An alloy data extraction method, apparatus, device, and medium

Abstract

This application discloses a method, apparatus, device, and medium for alloy data extraction, relating to the field of alloy data extraction technology. The method includes: acquiring PDF documents such as literature and patents containing alloy data, extracting content from them to obtain extracted content in a first format and a second format; extracting alloy data and features from the first and second format extracted content based on a multimodal alloy information extraction model to obtain extraction results; and performing association processing on the extraction results based on alloy names to obtain associated data information corresponding to each alloy. This technical solution can automatically and accurately extract alloy association information from unstructured PDF files without limiting the format of the data in the PDF file. Furthermore, this technical solution is applicable to various alloy materials, improving its generalization ability.

Description

Technical Field

[0001] This invention relates to the field of alloy data extraction technology, and in particular to an alloy data extraction method, apparatus, equipment and medium. Background Technology

[0002] When extracting data from PDF (Portable Document Format) files, traditional NLP methods rely on precise rules, templates, and dictionaries to extract content with fixed formats, making it difficult to handle diverse document formats and expressions. They also suffer from the need for extensive manual definition of terminology, limitations in data extraction for specific alloy materials, and poor generalization. Summary of the Invention

[0003] This invention provides a method, apparatus, device, and medium for alloy data extraction, which can accurately extract alloy-related information from unstructured PDF files and is applicable to various alloy materials.

[0004] According to one aspect of the present invention, an alloy data extraction method is provided, the method comprising:

[0005] Obtain a PDF file containing alloy data and extract its content to obtain the extracted content in a first format and a second format.

[0006] Based on the multimodal alloy information extraction model, alloy data and features are extracted from the content extracted in the first format and the content extracted in the second format to obtain the extraction results.

[0007] The extraction results are processed based on the alloy name to obtain the associated data information corresponding to each alloy.

[0008] According to another aspect of the present invention, an apparatus for alloy data extraction method is provided, comprising:

[0009] The content extraction module is used to acquire a PDF file containing alloy data and extract its content to obtain extracted content in a first format and extracted content in a second format.

[0010] The data extraction module is used to extract alloy data and features from the extracted content in the first format and the extracted content in the second format based on the multimodal alloy information extraction model, and obtain the extraction results.

[0011] The data association module is used to associate the extraction results based on the alloy name to obtain the associated data information corresponding to each alloy.

[0012] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0013] At least one processor; and

[0014] A memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the alloy data extraction method according to any embodiment of the present invention.

[0016] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the alloy data extraction method according to any embodiment of the present invention.

[0017] The technical solution of this application includes: acquiring a PDF file containing alloy information, extracting its content to obtain extracted content in a first format and extracted content in a second format; extracting alloy data and features from the extracted content in the first and second formats based on a multimodal alloy information extraction model to obtain extraction results; and performing association processing on the extraction results based on alloy names to obtain associated data information corresponding to each alloy. This technical solution can automatically and accurately extract alloy association information from unstructured PDF files without limiting the format of the data in the PDF file. Furthermore, this technical solution is applicable to various alloy materials, improving its generalization ability.

[0018] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of an alloy data extraction method according to Embodiment 1 of this application;

[0021] Figure 2 This is a flowchart of an alloy data extraction method according to Embodiment 2 of this application;

[0022] Figure 3 This is a specific example of a layout analysis result provided in Embodiment 2 of this application;

[0023] Figure 4 This is a stress-strain curve diagram provided according to Embodiment 2 of this application;

[0024] Figure 5 This is a metallographic image provided according to Embodiment 2 of this application;

[0025] Figure 6 This is an electron microscope image provided according to Embodiment 2 of this application;

[0026] Figure 7 is a schematic flowchart of the overall process according to Embodiment 2 of this application, wherein Figure 7(a) is a flowchart of the preprocessing layer; Figure 7(b) is a flowchart of the standardization definition layer; Figure 7(c) is a flowchart of the agent application layer; and Figure 7(d) is a flowchart of the data postprocessing layer.

[0027] Figure 8 This is a schematic diagram of an alloy data extraction method apparatus according to Embodiment 3 of this application;

[0028] Figure 9 This is a schematic diagram of the structure of an electronic device that implements an alloy data extraction method according to an embodiment of this application. Detailed Implementation

[0029] To enable those skilled in the art to better understand the present invention, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "first," "second," "target," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0031] Example 1

[0032] Figure 1 This application provides a flowchart of an alloy data extraction method according to Embodiment 1. This embodiment is applicable to the extraction of alloy data. The method can be executed by an alloy data extraction method apparatus, which can be implemented in hardware and / or software and can be configured in an electronic device with data processing capabilities. Figure 1 As shown, the method includes:

[0033] S110: Obtain a PDF file containing alloy data and extract its content to obtain extracted content in a first format and extracted content in a second format.

[0034] The PDF file containing alloy information can be manufacturing process documents, patents, etc., and its format is PDF. The technical solution of this application embodiment extracts features such as alloy composition, heat treatment process, grain morphology, matrix and precipitate morphology, and mechanical property indicators from the PDF file containing alloy information to obtain the corresponding associated data for each alloy.

[0035] The content extracted in the first format is different from that extracted in the second format, and the first format is different from the second format. For example, the first format is Markdown (Lightweight Markup Language), which is plain text format; the second format is image format, such as PNG format.

[0036] Since PDF files can contain various forms of data such as text, data, tables, and images, content recognition can be performed to extract the content in Markdown format, while charts such as stress-strain curves can be converted to PNG format.

[0037] S120, Based on the multimodal alloy information extraction model, alloy data and features are extracted from the extracted content in the first format and the extracted content in the second format to obtain the extraction results.

[0038] Among them, the multimodal alloy information extraction model is a multimodal large model, which is used to extract alloy-related data from the extracted content. Since the extracted content includes the first format and the second format, and the second format, for example, may include various information such as curves, text, and images, the alloy data needs to be extracted through the multimodal model.

[0039] For example, when extracting data based on a multimodal alloy information extraction model, a data extraction agent, a verification agent, and a graph understanding agent can be constructed based on the multimodal large model. The data extraction agent extracts alloy data from the first format of the extracted content, the graph understanding agent extracts features from the second format of the extracted content, and the verification agent verifies the extraction results. After the verification is passed, data such as the composition and heat treatment process of a certain alloy are obtained, as well as key index features in the stress-strain curve, grain shape and size in the metallographic image, and the proportion of porosity and cracks.

[0040] S130, the extraction results are processed based on the alloy name to obtain the associated data information corresponding to each alloy.

[0041] For example, after obtaining the extraction results, since a PDF file containing alloy information may contain multiple extraction results corresponding to alloys, and the data types contained in each extraction result may be different, this application embodiment uses the alloy name as a basis to associate the extracted data with the alloy name, thereby obtaining the associated data information corresponding to each alloy, which can be structured and stored in the database.

[0042] The technical solution of this application includes: acquiring a PDF file containing alloy information, extracting its content to obtain extracted content in a first format and extracted content in a second format; extracting alloy data and features from the extracted content in the first and second formats based on a multimodal alloy information extraction model to obtain extraction results; and performing association processing on the extraction results based on alloy names to obtain associated data information corresponding to each alloy. This technical solution can automatically and accurately extract alloy association information from unstructured PDF files without limiting the format of the data in the PDF file. Furthermore, this technical solution is applicable to various alloy materials, improving its generalization ability.

[0043] Example 2

[0044] Figure 2 This is a flowchart of an alloy data extraction method provided in Embodiment 2 of this application. This embodiment is an optimization based on the above embodiment.

[0045] like Figure 2 As shown, the method in this embodiment of the application specifically includes the following steps:

[0046] S210: Obtain the PDF file containing alloy data and perform layout analysis to obtain the text, table, and chart sections.

[0047] For example, a PDF file containing alloy data is obtained and converted to an image format. Then, layout analysis is performed on the image file. Specifically, this layout analysis involves marking different types of content within the image file. Figure 3 This is a specific example of a layout analysis result (where yellow boxes represent charts, green boxes represent numbers, red boxes represent text, and purple boxes represent tables). The type of content marked is not limited and can include text, tables, chart blocks, etc. Further, determine the correct document order for the text, table, and chart sections, and convert them to Markdown and PNG formats according to this order. (It should be noted that if a table contains text that can be extracted, then it is considered a table section; if a table contains an image, then it can be a chart section. The chart section refers to the image format, and the content of the image is not limited; it can be a graph, metallographic image, or a table from which text cannot be extracted, etc.)

[0048] S220, perform optical character recognition on the text portion and the table portion to obtain the extracted content in the first format, and obtain the extracted content in the second format from the chart portion.

[0049] For example, optical character recognition is performed on the text portion and the table portion to obtain the extracted content in Markdown format, and the extracted content in PNG format is determined based on the chart portion.

[0050] S230, based on the data extraction agent, alloy data is extracted from the extracted content of the first format to obtain alloy data.

[0051] S240, the graph-based understanding agent extracts features from the extracted content in the second format to obtain key features and / or indicator information.

[0052] For example, the LangChain and Langraph frameworks can be used to construct a multi-agent system (this multi-agent system includes a data extraction agent, a validation agent, and a graph understanding agent; each agent can be built based on a multimodal model, so prompt words need to be constructed for each agent to issue commands). The data extraction agent extracts alloy data from Markdown format content, and the graph understanding agent extracts key features and / or indicator information from image format content. This approach allows for targeted processing of Markdown and image format content to obtain more accurate extraction results.

[0053] Optionally, in this embodiment of the application, the method further includes: performing data integrity and original text consistency verification processing on the extraction results of the data extraction agent and the extraction results of the chart understanding agent based on the verification agent; if the verification fails, performing alloy data extraction again on the extracted content of the first format based on the data extraction agent, and performing feature extraction again on the extracted content of the second format based on the chart understanding agent.

[0054] For example, the extraction results of the data extraction agent and the chart understanding agent can be automatically verified by a verification agent. During verification, the completeness of the extracted data and its consistency with the original text can be checked. If the verification passes, it indicates that the extracted data has high reliability and can be stored in a structured manner. If the verification fails, extraction is restarted until the requirements are met or a certain number of extractions have been reached. This configuration avoids problems such as incomplete or erroneous data extraction, ensuring the accuracy of the extracted data.

[0055] In this embodiment of the application, optionally, the process of constructing the prompt words for the multimodal alloy information extraction model includes: setting data extraction levels for alloy composition, heat treatment process, and mechanical properties for the prompt words of alloy data extraction, and determining the prompt words for alloy data extraction based on the data extraction levels.

[0056] In this embodiment of the application, optionally, the process of constructing prompt words for the multimodal alloy information extraction model includes: for the prompt words of feature extraction, setting the definition of extraction features and corresponding examples for multiple types of charts, and constructing prompt words of feature extraction based on the extraction features and corresponding examples for each chart; the charts include: stress-strain curves, metallographic images and electron microscope images.

[0057] For example, when constructing a prompt, one can define the alloy composition, heat treatment process, mechanical properties, and data extraction hierarchy schema, and define stress-strain curves (such as...). Figure 4 The key indicators to be extracted from the metallographic diagram (as shown) are defined. Figure 5 The characteristics of the grain shape and size, the proportion of porosity and cracks, and the type of inclusions in the electron micrograph (SEM) are used to define the SEM image (e.g., ...). Figure 6 The matrix structure, shape and size of the precipitated phase (as shown) are analyzed, and an extraction result template is constructed to obtain standardized extraction results.

[0058] S250, the extraction results are processed based on the alloy name to obtain the associated data information corresponding to each alloy.

[0059] In this embodiment of the application, optionally, after performing association processing on the extraction results based on the alloy name to obtain the associated data information corresponding to each alloy, the method further includes: performing structured storage on the associated data information corresponding to each alloy.

[0060] For example, the extracted alloy composition, heat treatment process, mechanical properties, and features extracted from charts are associated with the corresponding alloy names and then stored in a structured database. This approach allows for quick retrieval of relevant information based on the alloy name during subsequent searches, improving the convenience of information retrieval.

[0061] Figure 7 is a general flowchart of an embodiment of this application. Figure 7 consists of a flowchart of the preprocessing layer (Figure 7(a), a flowchart of the standardization definition layer (Figure 7(b), a flowchart of the agent application layer (Figure 7(c), and a flowchart of the data post-processing layer (Figure 7(d)). The relationship between the flowcharts is as follows: The initial stage is: PDF document input; the PDF document enters the preprocessing layer shown in Figure 7(a), and after being processed by the preprocessing layer, it enters the standardization definition layer shown in Figure 7(b). After passing through the standardization definition layer, it enters the agent application layer shown in Figure 7(c), and after passing through the agent application layer, it enters the data post-processing layer shown in Figure 7(d), ultimately realizing the update of the knowledge base. The technical solution of this application embodiment is based on a multimodal large model, constructing an extraction format schema and a multi-agent architecture. The extraction agent extracts effective information from alloy material literature PDFs, the graph understanding agent extracts key features from graphs, and the verification agent checks the completeness and consistency of the extraction results. The effective information extracted from the alloy material literature PDFs includes data, descriptive text, and graphs, covering features such as alloy composition, heat treatment processes, grain morphology, matrix, precipitate morphology, and mechanical properties. These elements are linked together to form a complete data record, stored in a structured database, and form a materials knowledge base. This technical solution extracts and links information from "text, tables, and graphs," solving the problems of ineffective utilization of graph information and lack of association with data. The format conversion route designed in this technical solution—"PDF—image—markdown—json—csv—database"—can retain the original document information to the maximum extent. Furthermore, this technical solution can be executed fully automatically, with fast processing speed, requiring no manual intervention. The processing time for a single document is reduced from 30 minutes of manual processing to 3 minutes of automated extraction.

[0062] This solution addresses the limitations of traditional extraction methods, which are only applicable to data from literature on specific alloy materials and lack generalizability. It also resolves the cumbersome process of relying on extensive manual terminology and data standardization definitions, requiring manual intervention to review extraction results and continuously optimize prompts. The solution provides a fully automated, one-click output of the final extraction results, significantly improving efficiency. Furthermore, it addresses the issue of simply parsing text and tables, directly saving charts without processing or extracting effective information, and failing to effectively link text, tables, and charts. This improves the accuracy, completeness, and consistency of data extracted from alloy material PDFs.

[0063] Compared to traditional OCR+regular matching, machine learning, and general large-scale model extraction of alloy data, this solution significantly improves accuracy, achieving over 95% accuracy in table cells. Furthermore, it boasts strong generalization ability: applicable to different journals, document formats, and alloy materials.

[0064] Example 3

[0065] Figure 8 This is a schematic diagram of an alloy data extraction method apparatus provided in Embodiment 3 of this application. This apparatus can execute the alloy data extraction method provided in any embodiment of the present invention, and possesses the corresponding functional modules and beneficial effects for executing the method. Figure 8 As shown, the device includes:

[0066] The content extraction module 310 is used to acquire a PDF file containing alloy data and extract its content to obtain extracted content in a first format and extracted content in a second format.

[0067] The data extraction module 320 is used to extract alloy data and features from the extracted content in the first format and the extracted content in the second format based on the multimodal alloy information extraction model, and obtain the extraction results.

[0068] The data association module 330 is used to perform association processing on the extraction results based on the alloy name to obtain the association data information corresponding to each alloy.

[0069] The technical solution of this application includes: a content extraction module 310, used to acquire a PDF file containing alloy information and extract its content to obtain extracted content in a first format and extracted content in a second format; a data extraction module 320, used to extract alloy data and features from the extracted content in the first format and the extracted content in the second format based on a multimodal alloy information extraction model to obtain extraction results; and a data association module 330, used to associate the extraction results based on alloy names to obtain associated data information corresponding to each alloy. This technical solution can automatically and accurately extract alloy association information from unstructured PDF files without limiting the format of the data in the PDF file. Furthermore, this technical solution is applicable to various alloy materials, improving its generalization ability.

[0070] Optionally, in this embodiment of the application, the content extraction module 310 includes:

[0071] The layout analysis unit is used to acquire PDF files containing alloy data and perform layout analysis on them to obtain the text, table, and chart sections.

[0072] The content extraction unit is used to perform optical character recognition on the text portion and the table portion to obtain extracted content in a first format, and to obtain extracted content in a second format from the chart portion.

[0073] Optionally, in this embodiment of the application, the data extraction module 320 includes:

[0074] The alloy data extraction unit is used to extract alloy data from the extracted content of the first format based on the data extraction agent to obtain alloy data.

[0075] The feature extraction unit is used to extract features from the extracted content of the second format based on the graph understanding agent to obtain key features and / or indicator information.

[0076] Optionally, in this embodiment of the application, the device further includes:

[0077] The verification module is used to verify the data integrity and consistency with the original text based on the extraction results of the data extraction agent and the extraction results of the chart understanding agent.

[0078] The re-extraction module is used to perform alloy data extraction again on the extracted content of the first format based on the data extraction agent if the verification fails, and to perform feature extraction again on the extracted content of the second format based on the chart understanding agent.

[0079] Optionally, in this embodiment of the application, the apparatus further includes: a prompt word construction module, comprising:

[0080] The alloy data extraction prompt word construction module is used to set the data extraction level of alloy composition, heat treatment process, and mechanical properties for the prompt words of alloy data extraction, and to determine the prompt words of alloy data extraction based on the data extraction level.

[0081] Optionally, in this embodiment of the application, the apparatus further includes: a prompt word construction module, comprising:

[0082] The feature extraction prompt word construction module is used to set the definition of the extracted features and corresponding examples for multiple types of charts for the prompt words of feature extraction, and to construct the feature extraction prompt words based on the extracted features and corresponding examples of each chart.

[0083] The chart includes: stress-strain curves, metallographic images, and electron microscope images.

[0084] Optionally, in this embodiment of the application, the device further includes:

[0085] The structured storage module is used to store the associated data information corresponding to each alloy in a structured manner.

[0086] The alloy data extraction method apparatus provided in this application embodiment can execute the alloy data extraction method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method execution.

[0087] Example 4

[0088] Figure 9 A schematic diagram of an electronic device 10, which can be used to implement embodiments of the present invention, is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0089] like Figure 9As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0090] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0091] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as alloy data extraction methods.

[0092] In some embodiments, the alloy data extraction method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the alloy data extraction method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the alloy data extraction method by any other suitable means (e.g., by means of firmware).

[0093] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0094] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0095] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0096] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0097] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0098] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0099] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0100] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for extracting alloy data, characterized in that, include: Obtain a PDF file containing alloy data and extract its content to obtain the extracted content in a first format and a second format. Based on the multimodal alloy information extraction model, alloy data and features are extracted from the content extracted in the first format and the content extracted in the second format to obtain the extraction results. The extraction results are processed based on the alloy name to obtain the associated data information corresponding to each alloy.

2. The method according to claim 1, characterized in that, Obtain a PDF file containing alloy data and extract its content to obtain extracted content in two formats: a first format and a second format. Obtain the PDF file containing alloy data and perform layout analysis to obtain the text, table, and chart sections; Optical character recognition is performed on the text portion and the table portion to obtain the extracted content in the first format, and the extracted content in the second format is obtained from the chart portion.

3. The method according to claim 2, characterized in that, Based on the multimodal alloy information extraction model, alloy data and features are extracted from the content extracted in the first format and the content extracted in the second format to obtain the extraction results, including: Based on the data extraction agent, alloy data is extracted from the content extracted in the first format to obtain alloy data; The graph-based understanding agent extracts features from the content extracted in the second format to obtain key features and / or indicator information.

4. The method according to claim 3, characterized in that, The method further includes: The data integrity and consistency with the original text are verified based on the extraction results of the data extraction agent and the extraction results of the chart understanding agent. If the verification fails, the data extraction agent will perform alloy data extraction again on the extracted content in the first format, and the chart understanding agent will perform feature extraction again on the extracted content in the second format.

5. The method according to claim 1, characterized in that, The process of constructing prompt words for the multimodal alloy information extraction model includes: For the prompts used in alloy data extraction, data extraction levels are set for alloy composition, heat treatment process, and mechanical properties. Based on these data extraction levels, prompts for alloy data extraction are determined.

6. The method according to claim 1, characterized in that, The process of constructing prompt words for the multimodal alloy information extraction model includes: For feature extraction prompts, define the extracted features and corresponding examples for multiple types of charts, and construct feature extraction prompts based on the extracted features and corresponding examples for each chart. The chart includes: stress-strain curves, metallographic images, and electron microscope images.

7. The method according to claim 1, characterized in that, After performing association processing on the extraction results based on alloy names to obtain the associated data information corresponding to each alloy, the method further includes: The associated data information for each alloy is stored in a structured manner.

8. An apparatus for alloy data extraction, characterized in that, include: The content extraction module is used to acquire a PDF file containing alloy data and extract its content to obtain extracted content in a first format and extracted content in a second format. The data extraction module is used to extract alloy data and features from the extracted content in the first format and the extracted content in the second format based on the multimodal alloy information extraction model, and obtain the extraction results. The data association module is used to associate the extraction results based on the alloy name to obtain the associated data information corresponding to each alloy.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the alloy data extraction method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the alloy data extraction method according to any one of claims 1-7.

Images