using the period prediction model 12c, but provides the user with information about the period based on information about processes performed in the past, and accepts input of the period by the user. After accepting selection of a process from the user on the process selection screen shown in Fig. 10 or 12, the server device 1 according to the second embodiment displays on the display unit 34 of the terminal device 3 a period input screen for accepting input of the implementation period of the selected process.

[0093] FIG. 15 is a schematic diagram showing an example of a period input screen according to the second embodiment. The period input screen displayed by the terminal device 3 in the information processing system according to the second embodiment is a screen in which a dialogue box for inputting a period is superimposed on a process chart in which no process is entered. In the dialogue box for inputting a period, for example, a title character string of "Period Input" is displayed at the top, and a message prompting input such as "Please input the period of the selected process A" is displayed below it. Below this message, the terminal device 3 displays statistics such as "average number of days," "maximum number of days," and "minimum number of days" based on the past history of the process A selected by the user. In addition, the terminal device 3 displays an input box for receiving a numerical input of the period of the process A and a button labeled "OK" below these displays. The user can input the period by using the operation unit 35 of the terminal device 3 to input the number of days required for the implementation of the process A in the input box and performing a click operation or touch operation on the OK button. The terminal device 3 transmits information regarding the input period to the server device 1, and the server device 1, upon receiving this, stores the period and transmits candidates for the next process to the terminal device 3. After this, the user repeatedly selects candidates and inputs periods, and when the final process is selected, the server device 1 creates a process table.

[0094] In order for the terminal device 3 to display the illustrated period input screen, the server device 1 according to the second embodiment calculates statistical values ​​such as the average number of days, maximum number of days, and minimum number of days for the process selected by the user, and transmits the calculation results to the terminal device 3. At this time, the server device 1 extracts all the actual work days of the target process from the information of the past process schedule stored in advance, for example the information shown in FIG. 6, and calculates the average, maximum, and minimum values ​​of the extracted actual work days. The terminal device 3 receives this information transmitted from the server device 1, and displays the average number of days, maximum number of days, and minimum number of days for the process on the period input screen. Note that in this embodiment, the terminal device 3 displays the average number of days, maximum number of days, and minimum number of days for the past period, but this is not limited thereto, and any other statistical values ​​may be displayed.

[0095] Furthermore, the terminal device 3 may generate a histogram for the number of working days extracted for the target process, with the horizontal axis representing the number of days and the vertical axis representing the frequency, and display the generated histogram as a statistical value on the display unit 34. Furthermore, the terminal device 3 may display statistical values ​​such as the average value, maximum value, and minimum value described above along with the display of the histogram. Note that the terminal device 3 may display various graphs related to the number of working days for a process, not limited to histograms.

[0096] Furthermore, in the information processing system according to the second embodiment, the terminal device 3 accepts operations such as clicking or touching on the maximum number of days or the minimum number of days displayed on the period input screen. For example, when an operation on the maximum number of days is performed, the terminal device 3 notifies the server device 1 of the operation, the server device 1 transmits information on a process schedule including the process carried out in the maximum number of days to the terminal device 3, and the terminal device 3 displays the process schedule on the display unit 34. When there are multiple process schedules including a process carried out in the maximum number of days, the server device 1 may, for example, select the most recent process schedule and transmit it to the terminal device 3, or may display a list of the titles or dates of the corresponding process schedules on the terminal device 3 and accept a selection from the user to display the process schedule.

[0097] In the information processing system according to the second embodiment configured as above, the server device 1 calculates at least one statistical value of the average number of days, the maximum number of days, or the minimum number of days in the past for a process selected by a user based on the schedule of a construction work previously carried out, and transmits (outputs) the calculated value to the terminal device 3, and causes the terminal device 3 to display (output) these statistical values. This allows the user to know from the past performance how long it will take to carry out the process selected by the user, so the information processing system according to this embodiment is expected to support the user in creating a schedule.

[0098] Other configurations of the information processing system according to the second embodiment are similar to those of the information processing system according to the first embodiment, so the same reference numerals are used for the same parts and detailed description will be omitted.

[0099] The embodiments disclosed herein are illustrative in all respects and should not be considered as limiting. The scope of the present invention is defined by the claims, not by the above meaning, and is intended to include all modifications within the scope and meaning equivalent to the claims. [Explanation of symbols]

[0100] 1. Server device 3 Terminal Equipment 11 Processing section 11a Input information acquisition unit 11b Next Process Prediction Department 11c Period Prediction Section 11d Prediction result output section 11e Schedule Creation Department 11f Learning processing section 12 Storage section 12a Server Program 12b Process Prediction Model 12c Period Forecast Model 13. Communications Department 31 Processing section 31a Selection Reception Section 31b Display processing unit 32 Storage section 32a Program 33 Communications Department 34 Display section 35 Control section 98,99 Recording media N Network

Claims

1. An information processing device, Accepts process input, inputting the received process into a process prediction model that has been trained to output candidates for a next process in response to a process input, and acquiring a plurality of candidates for the next process output by the process prediction model; outputting the obtained candidates for the next process; Accept the selection of one process from multiple candidates, Inputting the received process into the process prediction model and acquiring a candidate for a next process output by the process prediction model; Outputting the obtained candidates for the next process; the candidates output by the process prediction model include the end of the process, and output of candidates for the next process by the process prediction model is repeated until a selection input for the end of the process is received from among the multiple output candidates; Information processing methods.

2. The information processing device, Obtaining a plurality of stored information regarding the duration of previously performed steps; Calculating statistics regarding the period based on the acquired information; outputting the calculated statistical value; The information processing method according to claim 1 .

3. the information processing device calculates, as the statistical value, at least one of an average value, a maximum value, or a minimum value of the period required to perform the process; The information processing method according to claim 2 .

4. The information processing device, inputting the received process into a duration prediction model that has been trained to output a period required for carrying out the process in response to an input of the process, and acquiring a period output by the duration prediction model; Output the acquired period. The information processing method according to claim 1 .

5. The information processing device, inputting the received process into a duration prediction model that has been trained to output a period required for performing a last process included in a plurality of processes that are sequentially performed, and acquiring a period output by the duration prediction model; Output the acquired period. The information processing method according to claim 1 .

6. The information processing device outputs a plurality of candidates for the period required for carrying out the process.

6. An information processing method according to claim 2, wherein the information processing method comprises:

7. The information processing device, Accept input of multiple processes to be carried out, Accepts input of duration for each process, Based on the received process and period, a schedule is created showing the order in which multiple processes are carried out. Output the created schedule.

7. An information processing method according to claim 1.

8. The information processing device, Accepting input of a plurality of processes and a period of each process for each of a plurality of work divisions or a plurality of building sections; Creating a schedule showing the steps in the order of execution for each of the plurality of work divisions or each of the plurality of building sections; The information processing method according to claim 7.

9. An information processing device, Acquire learning data including a plurality of steps and information regarding the order in which the steps are performed; generating a process prediction model that outputs candidates for a process to be performed next in response to a process input by machine learning using the acquired learning data; How to generate a learning model.

10. The information processing device, Acquire data in which a plurality of processes are associated with the start and end times of each process; Based on the acquired data, a plurality of steps are arranged in an order of execution to generate learning data; generating the process prediction model using the generated learning data; The method for generating a learning model according to claim 9 .

11. An information processing device, Acquire learning data including a plurality of processes and an implementation period of each process; generating a time period prediction model that outputs a time period required for performing a process in response to an input of a process, by machine learning using the acquired learning data; How to generate a learning model.

12. An information processing device, Acquire learning data including a plurality of steps, information regarding the order in which the steps are performed, and a period during which each step is performed; generating a time period prediction model that outputs a time period required for performing a final process among a plurality of processes that are sequentially performed in response to an input of the plurality of processes through machine learning using the acquired learning data; How to generate a learning model.

13. On the computer, Accepts process input, inputting the received process into a process prediction model that has been trained to output candidates for a next process in response to a process input, and acquiring a plurality of candidates for the next process output by the process prediction model; outputting the obtained candidates for the next process; Accept the selection of one process from multiple candidates, Inputting the received process into the process prediction model and acquiring a candidate for a next process output by the process prediction model; Outputting the obtained candidates for the next process; The candidates output by the process prediction model include the end of the process, and the output of candidates for the next process by the process prediction model is repeated until a selection input for the end of the process is received from among the multiple output candidates. A computer program that executes a process.

14. On the computer, Acquire learning data including a plurality of processes and information regarding the order of execution of the plurality of processes, and generate a process prediction model that outputs candidates for the next process to be executed in response to a process input by machine learning using the acquired learning data. A computer program that executes a process.

15. A processing unit is provided, The processing unit, Accepts process input, inputting the received process into a process prediction model that has been trained to output candidates for a next process in response to a process input, and acquiring a plurality of candidates for the next process output by the process prediction model; outputting the obtained candidates for the next process; Accept the selection of one process from multiple candidates, Inputting the received process into the process prediction model and acquiring a candidate for a next process output by the process prediction model; Outputting the obtained candidates for the next process; the candidates output by the process prediction model include the end of the process, and output of candidates for the next process by the process prediction model is repeated until a selection input for the end of the process is received from among the multiple output candidates; Information processing device.

16. An acquisition unit that acquires learning data including a plurality of steps and information regarding an order in which the plurality of steps are performed; a generation unit that generates a process prediction model that outputs candidates for a process to be performed next in response to an input of a process by machine learning using the acquired learning data; An information processing device comprising:

Images