Building construction scheme key information extraction system and method thereof

Abstract

The invention relates to a building construction scheme key information extraction system and method, and belongs to the technical field of building construction, the system comprises a building construction scheme identification system, an information association construction system and a key information structured packaging system, the building construction scheme identification system is mainly used for analysis of unstructured data; the information association construction system is mainly used for constructing an association relationship between texts detected by the recognition system; and the key information structured packaging system is mainly used for further packaging the output result of the information association construction system and unifying the packaging format. According to the building construction scheme key information extraction system and method, the problems that the utilization rate of a building construction scheme is low and file content is difficult to retrieve and query are solved based on word embedding and large oracle model technologies.

Description

Technical Field

[0001] The invention belongs to the technical field of building construction, and in particular relates to a system and method for extracting key information of a building construction plan. Background Art

[0002] As the technical experience accumulated by construction companies over the years, construction plans are the most valuable knowledge assets of construction companies. Before the full electronicization of documents, a large number of construction plans could only be stored in the form of paper materials. In recent years, as companies move towards digital transformation, a large number of paper construction plans have been scanned into electronic documents for storage.

[0003] However, these electronic documents of construction plans stored in PDF format are unstructured data, which not only makes them difficult to parse and use, but also cannot provide a reference for the current corporate construction plan preparation.

[0004] At present, there is a lack of effective construction plan analysis and key information extraction methods to extract, understand and store such unstructured text data, which makes it difficult for construction plans within enterprises to play their due value.

[0005] Therefore, how to provide a system and method for extracting key information of a building construction plan is a technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0006] Aiming at the problems of low utilization rate of construction plans and difficulty in searching and retrieving file contents, a system and method for extracting key information of construction plans was proposed based on word embedding and large language model technology. The system and method can parse unstructured construction plan files, establish the correlation of text contents, and encapsulate them into structured data.

[0007] In order to solve the above technical problems, the present invention includes the following technical solutions:

[0008] A key information extraction system for a building construction plan, comprising:

[0009] A construction plan recognition system, which uses OCR recognition and a page detection module to parse unstructured data;

[0010] An information association construction system, wherein the information association construction system uses the context association understanding ability and content summarization ability of the model to establish context information association and summarize and generalize complex information;

[0011] The key information structured packaging system files the existing information in a unified manner by setting prompt words, so that the packaging format is unified.

[0012] Furthermore, the unstructured data is in PDF format, and it is difficult to directly extract the text content of the file.

[0013] Furthermore, the page detection module detects page by page and records and stores the title, text and chart respectively.

[0014] Furthermore, the key information structured packaging system automatically and uniformly archives the existing information in the form of attributes and contents.

[0015] The present invention also provides a method for extracting key information of a building construction plan, the method comprising:

[0016] Step S1, using the building construction plan key information extraction system;

[0017] Step S2: parsing unstructured data through a building construction plan recognition system;

[0018] Step S3, constructing the association relationship between the texts detected by the building construction plan recognition system through the information association construction system;

[0019] Step S4: further encapsulate the output result of the information association construction system through the key information structured encapsulation system, and unify the encapsulation format.

[0020] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0021] The present invention provides a system and method for extracting key information of a construction plan. The system includes a construction plan recognition system, an information association construction system and a key information structured packaging system. The construction plan recognition system is mainly used for parsing unstructured data; the information association construction system is mainly used for constructing the association relationship between texts detected by the above recognition system; the key information structured packaging system mainly further packages the output results of the above information association construction system and unifies the packaging format.

[0022] Compared with the traditional method, the beneficial effects of the present invention are mainly as follows:

[0023] (1) Construction plan analysis taking into account positional relationships. When analyzing the construction plan, the positional relationships among titles, text, and diagrams are retained.

[0024] (2) Build information associations. Use big model technology to build information associations among the various parts of the construction plan. At the same time, use the summarizing capabilities of the big model to professionally summarize and generalize the text, thereby reducing the cost of structured storage.

[0025] (3) Automatic packaging of structured data: Automatic and intelligent packaging of text data is achieved through specific prompt words, thereby improving the utilization value of construction plan assets. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is an architecture diagram of a system for extracting key information from a construction plan in one embodiment of the present invention. DETAILED DESCRIPTION

[0027] The following is a further detailed description of a system and method for extracting key information of a construction plan provided by the present invention in conjunction with the accompanying drawings and specific embodiments. The advantages and features of the present invention will become more apparent from the following description.

[0028] Embodiment 1

[0029] Combine the following Figure 1 , a detailed description of the key information extraction system for building construction plans of the present invention is given.

[0030] Please refer to Figure 1 A key information extraction system for a construction plan includes a construction plan recognition system, an information association construction system and a key information structured packaging system. The construction plan recognition system is mainly used for parsing unstructured data; the information association construction system is mainly used to construct the association relationship between the texts detected by the above recognition system; the key information structured packaging system mainly further encapsulates the output results of the above information association construction system and unifies the packaging format.

[0031] Specifically, the construction plan recognition system is mainly used for the analysis of unstructured data. Since most scanned files are unstructured data such as PDF, it is difficult to directly extract the text content of the file. In this embodiment, OCR recognition is used to perform preliminary recognition of the construction plan. Since the current OCR recognition technology cannot well preserve the positional relationship between texts, in this embodiment, a page detection module is used to detect page by page, and the title, text, charts, etc. are recorded and stored separately.

[0032] In this embodiment, more preferably, the information association construction system is mainly used to construct the association relationship between the texts detected by the above recognition system. By using the context association understanding ability and content summarization ability of the large model, while considering more significant features such as titles, texts, and charts, the context information association is established and the complex information is summarized.

[0033] In this embodiment, more preferably, the key information structured packaging system mainly further packages the output results of the above-mentioned information association construction system and unifies the packaging format. Although similar methods have been used to classify and archive construction plans in a unified manner, the archiving mainly relies on manual statistics. This embodiment proposes a packaging system based on a large model, which can automatically and uniformly archive existing information in the form of attributes and content by setting special prompt words, which can greatly improve the efficiency of packaging while ensuring the uniformity of the packaging format.

[0034] Please continue to refer to Figure 1 The present invention also provides a method for extracting key information of a building construction plan, the method comprising:

[0035] Step S1, using the building construction plan key information extraction system;

[0036] Step S2: parsing unstructured data through a building construction plan recognition system;

[0037] Step S3, constructing the association relationship between the texts detected by the building construction plan recognition system through the information association construction system;

[0038] Step S4: further encapsulate the output result of the information association construction system through the key information structured encapsulation system, and unify the encapsulation format.

[0039] The above examples are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. The above embodiments only express several embodiments of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for those of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A key information extraction system for building construction plans, characterized in that: include: A construction plan recognition system, which uses OCR recognition and a page detection module to parse unstructured data; An information association construction system, wherein the information association construction system uses the context association understanding ability and content summarization ability of the model to establish context information association and summarize and generalize complex information; The key information structured packaging system files the existing information in a unified manner by setting prompt words, so that the packaging format is unified.

2. The key information extraction system for building construction scheme according to claim 1 is characterized in that: The unstructured data is in PDF format.

3. The key information extraction system for building construction scheme according to claim 1 is characterized in that: The page detection module detects page by page, and records and stores the title, text and chart respectively.

4. The key information extraction system for building construction scheme according to claim 1 is characterized in that: The key information structured packaging system automatically and uniformly archives the existing information in the form of attributes and contents.

5. A method for extracting key information of a construction plan, characterized in that: include: Step S1, using the key information extraction system for building construction scheme according to any one of claims 1 to 4; Step S2: parsing unstructured data through a building construction plan recognition system; Step S3, constructing the association relationship between the texts detected by the building construction plan recognition system through the information association construction system; Step S4: further encapsulate the output result of the information association construction system through the key information structured encapsulation system, and unify the encapsulation format.

Images