Intelligent technology dynamic tracking system for science field

Abstract

The invention provides an intelligent technology dynamic tracking system oriented to the scientific field, which relates to the technical field of language models and comprises a full-automatic field literature library maintenance module, an information extraction module based on a language processing model and a technology dynamic development diagram generation module. The full-automatic domain literature library maintenance module is used for performing text conversion on the screened literature by utilizing a data processing tool and storing the converted literature into a scientific domain literature library; the information extraction module based on the language processing model is used for extracting attribute information of literatures in the scientific field; and the technology dynamic development diagram generation module is used for drawing and graphically displaying a technology dynamic development diagram in the scientific field. According to the method, dynamic updating of a domain literature database is realized through a full-automatic crawler tool and a domain paper correlation judgment model, and key points in the literature are extracted and summarized based on a large language model, so that the requirement of researchers for quickly tracking the development of a domain technology is met.

Description

Technical Field

[0001] The present invention relates to the field of language model technology, and in particular to a dynamic tracking system of intelligent technology oriented to the scientific field. Background Art

[0002] At present, in the field of technology dynamics tracking, the most commonly used method is for researchers to conduct manual research and collation by combining literature search tools, knowledge graph methods, and technology trees. Existing literature search tools can realize the automatic retrieval and download functions of academic literature. Some open source tools, such as PySciHub and Scholarly, can also achieve similar functions. The main goal of these systems is to automatically collect relevant literature.

[0003] In terms of knowledge graphs, the ConnectedPapers website provides a graph function for literature association analysis. Users only need to enter the literature information to be analyzed, and the website will automatically generate a grid graph for literature analysis, and visualize information such as the number of citations through elements such as the size and color of the balls, and the thickness and length of the lines.

[0004] The technology tree is a graphical tool used to display the technological evolution and innovation routes in a certain field. It can help researchers understand the path of technological development, key nodes, and possible future research directions.

[0005] With the rapid development of various scientific fields, technology tracking has become an important part of researchers' daily work. Researchers need to constantly track the latest technological advances and research trends to ensure that their research work is "in line" with the highest points in the field. By understanding the latest scientific progress, researchers can identify new research hotspots and breakthroughs, adjust research directions in a timely manner, and make correct decisions in research directions, technology development, and scientific research project planning. However, faced with a huge amount of scientific research literature and information, researchers often find it difficult to quickly track the latest technological advances in the field.

[0006] At present, technical summaries in various scientific fields are usually manually combed by researchers. For example, the field technology tree is a graphical tool for displaying technological evolution and innovation routes. This work is very hard and time-consuming. In order to improve the efficiency of technology tracking, some literature association analysis map tools based on representation models and text similarity have appeared on the market. However, these maps are only drawn based on a single document, and cannot widely cover the literature in the entire field, and usually only provide the context of the development of the field, which requires researchers to still read the content of the article. At the same time, compared with manual combing methods, these tools have failed to significantly speed up the speed of researchers tracking technological development. Therefore, in the face of massive scientific research literature and information, researchers often find it difficult to quickly track the latest technological advances in the field, and the summary of technology in the scientific field is usually manually combed by researchers, which is very laborious and time-consuming.

[0007] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0008] In view of this, the present invention provides an intelligent technology dynamic tracking system for the scientific field to solve the above-mentioned problems.

[0009] In order to solve the above problems, the specific technical solutions adopted by the present invention are as follows: An intelligent technology dynamic tracking system for the scientific field, including: a fully automated field literature library maintenance module, an information extraction module based on a language processing model, and a technology dynamic development map generation module; A fully automated field document library maintenance module is used to screen documents that meet the preset scientific fields through document classifiers, and use data processing tools to convert the screened documents into text and store them in the scientific field document library; An information extraction module based on a language processing model is used to extract attribute information of scientific literature by using the language processing model and combining it with information used to characterize the development attributes and technical points of the scientific field; The technology dynamic development diagram generation module is used to draw and graphically display the technology dynamic development diagram in the scientific field based on the attribute information of the extracted scientific field literature.

[0010] Preferably, the fully automated field document library maintenance module includes: a document crawling submodule, a document processing submodule and a document storage submodule; The document crawling submodule is used to crawl documents from the online document library regularly using a fully automated crawler tool and retain the document abstract text; The document processing submodule is used to screen and classify the crawled documents based on the document abstract text through data processing tools, and convert the documents related to the scientific field into text form; The document storage submodule is used to store the text-converted documents into a pre-set scientific field document library.

[0011] Preferably, the crawled literature is screened and classified by data processing tools, and the literature related to the scientific field is converted into text form including: For the crawled documents, use data processing tools to read the document's meta information; Based on the metadata of the document and the document abstract text as the classification basis, the crawled documents are classified using the document classifier; based on the classification results of the document, the documents that meet the scientific field are screened out; The screened documents that meet the scientific field are converted into text form, and the reference information and random numerical values ​​of the text documents are matched using a pre-set regular expression; the matched reference information and random numerical values ​​are deleted and replaced to obtain the text for storing in the scientific field literature library.

[0012] Preferably, the document classifier is obtained by training a classification model using pre-set document title and abstract data as a training set.

[0013] Preferably, the information extraction module based on the language processing model includes: a representation attribute extraction submodule and a technical detail extraction submodule; The representation attribute extraction submodule is used to analyze the attributes of scientific literature based on the language processing model, and evaluate and select the optimal representation attributes of the literature based on the text similarity calculation results; The technical details extraction submodule is used to obtain the technical details of the document through technical key point analysis based on the language processing model.

[0014] Preferably, based on the language processing model, attribute analysis is performed on scientific documents, and the optimal representation attributes of the documents are evaluated and selected according to the text similarity calculation results, including: Perform word segmentation and segmentation on scientific literature to obtain literature paragraphs; The document paragraph is spliced ​​with the pre-set model extraction prompt words, and input into the language processing model to determine whether the document paragraph contains data of the preset attribute; Perform structured data extraction on the document paragraphs containing preset attribute data to obtain structured representation attributes; Based on the document abstract text of each scientific field document, the cosine similarity is used to calculate the similarity between the document abstract text and the structured representation attribute text; According to the similarity calculation results, the confidence is calculated by multi-dimensional scoring of the summary text and attribute text, and the representation attribute with the highest confidence is selected as the optimal representation attribute.

[0015] Preferably, the formula for calculating the similarity between the document abstract text and the structured attribute representation text using cosine similarity is: ; In the formula, Represents the similarity between the document abstract text and the structured attribute text; E abstract It represents the mapping of the document abstract text in the vector space obtained by encoding the abstract text through the text encoding model, also known as the embedding vector of the abstract text; E property It represents the mapping of structured attribute text in vector space obtained by encoding structured attribute text through text encoding model, also known as embedding vector of structured attribute text.

[0016] Preferably, by scoring the summary text and the attribute text in multiple dimensions, the confidence is calculated as follows: ; Represents the similarity between the document abstract text and the structured attribute text; Represents the multi-dimensional ratings of summary text and attribute text; Indicates confidence; ω 1 represents the weight of the text similarity between the summary text and the attribute text; ω 2 represents the weights of the multi-dimensional scores of the summary text and the attribute text, and the weighted calculation of the confidence is performed based on the two weights using the above formula.

[0017] Preferably, based on the language processing model, through technical point analysis, the technical details of the document include: S1. Encode the literature paragraphs and preset technical key query statements into text vectors through the embedding model; S2, calculating the cosine similarity between the query statement and each document paragraph, and sorting them according to the similarity; S3. Based on the similarity ranking results, select the document paragraphs that meet the preset number as the initial boundary, and determine whether the selected document paragraphs contain the required technical details through the language processing model; S4. If the selected literature paragraph contains the required technical details, the boundary is gradually expanded and S3 is repeated until the model determines that it does not contain technical details and the final text boundary is obtained; S5, splicing the document paragraph text and the abstract text within the final boundary, extracting structured data through the language processing model, and filling the technical detail template to obtain technical detail data; S6. Construct a statement to inquire about the correctness of the technical details, and use the language processing model to determine the correctness of the technical details. If incorrect, make corrections based on the modification suggestions of the model and re-execute S5 until the model verifies the correctness of the technical details.

[0018] Preferably, the technology dynamic development diagram generation module includes: a representation attribute drawing submodule, a technology detail embedding submodule and a technology detail display submodule; The characterization attribute drawing submodule is used to draw the characterization attributes into a technology development curve in chronological order; The technology details embedding submodule is used to embed the technology details data of scientific literature into the technology development graph; The technical details display submodule is used to display the technical details at the corresponding time nodes of the technology development diagram in the form of marks or annotations.

[0019] The beneficial effects of the present invention are: 1. The present invention realizes the dynamic update of the field literature database through a fully automated crawler tool and a field paper relevance judgment model, and extracts and summarizes the key points in the literature based on a large language model. In this way, the latest technologies can be continuously tracked and summarized, and a dynamic development map of the technology in the field can be drawn to meet the needs of researchers to quickly track the development of field technology.

[0020] 2. The present invention realizes the dynamic maintenance of the scientific literature database, which can not only continuously track the latest technological progress, but also ensure the timeliness and comprehensiveness of the literature database, reducing the workload of researchers to manually update and screen literature.

[0021] 3. The present invention uses a large language model to automatically extract and summarize the key technical points in the literature. This process can quickly identify and extract key technical information from a large amount of literature, and integrate it into key points in tabular form. In this way, the system can efficiently extract core technical content from massive data, greatly improving the speed and accuracy of literature analysis.

[0022] 4. The present invention extracts and summarizes a large amount of scientific literature, can quickly draw a dynamic technology development map in the field, and update it according to the dynamic literature library, which not only significantly improves the efficiency of researchers in tracking technological development, but also provides a detailed technological development context, which helps to make more accurate decisions in research directions, technology development and scientific research project planning. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings: Figure 1 is a principle block diagram of a dynamic tracking system for intelligent technology in the scientific field according to an embodiment of the present invention; Figure 2 It is a schematic diagram of the principle of a fully automated document pool maintenance module of a dynamic tracking system of intelligent technology for scientific fields according to an embodiment of the present invention; Figure 3 is a schematic diagram of a structure of optimal attribute extraction in a dynamic tracking system of intelligent technology for scientific fields according to an embodiment of the present invention; Figure 4 It is a structural schematic diagram of technical detail extraction in a dynamic tracking system of intelligent technology for scientific fields according to an embodiment of the present invention.

[0024] In the figure: 1. Fully automated field document library maintenance module; 2. Information extraction module based on language processing model; 3. Technology dynamic development diagram generation module. DETAILED DESCRIPTION

[0025] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0026] According to an embodiment of the present invention, a scientific intelligent technology dynamic tracking system is provided.

[0027] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1-Figure 4 As shown, the intelligent technology dynamic tracking system for the scientific field according to the embodiment of the present invention includes: a fully automated field document library maintenance module 1, an information extraction module 2 based on a language processing model, and a technology dynamic development diagram generation module 3; The fully automated field document library maintenance module 1 is used to screen documents that meet the preset scientific field through a document classifier, and use data processing tools to convert the screened documents into text and store them in the scientific field document library; As a preferred embodiment, the fully automated field document library maintenance module includes: a document crawling submodule, a document processing submodule and a document storage submodule; The document crawling submodule is used to crawl documents from the online document library regularly using a fully automated crawler tool and retain the document abstract text; The document processing submodule is used to screen and classify the crawled documents based on the document abstract text through data processing tools, and convert the documents related to the scientific field into text form; As a preferred implementation, the crawled documents are screened and classified by data processing tools, and the documents related to the scientific field are converted into text forms, including: For the crawled documents, use data processing tools to read the document's meta information; Based on the metadata of the document and the document abstract text as the classification basis, the crawled documents are classified using the document classifier; based on the classification results of the document, the documents that meet the scientific field are screened out; As a preferred embodiment, the document classifier is obtained by training a classification model using pre-set document titles and abstract data as a training set.

[0028] The screened documents that meet the scientific field are converted into text form, and the reference information and random numerical values ​​of the text documents are matched using a pre-set regular expression; the matched reference information and random numerical values ​​are deleted and replaced to obtain the text for storing in the scientific field literature library.

[0029] The document storage submodule is used to store the text-converted documents into a pre-set scientific field document library.

[0030] Specifically, in the fully automated field document library maintenance module, a fully automated document library maintenance strategy is formulated: the fully automated crawler tool will automatically crawl and return specific documents in the field from the network document library at regular intervals; at the same time, the PDF of the document is read using tools such as pymupdf and python, which can read the metadata of the document PDF. The metadata and the document abstract text retained during the crawling process are used as the basis for document classification, and the returned documents are classified using the document classifier to screen out the documents in the corresponding scientific field. If the document is judged to be related to the scientific field by the classifier, it will be automatically added to the document library, thereby dynamically updating the scientific field document library.

[0031] The document classifier is based on the BERT model, which is a classification model trained using pre-processed data of specific document titles and abstracts as well as document titles and abstracts in other fields as training data. Finally, data processing tools such as pymupdf are used to pre-process and clean the documents, including converting the documents from PDF to text format. At the same time, regular expression templates are pre-set, and regular expressions are used to match reference information and messy values ​​in the text. The matched information is deleted and replaced using a python program, and the final text is stored in the document library.

[0032] The information extraction module 2 based on the language processing model is used to extract the attribute information of the scientific field literature by using the language processing model and combining the information used to characterize the development attributes and technical points of the scientific field; As a preferred implementation, the information extraction module based on the language processing model includes: a representation attribute extraction submodule and a technical detail extraction submodule; As a preferred implementation, based on the language processing model, attribute analysis is performed on scientific documents, and the optimal representation attributes of the documents are evaluated and selected according to the text similarity calculation results, including: Perform word segmentation and segmentation on scientific literature to obtain literature paragraphs; The document paragraph is spliced ​​with the pre-set model extraction prompt words, and input into the language processing model to determine whether the document paragraph contains data of the preset attribute; Perform structured data extraction on the document paragraphs containing preset attribute data to obtain structured representation attributes; Based on the document abstract text of each scientific field document, the cosine similarity is used to calculate the similarity between the document abstract text and the structured representation attribute text; As a preferred implementation, the formula for calculating the similarity between the document abstract text and the structured attribute characterization text using cosine similarity is: ; In the formula, Represents the similarity between the document abstract text and the structured attribute text; E abstract It represents the mapping of the document abstract text in the vector space obtained by encoding the abstract text through the text encoding model, also known as the embedding vector of the abstract text; E property The mapping of the structured attribute text encoded by the text encoding model in the vector space is also called the embedding vector of the structured attribute text.

[0033] According to the similarity calculation results, the confidence is calculated by multi-dimensional scoring of the summary text and attribute text, and the representation attribute with the highest confidence is selected as the optimal representation attribute.

[0034] The representation attribute extraction submodule is used to analyze the attributes of scientific literature based on the language processing model, and evaluate and select the optimal representation attributes of the literature based on the text similarity calculation results; As a preferred implementation, by scoring the summary text and the attribute text in multiple dimensions, the confidence calculation formula is: ; Represents the similarity between the document abstract text and the structured attribute text; Represents the multi-dimensional ratings of summary text and attribute text; Indicates confidence; ω 1 represents the weight of the text similarity between the summary text and the attribute text; ω 2 represents the weights of the multi-dimensional scores of the summary text and the attribute text, and the weighted calculation of the confidence is performed based on the two weights using the above formula.

[0035] The technical details extraction submodule is used to obtain the technical details of the document through technical key point analysis based on the language processing model.

[0036] As a preferred implementation, based on the language processing model, through technical point analysis, the technical details of the document include: S1. Encode the literature paragraphs and preset technical key query statements into text vectors through the embedding model; S2, calculating the cosine similarity between the query statement and each document paragraph, and sorting them according to the similarity; S3. Based on the similarity ranking results, select the document paragraphs that meet the preset number as the initial boundary, and determine whether the selected document paragraphs contain the required technical details through the language processing model; S4. If the selected literature paragraph contains the required technical details, the boundary is gradually expanded and S3 is repeated until the model determines that it does not contain technical details and the final text boundary is obtained; S5, splicing the document paragraph text and the abstract text within the final boundary, extracting structured data through the language processing model, and filling the technical detail template to obtain technical detail data; S6. Construct a statement to inquire about the correctness of the technical details, and use the language processing model to determine the correctness of the technical details. If incorrect, make corrections based on the modification suggestions of the model and re-execute S5 until the model verifies the correctness of the technical details.

[0037] Specifically, in the information extraction module based on the language processing model, researchers need to specify the attributes that can characterize the development of a specific scientific field and the corresponding technical points as the information that should be extracted from the literature. The details are as follows: (1) Representation attribute extraction Jieba model and Python program are used to segment documents, calculate the number of tokens in the documents, and segment documents according to the order of 300 tokens. The documents are divided into document paragraphs of 300 words each. The document paragraphs to be extracted are concatenated with the prompt words extracted by the pre-set model to obtain the corresponding string fragments as the input of the language processing model. The model judges whether the segmented document paragraphs contain data (values ​​and units) related to the attributes under discussion based on the above prompts, for example: "Please judge whether the current text contains the {} attribute? Please only answer yes if it does, and answer no if it does not". For the document paragraphs judged as yes, structured data extraction is further performed, that is, template filling, "There is the {} attribute in this paragraph of text. Please extract the attribute and organize it into the following structure: { 'type': '', 'value': '', 'unit': '', } ” Thus, the structured representation attributes are obtained. Among the obtained representation attributes, there are two categories: optimal attributes and test attributes. The optimal attributes are selected based on the similarity between the summary and the structured representation attribute text and the weighted confidence of the model's multi-dimensional scores of the summary text and the attribute text.

[0038] Calculate the similarity between attribute text and summary text based on cosine similarity. First, perform text segmentation, remove stop words, and lowercase. Then load the pre-trained embedding model, which is a pre-trained text encoding model. Then convert the processed text and the vocabulary file that comes with the model into the tokenids required for model input, that is, the corresponding id of the token (word) in the vocabulary, and add special tags [CLS] and [SEP] to the text. After these tokenids are input into the model, the model maps the text to a numerical vector in the vector space, and then calculates the cosine similarity to represent the similarity between the two texts.

[0039] Using the pre-set model scoring prompt words, the language processing model is prompted to perform multi-dimensional scoring on the summary text and the corresponding attribute text to obtain similarity, which includes multiple levels: the semantic similarity between the summary text and the corresponding text, the consistency of the referential attributes in the summary and the attribute text, and the consistency between the attribute and the standard representation attribute. For example, "Please score the semantic similarity between the summary text and the corresponding text based on the given summary text {} and the attribute text {}, as well as the attribute {} extracted from the attribute text, and the score result is within the range of 0-10." is a scoring prompt word for one dimension. After obtaining the scores from the three angles, the average weighted score of the model for the representation attribute extraction result is obtained. In this way, it is evaluated whether the two texts describe the same attribute and whether the attribute is in the format of the representation attribute, and finally the representation attribute in the document is obtained.

[0040] (2) Extraction of technical details Use a query statement with pre-set technical points. For example, in the field of electrolytes, set "What are the components of an electrolyte that achieves a Coulomb efficiency of 99.8%?" as the query statement for the technical points. The statement and the segmented documents are encoded by the embedding model and converted into text vectors. The text similarity is calculated based on cosine similarity, and a text list sorted by similarity is returned. The top 5 paragraphs are selected as the first-level boundary, prompting the language processing model to determine whether the top 5 paragraphs contain the required technical details, that is, "Does the current text contain all {} technical points?" If it does, expand the boundary and select the next paragraph with less similar semantics as the new boundary. Repeat the above steps until the model determines that it does not contain technical details, which is the final text boundary.

[0041] The text of the TopN paragraphs is concatenated with the summary text, and structured data is extracted to fill in the technical point template to obtain the specific data of the technical point. For example, the template that prompts the model to fill in the technical point: "Please find the specific technical point of the attribute {} based on the above text and fill it in the following template: { "name": "", "consist": "", "condiction" : "", } ” The specific details of the template are based on the specified scientific field. At the same time, based on the technical point template, with the technical point name as the identifier, construct a query to inquire about the correctness of other attributes, such as "Based on the above text, is the composition of {name} {consist}? Please output yes or no. If the output is yes, no processing will be done; if the output is no, modification suggestions will be given." In this way, the correctness of the final technical points is judged. If the language processing model returns "no", the modification suggestions output by the language processing model are spliced ​​with the original text, and the steps of extracting structured data of the technical points are repeated. The above steps are repeated until the language processing model returns "yes" to verify that the technical points are correct. If not, make corrections. Repeat this process until the model verifies the correctness of the technical details.

[0042] The technology dynamic development diagram generation module 3 is used to draw and graphically display the technology dynamic development diagram in the scientific field based on the extracted attribute information of the scientific field literature.

[0043] As a preferred implementation, the technology dynamic development diagram generation module includes: a representation attribute drawing submodule, a technology detail embedding submodule and a technology detail display submodule; The characterization attribute drawing submodule is used to draw the characterization attributes into a technology development curve in chronological order; The technology details embedding submodule is used to embed the technology details data of scientific literature into the technology development graph; The technical details display submodule is used to display the technical details at the corresponding time nodes of the technology development diagram in the form of marks or annotations.

[0044] Specifically, in the technology dynamic development diagram generation module, the system generates a technology dynamic development diagram based on the representation attributes and technical details, aiming to intuitively display the technology dynamics in the scientific field in a graphical way. Specifically, the module first plots the representation attributes into a technology development curve in chronological order to show the trend of technology evolution over time. At the same time, the system will embed the specific technical details in the relevant literature into the diagram and display them at the corresponding time nodes in the form of marks or annotations.

[0045] The present invention proposes an intelligent technology dynamic tracking system for the scientific field, which is practical in the electrolyte field, and aims to dynamically track and analyze the technology development trend in the electrolyte field. The system can dynamically obtain the latest literature in the field and extract the key technical points, and realize the dynamic tracking of technology in the electrolyte field. The intelligent technology dynamic tracking system for the electrolyte field consists of three main modules: a fully automated field literature library maintenance module, an information extraction module based on language processing model technology, and a technology dynamic development diagram generation module.

[0046] The present invention constructs a fully automated field literature database dynamic maintenance subsystem, the main purpose of which is to construct and dynamically maintain the literature database in the electrolyte field. Specifically, it consists of two parts: The first is the fully automated literature crawling part. Researchers specify literature keywords in the electrolyte field. The fully automated crawling system will regularly crawl and return specific literature in related fields from the online literature library, and use OCR and Python technologies to pre-process and clean the literature. The second is the judgment of the relevance of the literature field. The present invention implements a BERT-based electrolyte field classification model, which can judge the relevance to the electrolyte field based on the title and abstract of the literature, and process the returned literature using the classification model, thereby dynamically updating the electrolyte field literature library.

[0047] In order to facilitate understanding of the above technical solutions of the present invention, the specific implementation of the present invention in the field of electrolyte is described in detail below.

[0048] The present invention builds an electrolyte information extraction system based on language processing model technology. Its main purpose is to extract coulombic efficiency, capacity retention rate, cycle life and other properties that can characterize the development of the electrolyte field and the corresponding technical points - electrolyte formula from the literature.

[0049] The specific process is as follows: 1. Optimal Attribute Extraction Module 1. Process the text of the document in segments and input them into the language processing model in turn to determine whether they have the specified attribute value. If they do, extract them and save the paragraph text corresponding to the attribute value. If they do not have the attribute value, skip them and organize them into the following format: { 'type': 'coulombic efficiency', 'value': '98.2% within 450 cycles', 'unit': None, 'sentence_id': 2 } 2. Filter the top 5 attribute values ​​and their corresponding paragraph texts, and calculate the similarity between the paragraph text corresponding to each attribute value and the abstract text. At the same time, use the language processing model to determine the similarity between the paragraph text corresponding to the attribute value and the optimal attribute of the document mentioned in the abstract text. If it is determined to be similar, a score of 1-10 is given to describe the similarity. If it is determined to be dissimilar, the score is 0. The confidence of the optimal attribute is finally determined according to the specific weight combination similarity and the judgment score of the language processing model. The attribute with the highest confidence is selected as the optimal attribute, which is regarded as representing the technical development brought by the technology of this document.

[0050] 2. Technical Points Extraction Module 1. Calculate the similarity between the description of the electrolyte composition and the paragraphs in the literature, return the top 5 most similar paragraphs, and use the language processing model to determine whether the last paragraph contains the composition of the electrolyte. If it does, then determine the top 6 paragraph, and repeat this process until the model determines that it does not contain the composition.

[0051] 2. Connect the description of the electrolyte composition with the largest paragraph containing the electrolyte composition found by the language processing model, and use the language processing model to extract the structured expression of the electrolyte composition. The example is as follows: { "name": "CNE-30", "solute": "LiFSI", "solute concentration" : "2 M", "solvents": "EC / DMC", "diluents": "CTAB", "proportion": "EC / DMC 3:7 by volume; CTAB 30mM" } 3. The large language model verifies the correctness of the extracted electrolyte composition by using a preset list that can verify the correctness of the answer to the question, that is, a list of questions asking whether the electrolyte composition is correct, such as "Is the salt solute of the CNE-30 electrolyte LiFSI?", to verify whether there is an illusion problem in the electrolyte composition. The verification answer, the description statement of the electrolyte composition and the largest paragraph containing the electrolyte composition found by the language processing model are provided to the language processing model for correction or confirmation of the answer, and the cycle is repeated until the verification stage is passed to obtain the final electrolyte composition.

[0052] The present invention constructs a module for generating a technical dynamic development diagram in the field of electrolytes based on optimal properties and technical key points. Its main purpose is to graphically display the technical dynamics in the field of electrolytes. Specifically, it consists of two parts: The first is the technology development trend display part, which presents the technology dynamics in the electrolyte field in a visual form by drawing a scattered line graph of technology development trends. Each point in the line graph corresponds to the electrolyte properties and technical points of a document; The second is the technical information display section, which displays the specific technical details of the corresponding literature. Researchers can quickly browse the detailed information of each technical point, so as to fully understand the technical development status in the electrolyte field and the technical details of the latest technology.

[0053] In summary, with the help of the above technical solutions of the present invention, the present invention realizes the dynamic update of the field literature database through a fully automated crawler tool and a field paper relevance judgment model, and extracts and summarizes the key points in the literature based on a large language model, so that the latest technology can be continuously tracked and summarized, and a technical dynamic development diagram of the field can be drawn to meet the needs of researchers to quickly track the development of field technology. The present invention realizes the dynamic maintenance of the scientific field literature database, which can not only continuously track the latest technological progress, but also ensure the timeliness and comprehensiveness of the literature database, reducing the workload of researchers to manually update and screen literature. The present invention uses a large language model to automatically extract and summarize the technical key points in the literature. This process can quickly identify and refine key technical information from a large number of documents, and integrate it into key points in the form of a table. In this way, the system can efficiently extract core technical content from massive data, greatly improving the speed and accuracy of literature analysis. The present invention extracts and summarizes a large amount of scientific literature, can quickly draw a dynamic technology development map in the field, and update it according to the dynamic literature library, which not only significantly improves the efficiency of researchers in tracking technological development, but also provides a detailed technological development context, which helps to make more accurate decisions in research directions, technology development and scientific research project planning.

[0054] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, optical storage, etc.) containing computer-usable program code.

[0055] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A dynamic tracking system for intelligent technology in the scientific field, characterized in that: include: Fully automated field document library maintenance module, information extraction module based on language processing model, and technology dynamic development diagram generation module; The fully automated domain document library maintenance module is used to screen documents that meet the preset scientific fields through a document classifier, and use a data processing tool to convert the screened documents into text and store them in the scientific domain document library; The information extraction module based on the language processing model is used to extract attribute information of scientific field literature by using the language processing model and combining it with information used to characterize the development attributes and technical points of the scientific field; The technology dynamic development diagram generation module is used to draw and graphically display the technology dynamic development diagram in the scientific field based on the extracted attribute information of the scientific field literature.

2. According to claim 1, a dynamic tracking system for intelligent technology in the scientific field is characterized in that: The fully automated domain document library maintenance module includes: a document crawling submodule, a document processing submodule and a document storage submodule; The document crawling submodule is used to crawl documents from the network document library regularly using a fully automated crawler tool and retain document abstract texts; The document processing submodule is used to screen and classify the crawled documents based on the document abstract text through a data processing tool, and convert the documents related to the scientific field into text form; The document storage submodule is used to store the text-converted documents into a pre-set scientific field document library.

3. The intelligent technology dynamic tracking system for the scientific field according to claim 2 is characterized in that: The method of screening and classifying the crawled documents by using data processing tools and converting the documents related to the scientific field into text form includes: For the crawled documents, use data processing tools to read the document's meta information; Based on the metadata of the document and the document abstract text as the classification basis, the crawled documents are classified using the document classifier; based on the classification results of the document, the documents that meet the scientific field are screened out; The screened documents that meet the scientific field are converted into text form, and the reference information and random numerical values ​​of the text documents are matched using a pre-set regular expression; the matched reference information and random numerical values ​​are deleted and replaced to obtain the text for storing in the scientific field literature library.

4. The intelligent technology dynamic tracking system for the scientific field according to claim 3 is characterized in that: The document classifier is obtained by training a classification model using pre-set document titles and abstract data as a training set.

5. The intelligent technology dynamic tracking system for the scientific field according to claim 1 is characterized in that: The information extraction module based on the language processing model includes: a representation attribute extraction submodule and a technical detail extraction submodule; The representation attribute extraction submodule is used to perform attribute analysis on scientific documents based on the language processing model, and evaluate and select the optimal representation attribute of the document according to the text similarity calculation results; The technical details extraction submodule is used to obtain the technical details of the document through technical key point analysis based on the language processing model.

6. The intelligent technology dynamic tracking system for the scientific field according to claim 5, characterized in that: The above-mentioned language processing model is used to analyze the attributes of scientific documents, and the optimal representation attributes of the documents are evaluated and selected according to the similarity calculation results of the texts, including: Perform word segmentation and segmentation on scientific literature to obtain literature paragraphs; The document paragraph is spliced ​​with the pre-set model extraction prompt words, and input into the language processing model to determine whether the document paragraph contains data of the preset attribute; Perform structured data extraction on the document paragraphs containing preset attribute data to obtain structured representation attributes; Based on the document abstract text of each scientific field document, the cosine similarity is used to calculate the similarity between the document abstract text and the structured representation attribute text; According to the similarity calculation results, the confidence is calculated by multi-dimensional scoring of the summary text and attribute text, and the representation attribute with the highest confidence is selected as the optimal representation attribute.

7. The intelligent technology dynamic tracking system for the scientific field according to claim 6 is characterized in that: The calculation formula for calculating the similarity between the document abstract text and the structured attribute text using cosine similarity is: ； In the formula, Represents the similarity between the document abstract text and the structured attribute text; E abstract Represents the mapping of the document abstract text encoded by the text encoding model in the vector space; E property Represents the mapping of structured representation attribute text in vector space obtained by encoding structured representation attribute text through text encoding model.

8. The intelligent technology dynamic tracking system for the scientific field according to claim 6 is characterized in that: The confidence is calculated by scoring the summary text and the attribute text in multiple dimensions as follows: ； Represents the similarity between the document abstract text and the structured attribute text; Represents the multi-dimensional ratings of summary text and attribute text; Indicates confidence; ω 1 represents the weight of the text similarity between the summary text and the attribute text; ω 2 represents the weights of the multi-dimensional scores of the summary text and the attribute text, and the weighted calculation of the confidence is performed based on the two weights using the above formula.

9. The intelligent technology dynamic tracking system for the scientific field according to claim 1 is characterized in that: Based on the language processing model, the technical details of the document obtained through technical point analysis include: S1. Encode the literature paragraphs and preset technical key query statements into text vectors through the embedding model; S2, calculating the cosine similarity between the query statement and each document paragraph, and sorting them according to the similarity; S3. Based on the similarity ranking results, select the document paragraphs that meet the preset number as the initial boundary, and determine whether the selected document paragraphs contain the required technical details through the language processing model; S4. If the selected literature paragraph contains the required technical details, the boundary is gradually expanded and S3 is repeated until the model determines that it does not contain technical details and the final text boundary is obtained; S5, splicing the document paragraph text and the abstract text within the final boundary, extracting structured data through the language processing model, and filling the technical detail template to obtain technical detail data; S6. Construct a statement to inquire about the correctness of the technical details, and use the language processing model to determine the correctness of the technical details. If incorrect, make corrections based on the modification suggestions of the model and re-execute S5 until the model verifies the correctness of the technical details.

10. The intelligent technology dynamic tracking system for the scientific field according to claim 1, characterized in that: The technology dynamic development diagram generation module includes: a representation attribute drawing submodule, a technology detail embedding submodule and a technology detail display submodule; The characterization attribute drawing submodule is used to draw the characterization attributes into a technology development curve in chronological order; The technical details embedding submodule is used to embed the technical details data of scientific field literature into the technology development map; The technical details display submodule is used to display the technical details at the corresponding time nodes of the technology development diagram in the form of marks or annotations.

Images