JP7087256B2 - Construction system ledger inspection support device and construction system ledger inspection support model learning device - Google Patents

Description

The present invention relates to a construction system ledger inspection support device and a construction system ledger inspection support model learning device.

Conventionally, a system for managing a construction system at a construction site is known. For example, when a construction system consisting of cooperating companies is constructed at a work site such as a construction site, a management system that assists labor management at the work site is known (for example, Patent Document 1).

Japanese Unexamined Patent Publication No. 2005-242849

In the management system described in Patent Document 1, the work of creating various safety documents, construction system ledgers, etc. at the work place is assisted.

However, the technique described in Patent Document 1 does not consider the inspection of the construction system ledger. When creating a construction system ledger, it is necessary to check the consistency of each document, for example, the construction details, construction period, construction permit, qualification, etc. of the construction work described in each document. Inspection work is done manually. Therefore, there is a problem that the inspection work of the construction system ledger is costly.

It is an object of the present invention to appropriately support the inspection work of the construction system ledger based on the actual data of the inspection of the construction system ledger of the past construction work in consideration of the above facts.

In order to achieve the above object, the construction system ledger inspection support device of the present invention is acquired by the character information acquisition unit that acquires the character information of the documents related to the construction system ledger of the target construction work, and the character information acquisition unit. Learning learned in advance from the learning data representing the character information of the document, the character information of the document related to the construction system ledger for learning, and the evaluation result of the document related to the construction system ledger. It is configured to include an information output unit that outputs evaluation information of the document of the target construction work based on the completed model. As a result, it is possible to appropriately support the inspection work of the construction system ledger based on the actual data of the inspection of the construction system ledger of the past construction work.

The character information acquisition unit of the present invention acquires the character information of the document related to the business operator involved in the target construction work as the character information of the document related to the construction system ledger, and the information output unit , The character information of the document related to the business acquired by the character information acquisition unit, the character information of the document related to the business involved in the construction work for learning, and the document related to the business. It is possible to output the evaluation information of the documents related to the business operator of the target construction work based on the trained model learned in advance from the training data representing the evaluation result. As a result, it is possible to appropriately support the inspection work of the documents related to the business operator based on the actual data of the inspection results of the documents related to the business operator in the construction system ledger.

Documents related to the business operator of the present invention may include at least one of a document relating to a contract between business operators and a construction business permit of the business operator. As a result, it is possible to appropriately support the inspection work of at least one of the documents related to the contract between the business operators and the construction business permit of the business operator.

The character information acquisition unit of the present invention uses the character information of the construction business permit of the business operator involved in the target construction work as the character information of the documents related to the construction system ledger, and the individual belonging to the business operator. The information output unit acquires the character information of the documents related to the above, and the information output unit uses the character information acquired by the character information acquisition unit and the characters of the construction business permit of the business operator involved in the construction work for learning. Based on the trained model pre-learned from the learning data representing the information and the textual information of the documents related to the individual belonging to the business operator and the evaluation results of the construction business permit and the documents related to the individual. Therefore, it is possible to output the evaluation information of the construction business permit of the business operator of the target construction work and the documents related to the individual belonging to the business operator. As a result, it is possible to appropriately support the inspection work between the construction business permit, which is a document related to the business operator, and the document related to the individual.

Documents relating to the individual of the present invention may include at least one of the individual's chief engineer qualification and the individual's work experience certificate. Thereby, for example, it is possible to appropriately support the inspection work of the consistency between the construction business permit and at least one of the chief engineer qualification certificate and the work experience certificate.

The construction system ledger inspection support model learning device of the present invention is based on learning data representing the text information of documents related to the construction system ledger for learning and the evaluation results of the documents related to the construction system ledger. , Learn the model for outputting the evaluation information of the documents related to the construction system ledger from the text information of the documents related to the construction system ledger of the construction work, and the textual information of the documents related to the construction system ledger of the construction work. It is configured to include a learning unit for obtaining a trained model that outputs evaluation information of documents related to the construction system ledger. As a result, it is possible to obtain a learned model that outputs evaluation information of documents related to the construction system ledger based on the learning data obtained from the actual data of the inspection results of the construction system ledger of the past construction work.

According to the present invention, it is possible to appropriately support the inspection work of the construction system ledger based on the actual data of the inspection of the construction system ledger of the past construction work.

It is explanatory drawing for explaining the inspection of the document related to the construction system ledger. It is a block diagram which shows the schematic structure of the construction system ledger inspection support device which concerns on this embodiment. It is a figure which shows an example of the data set for training. It is a figure which shows an example of the trained model used in 1st Embodiment. It is a figure which shows an example of the preprocessing routine of this embodiment. It is a figure which shows an example of the learning processing routine of this embodiment. It is a figure which shows an example of the inspection support processing routine of this embodiment. It is a figure which shows an example of the trained model used in the 2nd Embodiment. It is a figure which shows an example of the trained model used in 3rd Embodiment. It is a figure which shows an example of the trained model used in 4th Embodiment. It is a figure which shows an example of the trained model used in 5th Embodiment.

<Outline of this embodiment>

When creating a construction system ledger in construction work, it is necessary to check the consistency of each document related to the construction system ledger. These documents are, for example, submitted to the prime contractor by the construction cooperating company. FIG. 1 shows an explanatory diagram for explaining the inspection of the consistency of documents in the construction system ledger.

For example, as shown in FIG. 1, the person in charge of inspecting the construction system ledger, as an example of the documents related to the construction system ledger, obtains the document A regarding the contract between the contractors of the construction work and the permit B of the construction industry. By comparison, it is confirmed whether or not the description of the construction contents included in the document A regarding the contract and the type of the construction industry of the construction industry permit B match. For example, if the description of the construction content is "electrical work business", the type of permit for the construction industry must include "electrical work business". Therefore, the person in charge checks whether those descriptions in each document are consistent.

However, there are various ways in which each document related to the construction system ledger is described. For example, the description of the construction content of the document A relating to the contract may be a description such as "X wiring construction" or "X power supply construction". Again, the type of construction permit required is "electrical work". Further, when the description of the construction content of the document A relating to the contract is "X air conditioning equipment construction" or "X water supply / drainage equipment construction", the type of permission for the required construction industry is "pipe construction business". Furthermore, in addition to these items, it is necessary to confirm the consistency of the construction period described in each document.

As described above, since the method of describing the documents related to the construction system ledger is not uniform, it is difficult to automatically judge whether or not each document is consistent. For this reason, the inspection work of each document of the construction system ledger in the construction work must be performed manually, which requires a huge cost.

Therefore, in this embodiment, the inspection work of the construction system ledger is supported by the learned model learned by using the past inspection record data. Specifically, the actual data of the inspection of the construction system ledger of the past construction work is set as the learning data, and the evaluation of the document data from the document data related to the construction system ledger to be inspected based on the learning data. Generate a trained model that outputs information. Then, using the obtained learned model, the evaluation information of the documents related to the construction system ledger is output, and the inspection work of each document of the construction system ledger is supported. According to this embodiment, it is possible to appropriately support the inspection of the construction system ledger by using the learned model in which the inspection record data of the past construction system ledger is reflected.

In this embodiment, as an example of inspection of the construction system ledger, inspection support for the contents of the following documents (1) to (3) is provided.

(1) Inspection of contents related to the company (1-1) Construction permit (1-2) Order form and contract or construction contract contract (1-3) Basic contract or contract

(2) Inspection of personal information (2-1) Chief engineer qualification (2-2) Work experience certificate

(3) Consistency of multiple other documents

Hereinafter, each embodiment of the present invention will be described in detail.

<System configuration of construction system ledger inspection support device according to the first embodiment>

FIG. 2 is a block diagram showing an example of the configuration of the construction system ledger inspection support device 100 according to the first embodiment. As shown in FIG. 2, the construction system ledger inspection support device 100 can be functionally represented by a configuration including a data receiving unit 10, a computer 20, and an output unit 30.

The data reception unit 10 represents a combination of document data related to the construction system ledger of construction work for learning (hereinafter, simply referred to as "document data for learning") and the evaluation result of the document data for learning. Accept inspection record data. The data receiving unit 10 is realized by, for example, a keyboard, a mouse, an input / output device that receives input from an external device, or the like. In the present embodiment, a case where the data receiving unit 10 is a predetermined image reading device (for example, a scanner) and the document data is input as image data will be described as an example.

In the first embodiment, the document relating to the contract between the business operators involved in the construction work and the construction business permit of the business operator are used as the documents related to the construction system ledger. In the first embodiment, the inspection work as to whether the documents related to the contract and the construction business permit are consistent is supported. Documents related to contracts between business operators involved in construction work and construction business permits of business operators are examples of documents related to business operators involved in construction work.

Further, the data receiving unit 10 receives document data related to the construction system ledger of the construction work to be inspected (hereinafter, simply referred to as "document data to be inspected").

The computer 20 includes a CPU (Central Processing Unit), a ROM (Read Only Memory) that stores programs for realizing each processing routine, a RAM (Random Access Memory) that temporarily stores data, and a memory as a storage means. , Network interface, etc. are included. The computer 20 functionally includes a character information acquisition unit 22, a learning data storage unit 24, a learning unit 26, a learned model storage unit 28, and an information output unit 29.

The character information acquisition unit 22 acquires a plurality of inspection record data received by the data reception unit 10. Further, the character information acquisition unit 22 acquires the document data to be inspected received by the data reception unit 10. The character information acquired from the document data may be the character information of a specific part of the document data or the entire character information.

Then, the character information acquisition unit 22 acquires character information from the document data for learning among the inspection record data. Further, the character information acquisition unit 22 acquires character information from the document data to be inspected. For example, the character information acquisition unit 22 acquires character information from document data, which is image data, by OCR (Optical Character Recognition).

Further, the character information acquisition unit 22 uses a combination of the character information acquired from the learning document data of the inspection record data and the evaluation information of the learning document data of the inspection record data as learning data. It is stored in the learning data storage unit 24. The evaluation result of the document data for learning is given manually in the inspection process of the construction system ledger in the past construction work.

The learning data storage unit 24 stores a plurality of learning data acquired by the character information acquisition unit 22. For example, as shown in FIG. 3, the character information of the document data for learning and the evaluation information of the document data for learning are stored in association with each other.

In the learning data of the achievement ID "00001" shown in FIG. 3, in the construction system ledger of the already completed construction work, the character information of the purchase order and the contract is "XX1", and the character information of the construction business permit is "YY1". If, it means that the evaluation result regarding the consistency of those documents was "OK".

For example, when the text information of the construction content read from the order form and the confirmation is "XX air conditioning equipment construction", and the text information of the construction industry read from the construction business permit includes "pipe construction business". Is given an OK label to the training data. In addition, for example, if the construction content of the order form and the contract is "XX electrical equipment construction" and the construction industry of the construction business permit does not include "electrical construction business", the learning data is NG. Is labeled as.

In this case, for example, character information is expressed numerically so that it can be read by a computer. For example, as the character information, a set of identification information for identifying each word included in the character information, or an appearance frequency vector of each word included in the character information can be used.

The learning unit 26 is based on a plurality of learning data stored in the learning data storage unit 24, and from the text information of the document data related to the construction system ledger to be inspected, the document related to the construction system ledger to be inspected. A trained model is obtained by training a model for outputting data evaluation information. Since the learning data is the inspection record data corresponding to the already completed construction work, the inspection record data in the past construction system ledger is reflected in the trained model.

In this embodiment, a trained model as shown in FIG. 4 is generated. For example, as shown in FIG. 4, a neural network can be used as an example of a model, and deep learning can be used as an example of a learning algorithm.

The trained model storage unit 28 stores the trained model obtained by the learning unit 26.

The information output unit 29 constructs the construction work to be inspected based on the character information of the document data to be inspected acquired by the character information acquisition unit 22 and the learned model stored in the trained model storage unit 28. Output the evaluation information of the document data of the system ledger.

As shown in FIG. 4, when the character information of the document data to be inspected is input to the trained model, the trained model performs arithmetic processing and outputs the evaluation information of the target document data. This makes it possible to obtain information on whether or not the document data input to the trained model is consistent.

From the trained model, for example, the probability that the document data of the construction work order form and contract to be inspected and the document data of the construction business permit match (probability of OK) is 70%, and they do not match. Information in the format that the probability (probability of being NG) is 30% is output.

The output unit 30 outputs the evaluation information of the document data to be inspected, which is output by the information output unit 29, as a result. Then, for example, when the probability of being NG is 50% or more, an alert is output. As a result, the construction permit (29 types) required for the actual construction work is determined, and the consistency between the purchase order and the contract and the construction permit is checked. The output unit 30 is realized by, for example, a display.

The person in charge of inspection who inspects the documents of the construction system ledger confirms the evaluation information output by the output unit 30. The person in charge of inspection, who is the user of the construction system ledger inspection support device, refers to the evaluation information output from the trained model.

<Operation of construction system ledger inspection support device>

Next, the operation of the construction system ledger inspection support device 100 will be described. The construction system ledger inspection support device 100 executes a preprocessing routine, a learning processing routine, and an inspection support processing routine.

<Preprocessing routine>

When the data receiving unit 10 of the construction system ledger inspection support device 100 receives the input of a plurality of inspection record data, the computer 20 of the construction system ledger inspection support device 100 executes the preprocessing routine shown in FIG.

In step S50, the character information acquisition unit 22 acquires a plurality of inspection record data received by the data reception unit 10.

In step S52, the character information acquisition unit 22 acquires character information from the learning document data of the inspection record data for each of the plurality of inspection record data acquired in step S50.

In step S54, the character information acquisition unit 22 stores the combination of the character information acquired in step S52 and the evaluation result of the document data for learning as learning data for each of the plurality of inspection record data. It is stored in the unit 24 and the preprocessing routine is terminated.

<Learning processing routine>

Next, when the computer 20 of the construction system ledger inspection support device 100 receives the instruction signal for executing the learning process, the computer 20 executes the learning process routine shown in FIG.

In step S100, the learning unit 26 acquires a plurality of learning data stored in the learning data storage unit 24 in the preprocessing routine.

In step S102, the learning unit 26 learns a model for outputting the evaluation information of the document data from the document data of the construction system ledger based on the plurality of learning data acquired in step S100. Generate a finished model.

In step S104, the learning unit 26 stores the trained model generated in step S102 in the trained model storage unit 28, and ends the learning processing routine.

<Inspection support processing routine>

When the trained model is stored in the trained model storage unit 28 and the data reception unit 10 receives the document data of the construction system ledger to be inspected, the construction system ledger inspection support device 100 receives the document data of the construction system ledger to be inspected, and the inspection support process shown in FIG. Execute the routine.

In step S200, the information output unit 29 acquires the document data to be inspected received by the data reception unit 10.

In step S202, the character information acquisition unit 22 acquires character information from the document data to be inspected acquired in step S200.

In step S204, the information output unit 29 reads out the trained model stored in the trained model storage unit 28.

In step S206, the information output unit 29 performs the inspection acquired in step S200 based on the character information of the document data to be inspected acquired in step S202 and the learned model read in step S204. Output the evaluation information of the target document data as a result.

The output unit 30 outputs the evaluation information output by the information output unit 29 as a result. The person in charge of inspecting the construction system ledger confirms and refers to the evaluation information output by the output unit 30. For example, if the probability of being OK is 90% or more, the person in charge of inspection determines that the documents are consistent. On the other hand, for example, when the probability of being NG is 50% or more, the person in charge of inspection manually inspects those documents again.

As described in detail above, in the present embodiment, the text information of the documents related to the construction system ledger of the target construction work, the text information of the documents related to the construction system ledger of the construction work for learning, and the construction system. The evaluation information of the construction work documents to be inspected is output based on the trained model learned in advance from the learning data representing the evaluation results of the documents related to the ledger. As a result, it is possible to appropriately support the inspection of each document of the construction system ledger in the construction work based on the inspection record data of the past construction work.

Further, in the present embodiment, a model for outputting the evaluation information of the document is trained from the character information of the document data based on the learning data, and the trained model is obtained. As a result, it is possible to obtain a trained model that outputs the evaluation information of the document from the character information of the document.

<System configuration of construction system ledger inspection support device according to the second embodiment>

Next, the second embodiment will be described. The second embodiment differs from the first embodiment in that the documents related to the construction system ledger are only the documents related to the contract between the business operators involved in the construction work. Since the configuration of the second embodiment is the same as that of the first embodiment, the same reference numerals are given and the description thereof will be omitted.

When concluding a contract between contractors, it is necessary to conclude a contract that meets the prescribed 14 items stipulated in the Construction Business Act (for example, References (Ministry of Land, Infrastructure, Transport and Tourism, Land and Construction Industry Bureau, Construction Industry Division, "Construction". Guidelines for Compliance with Business Laws (5th Edition) -Points to Note Regarding the Relationship between the Main Contractor and the Subcontractor- ", March 2017).

However, since the wording described in the contract varies, the predetermined 14 items may be expressed by various wording. Therefore, there is a problem that the person in charge of inspection often manually confirms the contents of the contract, which is costly.

Therefore, in the second embodiment, based on the character information obtained from the document data related to the contract, it is checked by the trained model whether or not the predetermined 14 items are satisfied. Documents related to contracts between business operators include, for example, a combination of an order form and a contract with a basic contract or a contract, or a construction contract contract.

Specifically, in the second embodiment, a trained model as shown in FIG. 8 is used. For example, from the trained model, the probability that the predetermined 14 items are satisfied (probability of being OK) and the probability that the predetermined 14 items are not satisfied (probability of being NG) are output. The probability of NG may be output for each item.

Then, for example, when the probability of being NG is 50% or more, an alert is output. In addition, the construction business permit may be determined and whether or not a chief engineer is required may be output together.

As described in detail above, in the second embodiment, the contract to be inspected is based on the character information obtained from the document data related to the contract between the business operators and the learned model pre-learned from the learning data. Outputs the evaluation information of the documents related to. As a result, it is possible to appropriately support the inspection work of the documents related to the business operator based on the actual data of the inspection results of the documents related to the business operator in the construction system ledger.

<System configuration of construction system ledger inspection support device according to the third embodiment>

Next, a third embodiment will be described. The third embodiment differs from the first or second embodiment in that documents relating to individuals involved in construction work are handled as documents related to the construction system ledger. Since the configuration of the third embodiment is the same as that of the first embodiment, the same reference numerals are given and the description thereof will be omitted.

At the construction site, it is necessary to assign a chief engineer with appropriate qualifications according to the type of permit of the construction business permit. Whether or not a chief engineer is assigned according to the type of permit of the construction business permit is checked by comparing the construction business permit and the chief engineer qualification. Therefore, in the third embodiment, as shown in FIG. 9, the permission type of the construction business permit and the chief engineer qualification are based on the character information of the combination of the chief engineer license and the construction business license. Check if is consistent with the trained model.

Specifically, the character information acquisition unit 22 of the third embodiment acquires character information from the document data of the construction business permit of the business operator involved in the construction work. In addition, the character information acquisition unit 22 acquires the character information of the document data of the chief engineer qualification of the individual belonging to the business operator. In the third embodiment, the chief engineer qualification is an example of documents related to an individual belonging to a business operator.

In the learning data storage unit 24 of the third embodiment, the textual information of the document data of the construction business permit of the business operator involved in the construction work for learning and the chief engineer qualification certificate of the individual belonging to the business operator. Data for learning representing the textual information of the document data of the above and the evaluation result between the construction business license and the chief engineer qualification is stored.

The learning unit 26 of the third embodiment is based on the text information of the document data of the construction business permit and the chief engineer qualification to be inspected based on the plurality of learning data stored in the learning data storage unit 24. , A model for outputting the evaluation information of the document data is trained, and a trained model is obtained.

The information output unit 29 of the third embodiment inspects based on the character information of the document data to be inspected acquired by the character information acquisition unit 22 and the trained model stored in the trained model storage unit 28. Output the evaluation information of the document data of the target chief engineer qualification and the construction business permit.

As described in detail above, in the third embodiment, based on the character information obtained from the document data of the construction industry permit and the chief engineer qualification, and the trained model pre-learned from the learning data. , Output the evaluation information of the construction industry license and the chief engineer qualification to be inspected. This can appropriately support the inspection work between the construction permit and the chief engineer's qualification.

In addition, by using the learning data, which is the inspection record data of the construction industry permit and the chief engineer qualification, it is difficult or peculiar to determine the qualification name required to be assigned as the chief engineer. Inspection support is appropriately provided by using a trained model that reflects past inspection performance data.

In addition, the validity of the chief engineer qualification (for example, the training history of the chief engineer, etc.) is further confirmed, and the name, etc. (personal information) of the chief engineer qualification is collated with other systems to the relevant construction company. You may further confirm that you belong.

<System configuration of construction system ledger inspection support device according to the fourth embodiment>

Next, a fourth embodiment will be described. The fourth embodiment differs from the third embodiment in that it handles a work experience certificate as a document relating to an individual involved in construction work. Since the configuration of the fourth embodiment is the same as that of the first embodiment, the same reference numerals are given and the description thereof will be omitted.

If the chief engineer to be assigned to the construction site does not have the prescribed national qualifications, etc., it is necessary to meet the requirements of the chief engineer based on his educational background and practical experience. Whether or not this requirement is met is checked by the description of the individual's work experience certificate and construction permit and the description in the work experience certificate. Therefore, in the fourth embodiment, as shown in FIG. 10, the type of permit of the construction business permit and the content of the work experience certificate are based on the character information of the combination of the work experience certificate and the construction business permit. Check with the trained model to see if they are consistent with.

The character information acquisition unit 22 of the fourth embodiment acquires, for example, construction contents, years of experience, educational background, name, etc. as character information from the document data of the work experience certificate.

As shown in FIG. 10, for example, from the trained model, the construction content acquired from the document data of the work experience certificate and the type of construction industry acquired from the document data of the construction permit are consistent. The probability of being (OK) and the probability that the construction content and the type of construction industry do not match (NG) are output. Also, from the trained model, the probability that the required number of years of work experience and educational background are consistent (OK) and the probability that the required number of years of work experience and educational background are not consistent (NG). Is output.

As described in detail above, in the fourth embodiment, based on the character information obtained from the document data of the construction business permit and the work experience certificate, and the trained model pre-learned from the learning data. Output the evaluation information of the construction business permit and work experience certificate to be inspected. This can appropriately support the inspection work between the construction permit and the work experience certificate.

In addition, even if it is difficult or peculiar to judge the construction contents, by using the learned model that reflects the past inspection record data, it is possible to appropriately support the inspection of the documents in the construction system ledger. In addition, you may further confirm that you belong to the construction company by collating the name etc. (personal information) of the work experience certificate with other systems.

<System configuration of construction system ledger inspection support device according to the fifth embodiment>

Next, a fifth embodiment will be described. The fifth embodiment differs from the first to third embodiments in that a plurality of documents related to the construction system ledger are handled. Since the configuration of the fifth embodiment is the same as that of the first embodiment, the same reference numerals are given and the description thereof will be omitted.

In the first to fourth embodiments described above, the case of inspecting a specific document in the construction system ledger has been described as an example, but a learned model for inspecting all the documents related to the construction system ledger is generated. You can also do it. Therefore, in the fifth embodiment, a trained model as shown in FIG. 11 is generated.

In the trained model shown in FIG. 11, character information acquired from all document data related to the construction system ledger is input, and information regarding which item is inconsistent is output. For example, information on whether or not the construction period is consistent among a plurality of documents and information on whether or not the construction content and the type of permit of the construction industry are consistent are output.

As described in detail above, in the fifth embodiment, the inspection target is based on the character information obtained from a plurality of document data related to the construction system ledger and the learned model learned in advance from the learning data. Outputs the evaluation information of multiple documents in the construction system ledger. As a result, it is possible to appropriately support the inspection work of a plurality of documents related to the construction system ledger.

The present invention is not limited to the above-described embodiment, and various modifications and applications can be made without departing from the gist of the present invention.

For example, in the above embodiment, the case where the neural network model as an example of the model is trained by deep learning has been described as an example, but the present invention is not limited to this. For example, another model different from the neural network model may be trained by another learning method different from deep learning.

Further, in the above embodiment, the construction system ledger inspection support device has been described as an example of performing learning processing and inspection support processing, but the present invention is not limited to this. For example, the system may be configured by a construction system ledger inspection support model learning device that performs learning processing and a construction system ledger inspection support device that performs inspection support processing.

Further, the trained model described in each of the above embodiments is an example, and the trained model may be configured in any way as long as it is an inspection of documents related to the construction system ledger.

In addition, the inspection result of the construction system ledger of the actual construction work may be fed back to the learned model.

In addition, the trained model may be configured to output an answer to the gray zone in the construction system ledger. For example, a trained model is used to output OK or NG according to the text information obtained from the document information "Does the construction company need not have an electrical work permit for solar panel installation work?" It may be configured.

Further, in each of the above embodiments, the case where the character information acquired from each document is simply used has been described as an example, but the present invention is not limited to this, and preprocessing for the character information may be performed. For example, a trained model may be generated by extracting only predetermined character strings (for example, construction period, construction name, etc.) from a combination of a purchase order / confirmation and a construction business permit, and using those character strings as training data. ..

Further, the predetermined inspection process and the inspection process using the learned model generated from the inspection record data may be combined to support the inspection work of the documents in the construction system ledger. For example, in this case, whether the expiration date of the construction business permit meets the construction period, whether the construction content of the purchase order / contract matches the type of construction business permit, and whether the basic contract / terms are in the contract contract. Whether the minimum required items are met or whether the qualification is a chief engineer according to the type of permit of the construction industry is processed using a table in which predetermined information is stored. If the description of each document does not match the information stored in the predetermined table (for example, if the description of the contract document is an ambiguous expression, the construction content is not classified as the type of permit specified by the Construction Business Act. , Etc.), a trained model generated from past inspection record data may be used.

Further, although the mode in which the program is stored (installed) in advance in the storage unit (not shown) has been described above, the program is recorded on any recording medium such as a CD-ROM, a DVD-ROM, or a micro SD card. It is also possible to provide it in the form in which it is used.

10 Data reception unit 20 Computer 22 Character information acquisition unit 24 Learning data storage unit 26 Learning unit 28 Learned model storage unit 29 Information output unit 30 Output unit 100 Construction system ledger inspection support device

Claims (6)

The character information acquisition department that acquires the character information of documents related to the construction system ledger of the target construction work,
For learning showing the character information of the document acquired by the character information acquisition unit, the character information of the document related to the construction system ledger of the construction work for learning, and the evaluation result of the document related to the construction system ledger. An information output unit that outputs evaluation information of the document of the target construction work based on a trained model learned in advance from the data, and an information output unit.
Including
The character information acquisition unit acquires the character information of the document related to the business operator involved in the target construction work as the character information of the document related to the construction system ledger.
The information output unit includes the character information of the document related to the business operator acquired by the character information acquisition unit, the character information of the document related to the business operator involved in the construction work for learning, and the business operator. Based on the trained model learned in advance from the learning data representing the evaluation results of the documents related to the above, the evaluation information of the documents related to the business operator of the target construction work is output.
Construction system ledger inspection support device. Documents related to the business operator include at least one of the documents relating to the contract between the business operators and the construction business permit of the business operator.
The construction system ledger inspection support device according to claim 1 . The character information acquisition department that acquires the character information of documents related to the construction system ledger of the target construction work,
For learning showing the character information of the document acquired by the character information acquisition unit, the character information of the document related to the construction system ledger of the construction work for learning, and the evaluation result of the document related to the construction system ledger. An information output unit that outputs evaluation information of the document of the target construction work based on a trained model learned in advance from the data, and an information output unit.
Including
The character information acquisition unit relates to the character information of the construction business permit of the business operator involved in the target construction work and the individual belonging to the business operator as the character information of the documents related to the construction system ledger. Get the textual information of the document and
The information output unit relates to the character information acquired by the character information acquisition unit, the character information of the construction business permit of the business operator involved in the construction work for learning, and the individual belonging to the business operator. Based on the trained model learned in advance from the learning data representing the textual information of the documents and the evaluation results of the construction business permit and the documents related to the individual, the business operator of the target construction work Output the evaluation information of the construction business permit and the documents related to the individual belonging to the business operator,
Construction system ledger inspection support device. Documents relating to the individual include at least one of the individual's chief engineer qualification and the individual's work experience certificate.
The construction system ledger inspection support device according to claim 3 . Characters of documents related to the construction system ledger of construction work based on the learning data representing the character information of the documents related to the construction system ledger for learning and the evaluation results of the documents related to the construction system ledger. The model for outputting the evaluation information of the documents related to the construction system ledger is learned from the information, and the evaluation information of the documents related to the construction system ledger is obtained from the text information of the documents related to the construction system ledger of the construction work. Includes a learning unit to get the trained model to output
The character information of the document related to the construction system ledger is the character information of the document related to the business operator involved in the construction work.
The evaluation information of the documents related to the construction system ledger is the evaluation information of the documents related to the business operator of the construction work.
Construction system ledger inspection support model learning device. Characters of documents related to the construction system ledger of construction work based on the learning data representing the character information of the documents related to the construction system ledger for learning and the evaluation results of the documents related to the construction system ledger. The model for outputting the evaluation information of the documents related to the construction system ledger is learned from the information, and the evaluation information of the documents related to the construction system ledger is obtained from the text information of the documents related to the construction system ledger of the construction work. Includes a learning unit to get the trained model to output
The character information of the document related to the construction system ledger is the character information of the construction business permit of the business operator involved in the construction work and the character information of the document related to the individual belonging to the business operator.
The evaluation information of the documents related to the construction system ledger is the evaluation information of the construction business permit of the business operator of the construction work and the documents related to the individual belonging to the business operator.
Construction system ledger inspection support model learning device.

Images