A mineralization gene analysis method and system based on expert knowledge base

Abstract

The present application relates to the technical field of mineralization gene analysis, and in particular to a mineralization gene analysis method and system based on an expert knowledge base. The method comprises: obtaining authoritative theoretical information, analyzing the authoritative theoretical information, and determining a mineralization gene characteristic information set; performing a directional association analysis on mineralization genes based on the mineralization gene characteristic information set, and determining a pointing node set and a target node set; constructing a mineralization gene structure information library based on the pointing node set and the target node set; obtaining a mining area exploration data set, and analyzing the mining area exploration data set based on the mineralization gene structure information library to determine a valid node information set; analyzing the valid node information set and the mining area exploration data set based on the mineralization gene structure information library to determine and output a mineralization status analysis report. The present application enables users to better manage resources and make decisions, ensure sustainable development and utilization of mineral resources, and greatly improve the scientificity and economy of mineral exploration.

Description

Technical Field

[0001] The present application relates to the technical field of metallogenic gene analysis, and in particular to a metallogenic gene analysis method and system based on an expert knowledge base. Background Art

[0002] Metallogenic genes (also known as ore genesis) refer to the comprehensive set of geological, geochemical, and geophysical characteristics and their interactions associated with the formation of a mineral deposit. Metallogenic genes are crucial for understanding and describing the formation process of a mineral deposit, including the source, migration, and enrichment of ore-forming materials, as well as the geological environment and conditions of mineralization. Analysis of metallogenic genes provides an important scientific basis for mineral resource assessment.

[0003] Traditional mineralization gene analysis methods often rely on the personal experience and knowledge of experts, and have problems of strong subjectivity and low efficiency. However, with the development of geological exploration technology, the types of data collected are diverse and involve many fields. The complexity and diversity of these data make it difficult for traditional mineralization gene analysis methods to efficiently process and comprehensively analyze, resulting in low accuracy of mineralization gene analysis results. Summary of the Invention

[0004] This application provides a mineralization gene analysis method and system based on an expert knowledge base to solve the above technical problems.

[0005] In a first aspect, the present application provides a method for analyzing metallogenic genes based on an expert knowledge base, the method comprising:

[0006] Obtaining authoritative theoretical information, analyzing the authoritative theoretical information, and determining a metallogenic gene characteristic information set;

[0007] According to the metallogenic gene characteristic information set, a directional association analysis is performed on the metallogenic genes to determine a directional node set and a target node set;

[0008] Constructing a mineralization gene structure information library according to the pointing node set and the target node set;

[0009] Acquire a mining area exploration data set, analyze the mining area exploration data set based on the metallogenic gene structure information database, and determine a valid node information set;

[0010] Based on the metallogenic gene structure information database, the effective node information set and the mining area exploration data set are analyzed to determine and output a metallogenic status analysis report.

[0011] Through this solution, the authoritative theoretical information and mining area exploration data sets are combined in the process of analyzing mineralization genes, which significantly improves the accuracy and efficiency of mining area exploration, identifies and analyzes key mineralization characteristics, optimizes exploration strategies, ensures the rational allocation of resources, and provides systematic data support for the process of constructing a mineralization gene structure information database, so that the exploration process can quickly identify effective ore body characteristics, help reduce exploration risks, significantly improve exploration success rates, reduce unnecessary mining and resource waste, and thus effectively save costs. By outputting detailed mineralization status analysis reports, users can better manage resources and make decisions, ensure sustainable mineral resource development and utilization, and greatly improve the scientific and economic nature of mineral exploration.

[0012] Optionally, the authoritative theoretical information includes geological example image information and theoretical text information, and the analyzing the authoritative theoretical information to determine the metallogenic gene feature information set includes:

[0013] According to the geological example image information, performing a standardized vector dimensionality reduction process on the geological example image information to determine a geological structure feature vector set;

[0014] Analyzing the theoretical text information, performing entity recognition screening on the theoretical text information, and extracting a key text segment information set;

[0015] According to the key text segment information set, performing text feature vector analysis on the key text segment information set to determine a text feature vector set;

[0016] The metallogenic gene feature information set is constructed based on the geological structure feature vector set and the text feature vector set.

[0017] Through this scheme, feature analysis and extraction are performed on the geological example image information and theoretical text information in the authoritative theoretical information, so as to construct a geological structure feature vector information set and a key text segment information set, so as to facilitate the computer-level automated analysis of the authoritative theoretical information, while enhancing data quality, reducing data redundancy, and improving the efficiency and accuracy of subsequent feature analysis.

[0018] Optionally, performing standardized vector dimensionality reduction processing on the geological example image information to determine a geological structure feature vector set according to the geological example image information includes:

[0019] Analyzing the geological example image information and extracting the image mean and image standard deviation of each geological example image;

[0020] Based on the image mean and the image standard deviation, a feature vector dimensionality reduction extraction process is performed on each geological example image to determine the effective image feature vector corresponding to each geological example image, specifically the following formula:

[0021]

[0022] Among them, E i is the effective image feature vector corresponding to the i-th geological example image, PCA() is a preset dimensionality reduction function, CNN() is a preset convolution function, I i is the ith geological example image, mean(I i is the image mean of the i-th geological example image, std(I i ) is the image standard deviation of the i-th geological example image;

[0023] The geological structure feature vector set is constructed according to the effective image feature vector corresponding to each geological example image.

[0024] Through this solution, mathematical analysis methods are used to perform standardized vector dimensionality reduction processing on the image information of different geological examples, and mathematical formulas are used to automatically and accurately extract the effective image feature vectors in the geological example images. While effectively capturing the features within the geological example images, high-dimensional features are reduced to low dimensions, redundant features are removed, the burden of subsequent computational analysis is reduced, and the generalization ability of the subsequent analysis process is improved.

[0025] Optionally, performing text feature vector analysis on the key text segment information set according to the key text segment information set to determine the text feature vector set includes:

[0026] Analyzing the key text segment information set to extract a keyword set for each key text segment;

[0027] Based on the keyword set and the key text segment information set, an aggregate vector analysis is performed on each key text segment to extract the aggregate text vector corresponding to each key text segment, specifically the following formula:

[0028]

[0029] Among them, G m is the aggregated text vector corresponding to the mth key text segment, n is the total number of keywords in the keyword set corresponding to the current key text segment, V j is the jth keyword in the current keyword set, and Embed() is a preset word embedding function;

[0030] The text feature vector set is constructed according to the aggregated text vector corresponding to each key text segment.

[0031] Through this scheme, the keywords corresponding to each key text segment in the key text segment information are extracted to obtain the keyword set corresponding to each key text segment. On this basis, through mathematical analysis means, an aggregation vector analysis is performed on each key text segment to quantify the aggregation text vector corresponding to each key text segment. According to the aggregation text vector corresponding to each key text segment, a text feature vector set is constructed, so that the text feature vector set can comprehensively reflect the different semantic features in the theoretical text information, and provide a scientific data basis for the subsequent mineralization gene analysis process.

[0032] Optionally, performing a directional association analysis on the metallogenic genes according to the metallogenic gene feature information set to determine a directional node set and a target node set includes:

[0033] Analyzing the metallogenic gene feature information set to determine the feature complexity of each metallogenic gene feature;

[0034] Analyzing the metallogenic gene feature information set to determine the feature space center point vector;

[0035] Based on the feature complexity and the feature space center point vector, the mineralization gene feature information set is subjected to binary screening to determine the pointing node set and the target node set, respectively. Specifically, the mathematical expression is as follows:

[0036]

[0037] Among them, S T is the target node set, C p is the pth metallogenic gene feature in the metallogenic gene feature information set, β k is the preset mapping coefficient corresponding to the kth feature power, d is the preset target bias coefficient, w p is the characteristic complexity of the pth metallogenic gene feature, θ T is the preset target evaluation threshold, γ is the preset adjustment parameter, μ is the center point vector of the feature space, b is the preset pointing bias coefficient, θ I The preset pointing evaluation threshold.

[0038] Through this scheme, mathematical analysis methods are used, based on the feature space center point vector and the feature complexity of each metallogenic gene feature, and through clear mathematical expressions, binary screening of the metallogenic gene features in the metallogenic gene feature information set is performed to determine the target nodes and pointing nodes in the metallogenic gene feature information set, improve the accuracy and scientificity of node judgment, and provide an important data foundation for the subsequent construction of a structured information database.

[0039] Optionally, constructing a metallogenic gene structure information database according to the pointing node set and the target node set includes:

[0040] Analyzing the pointing node set and the target node set to determine the pointing association degree between each pointing node and each target node;

[0041] Based on the pointing association degree, a feature-weighted directed graph is constructed according to the pointing node set and the target node set;

[0042] The feature-weighted directed graph is used as the mineralization gene structure information database.

[0043] Through this scheme, on the basis of the pointing node set and the target node set, according to the pointing correlation between different pointing nodes and target nodes, a feature-weighted finite graph is constructed with the weighted directed graph as the basic data structure, and the feature-weighted directed graph is used as the metallogenic gene structure information library, so that the computer can quickly parse the node information in the metallogenic gene structure information library, providing structured data support for subsequent metallogenic analysis.

[0044] Optionally, constructing a feature-weighted directed graph based on the pointing association degree and according to the pointing node set and the target node set includes:

[0045] According to the directional correlation degree, the feature relationship adjacency matrix is ​​determined, which is specifically expressed as follows:

[0046]

[0047] Among them, A xy is the feature relationship adjacency matrix, Con(x,y) is the directional correlation between the xth pointing node and the yth target node, T x is the eigenvector corresponding to the x-th pointing node, R y is the feature vector corresponding to the y-th target node, and δ is the preset relationship threshold;

[0048] The feature weighted directed graph is constructed by taking the feature vectors corresponding to each node in the pointing node set and the target node set as directed graph nodes and taking each pointing association degree in the feature relationship adjacency matrix as a directed graph path.

[0049] Through this scheme, mathematical analysis methods are used to quantify the pointing correlation between different nodes, and through mathematical expressions, a characteristic adjacency matrix is ​​constructed based on the pointing correlation to reflect the relationship structure between all pointing nodes and target nodes. Based on the characteristic adjacency matrix, a characteristic weighted directed graph is constructed according to the pointing node set and the target node set, so that the characteristic weighted directed graph can clearly reflect the degree of correlation between different nodes, providing a scientific data basis for subsequent mineralization analysis.

[0050] Optionally, analyzing the mining area exploration dataset based on the metallogenic gene structure information database to determine the valid node information set includes:

[0051] Analyzing the mining area exploration data set to extract exploration gene feature information;

[0052] Matching the exploration gene feature information with the feature weighted directed graph, and determining an associated node set and an associated weight set based on the matching result;

[0053] Analyzing the mining area exploration data set based on the associated node set to extract a set of gene indicators;

[0054] According to the association weight set and the gene indicator set, the association node set is screened for effectiveness to determine the effective node information set.

[0055] Through this solution, the information in the mining exploration data set is converted into corresponding feature vector information, and the exploration gene feature information is matched with the feature weighted directed graph to obtain a highly correlated association node information set and a corresponding association weight set. On this basis, the effectiveness of the associated node information set is screened through the gene indicator set, and the effective node information set is determined to achieve rapid and automated matching of mining exploration data, further enhancing data accuracy and reducing data redundancy.

[0056] Optionally, the analyzing the valid node information set and the mining area exploration data set based on the metallogenic gene structure information database to determine and output a metallogenic status analysis report includes:

[0057] Analyze the mining area exploration data set to determine the regional distribution information of metallogenic genes;

[0058] Analyzing the metallogenic gene structure information database according to the valid node information set to determine the corresponding mining area target node set in the metallogenic gene structure information database;

[0059] Based on the regional distribution information of the metallogenic genes, analyzing the target node set of the mining area to determine the distribution information of the metallogenic area;

[0060] According to the mineralization area distribution information, the mineralization status analysis report is determined and output.

[0061] Through this solution, based on the distribution of different mineralization genes in different areas of the mining area, the mineralization status of the target minerals corresponding to the target nodes in the mining area within the effective nodes are analyzed in different areas of the mining area to construct the mineralization area distribution information corresponding to the current mining area, and according to the distribution of different target minerals in different areas within the mineralization area distribution information, the mineralization status analysis report is determined and output, so that the mineralization status analysis report clearly reflects the changes in the distribution of different target minerals between different areas of the mining area, thereby enabling users to accurately locate the mineralization-intensive areas within the mining area and provide data support for the formulation of subsequent mineral development plans.

[0062] In a second aspect, the present application provides a mineralization gene analysis system based on an expert knowledge base, the system comprising:

[0063] A feature analysis module is used to obtain authoritative theoretical information, analyze the authoritative theoretical information, and determine the metallogenic gene feature information set;

[0064] A node analysis module is used to perform a directional association analysis on the metallogenic genes according to the metallogenic gene feature information set, and determine a directional node set and a target node set;

[0065] A node construction module is used to construct a metallogenic gene structure information library according to the pointing node set and the target node set;

[0066] A node screening module is used to obtain a mining area exploration data set, analyze the mining area exploration data set based on the metallogenic gene structure information database, and determine a valid node information set;

[0067] The status output module is used to analyze the valid node information set and the mining area exploration data set based on the metallogenic gene structure information library, and determine and output a metallogenic status analysis report. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0069] Figure 1 A schematic diagram of an application scenario provided in one embodiment of the present application;

[0070] Figure 2 A flowchart of a method for analyzing mineralization genes based on an expert knowledge base provided in one embodiment of the present application;

[0071] Figure 3A schematic structural diagram of a mineralization gene analysis system based on an expert knowledge base provided in one embodiment of the present application. DETAILED DESCRIPTION

[0072] To make the purpose, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0073] In this document, the term "and / or" simply describes a relationship between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this document, unless otherwise specified, generally indicates an "or" relationship between the related objects.

[0074] The embodiments of the present application are described in further detail below with reference to the accompanying drawings.

[0075] With the development of geological exploration technology, the types of data collected are diverse and involve many fields. The complexity and diversity of these data make it difficult for traditional mineralization gene analysis methods to be efficiently processed and comprehensively analyzed, resulting in low accuracy of mineralization gene analysis results.

[0076] Based on this, the present application provides a mineralization gene analysis method and system based on an expert knowledge base. In the process of analyzing mineralization genes, authoritative theoretical information and mining area exploration data sets are combined to significantly improve the accuracy and efficiency of mining area exploration, identify and analyze key mineralization characteristics, optimize exploration strategies, ensure the rational allocation of resources, and build a mineralization gene structure information database to provide systematic data support, so that the exploration process can quickly identify effective ore body characteristics, help reduce exploration risks, significantly improve exploration success rates, reduce unnecessary mining and resource waste, thereby effectively saving costs. By outputting detailed mineralization status analysis reports, users can better manage resources and make decisions, ensure sustainable mineral resource development and utilization, and greatly improve the scientificity and economy of mineral exploration.

[0077] Figure 1 This is a schematic diagram of an application scenario provided by this application. In the process of analyzing mineralization genes, the method provided by this application is applied to automatically analyze the mining area exploration dataset and authoritative theoretical information, and provide users with a mineralization status analysis report that comprehensively reflects the mineralization status of each area within the mining area.

[0078] Specifically, the method of the present application is applied to any server, which communicates with the geological literature disclosure system, and obtains the authoritative theoretical information provided by the geological literature disclosure system and the mining area exploration information set provided by the user through the server. The authoritative theoretical information and the mining area exploration data set are combined in the process of analyzing the mineralization genes, which significantly improves the accuracy and efficiency of mining area exploration, identifies and analyzes key mineralization characteristics, optimizes exploration strategies, ensures the rational allocation of resources, and constructs a mineralization gene structure information database to provide systematic data support, so that the exploration process can quickly identify effective ore body characteristics, help reduce exploration risks, significantly improve exploration success rates, reduce unnecessary mining and resource waste, thereby effectively saving costs. By outputting detailed mineralization status analysis reports, users can better manage resources and make decisions, ensure sustainable mineral resource development and utilization, and greatly improve the scientificity and economy of mineral exploration.

[0079] For specific implementation methods, please refer to the following embodiments.

[0080] Figure 2 This is a flow chart of a method for analyzing mineralization genes based on an expert knowledge base provided in one embodiment of this application. The method of this embodiment can be applied to the server in the above scenario. Figure 2 As shown, the method includes:

[0081] S201. Obtain authoritative theoretical information, analyze the authoritative theoretical information, and determine a mineralization gene characteristic information set.

[0082] The authoritative theoretical information may be authoritative theoretical information on mineralization analysis in the field of geology, and the authoritative theoretical information may be obtained from a geological literature disclosure system.

[0083] The metallogenic gene characteristic information set may be a set of information on metallogenic gene characteristics extracted from authoritative theoretical information.

[0084] Specifically, authoritative theoretical information is the basis for constructing a scientific and reasonable mineralization model. These theoretical information have been publicly verified, ensuring the accuracy of the analysis of the complex mechanism of mineral formation. Through mathematical analysis, authoritative theoretical information is analyzed, the mineralization gene characteristics are extracted and quantified, and a mineralization gene characteristic information set is constructed. This can provide an overall understanding of the mineralization process, mineralization environment and mineral distribution. The determination of the mineralization gene characteristic information set is an important prerequisite for further analysis. It provides key factors and indicators in the mineralization process, helps to identify and establish the mineralization pattern and characteristics of the target mineral, ensures that subsequent analysis has solid theoretical and data support, and improves the accuracy and reliability of mineralization gene analysis.

[0085] S202. Based on the metallogenic gene characteristic information set, a directional association analysis is performed on the metallogenic genes to determine a directional node set and a target node set.

[0086] Directional association analysis can be an analysis process of the degree of directional association of different mineralization genes with target minerals.

[0087] The pointing node set may be a node set corresponding to the mineralization gene pointing to the target mineral.

[0088] The target node set may be a node set corresponding to the target mineral.

[0089] Specifically, mathematical analysis is used to conduct directional association analysis on different mineralization genes in the mineralization gene characteristic information set, and the evaluation criteria of the pointing nodes and target nodes are quantified. The main purpose of the directional association analysis is to determine the specific connection and interaction between the mineralization genes and the target minerals. This step can identify which gene characteristics have direct or indirect correlation with the target minerals by constructing a relationship model between the mineralization genes and the geological bodies, and clarify the pointing nodes and target nodes of the mineralization gene characteristic information set. The determination of the pointing node set and the target node set can clarify the direction in the subsequent construction process of the corresponding information database, effectively filter the mineralization information, highlight the key points, facilitate the subsequent rapid locking of valuable mining areas, and improve the success rate of exploration.

[0090] S203. Construct a mineralization gene structure information database based on the pointing node set and the target node set.

[0091] The metallogenic gene structure information database can be a structured data set containing different metallogenic gene characteristics.

[0092] Specifically, the pointing node set and the target node set are automatically structured to construct a metallogenic gene structure information database based on computer data structure. The metallogenic gene structure information database is the core tool in the entire metallogenic gene analysis process. It structures various nodes and corresponding data to form a systematic database. The metallogenic gene structure information database helps to store and retrieve information efficiently, and also makes the next step of data analysis more convenient and automated. The construction of the metallogenic gene structure information database can standardize data from different sources and properties, and support dynamic updates to adapt to the latest progress in geological and ore body research. The result of this step is the formation of a knowledge base that can be applied in real time, providing solid data support for subsequent analysis of the mineralization status of the mining area.

[0093] S204: Obtain a mining area exploration dataset, analyze the mining area exploration dataset based on the metallogenic gene structure information database, and determine a valid node information set.

[0094] The mining area exploration data set may be a set of geological exploration data within the mining area targeted for analysis by the current user, and the mining area exploration data set may be provided by the user.

[0095] The valid node information set may be a set of pointing nodes with analysis value within the current user target analysis area.

[0096] Specifically, the acquisition and analysis of mining exploration data sets is an important part of the application of metallogenic genes to actual mining areas. By combining on-site exploration data with the metallogenic gene structure information database, metallogenic characteristics that are consistent or different in theoretical and actual exploration can be identified. The determination of effective node information sets helps focus on key geological parameters and improve the accuracy of mining area analysis and prediction. This step is equivalent to the verification and screening of the metallogenic gene structure, which improves the effectiveness of the model in actual mining area application and the direction of improvement.

[0097] S205. Based on the metallogenic gene structure information database, analyze the effective node information set and the mining area exploration data set, determine and output the metallogenic status analysis report.

[0098] The mineralization status analysis report can be the current user's target analysis of mineralization status information in different areas within the mining area.

[0099] Specifically, by analyzing the metallogenic gene structure information database, the target node information corresponding to the valid node information set in the mining area exploration data set is extracted to reflect the metallogenic status corresponding to the target node. The data visualization technology is used to construct the metallogenic status analysis report, and the metallogenic status analysis report is provided to the user through the human-computer interaction device. The final metallogenic status analysis report is the analysis result obtained by integrating the aforementioned steps. By comprehensively analyzing the metallogenic gene structure information database, the valid node information set and the mining area exploration data, a comprehensive assessment report on the mineralization possibility, resource potential, mining value and even environmental impact of the mining area is obtained. The output of this part is not only the basis for the formulation of mineral exploration strategies, but also provides a scientific basis for investment and management decisions. Having a detailed and comprehensive metallogenic status analysis report can provide clear guidance for each stage of the mining development process, while improving the efficiency of exploration and development, reducing user blindness and resource waste.

[0100] Through this solution, the authoritative theoretical information and mining area exploration data sets are combined in the process of analyzing mineralization genes, which significantly improves the accuracy and efficiency of mining area exploration, identifies and analyzes key mineralization characteristics, optimizes exploration strategies, ensures the rational allocation of resources, and provides systematic data support for the process of constructing a mineralization gene structure information database, so that the exploration process can quickly identify effective ore body characteristics, help reduce exploration risks, significantly improve exploration success rates, reduce unnecessary mining and resource waste, and thus effectively save costs. By outputting detailed mineralization status analysis reports, users can better manage resources and make decisions, ensure sustainable mineral resource development and utilization, and greatly improve the scientific and economic nature of mineral exploration.

[0101] In some embodiments, based on the geological example image information, the geological example image information is subjected to standardized vector dimensionality reduction processing to determine the geological structure feature vector set; the theoretical text information is analyzed, the theoretical text information is subjected to entity recognition screening, and the key text segment information set is extracted; based on the key text segment information set, the key text segment information set is subjected to text feature vector analysis to determine the text feature vector set; based on the geological structure feature vector set and the text feature vector set, a mineralization gene feature information set is constructed.

[0102] Authoritative theoretical information includes geological example image information and theoretical text information.

[0103] The geological example image information may be an example image used to assist in explaining and expanding the authoritative theoretical information.

[0104] The theoretical text information may be text information in authoritative theoretical information used to elaborate on geological gene analysis.

[0105] The normalized vector dimensionality reduction process may be a process of performing normalized dimensionality reduction on vector information used to describe geological features.

[0106] The geological structure feature vector set may be a set of corresponding vector information containing geological structure features reflected in geological example image information.

[0107] Entity recognition and screening can be the process of identifying and screening entity information that maps real factors in theoretical text information.

[0108] The key text segment information set may be a set of text segments in the text information in the authoritative theoretical information that are highly relevant to the mineralization genes.

[0109] The text feature vector set may be a set of feature vectors corresponding to different text paragraph contents in the key text segment information set.

[0110] Specifically, geological example image information and theoretical text information are both forms of information conveyed in authoritative theoretical information that are easy for people to understand. In order to enable the above-mentioned forms of information to be automatically parsed by computer models, feature extraction algorithms in image processing technology, such as edge detection algorithms and texture analysis algorithms, are combined with feature dimension reduction algorithms in image processing technology, such as principal component analysis algorithms, to extract feature vectors from geological example images and perform dimension reduction processing on them to construct a geological structure feature vector set, thereby enhancing data quality and reducing data redundancy. At the same time, entity recognition technologies in natural language processing technology, such as named entity recognition technology, are used to extract key text segments from theoretical text information to construct a key text segment information set. Feature vector analysis technologies in natural language processing technology, such as word embedding technology, are combined to extract feature vectors corresponding to different key text segments to construct a text feature vector set. Based on the geological structure feature vector set and the text feature vector set, a metallogenic gene feature information set is constructed, so that the metallogenic gene feature information set covers the comprehensive information features about geological gene analysis in the authoritative theoretical information, thereby improving the scientificity and reliability of the metallogenic gene feature information set and providing a solid information foundation for the subsequent feature analysis process.

[0111] Through this scheme, feature analysis and extraction are performed on the geological example image information and theoretical text information in the authoritative theoretical information, so as to construct a geological structure feature vector information set and a key text segment information set, so as to facilitate the computer-level automated analysis of the authoritative theoretical information, while enhancing data quality, reducing data redundancy, and improving the efficiency and accuracy of subsequent feature analysis.

[0112] In some embodiments, geological example image information is analyzed to extract the image mean and image standard deviation of each geological example image; based on the image mean and image standard deviation, feature vector dimensionality reduction extraction processing is performed on each geological example image to determine the effective image feature vector corresponding to each geological example image, specifically the following formula (1):

[0113]

[0114] Among them, E i is the effective image feature vector corresponding to the i-th geological example image, PCA() is the preset dimensionality reduction function, CNN() is the preset convolution function, I i is the i-th geological example image, mean(I i ) is the image mean of the i-th geological example image, std(I i ) is the image standard deviation of the i-th geological example image; according to the effective image feature vector corresponding to each geological example image, a geological structure feature vector set is constructed.

[0115] The image mean can be the average value of all pixel values ​​in an image and can be used to represent the overall brightness level of the image.

[0116] The image standard deviation can be the standard deviation of all pixel values ​​in the image, indicating the degree of dispersion of pixel values ​​in the image and reflecting the contrast of the image.

[0117] The effective image feature vector may be a feature vector reflecting the metallogenic gene in the geological example image.

[0118] The preset dimensionality reduction function may be a preset function for performing dimensionality reduction processing on a feature vector, and the preset dimensionality reduction function may adopt a PCA (Principal Component Analysis) algorithm.

[0119] The preset convolution function may be a preset function for extracting features from an image. The preset convolution function may adopt a mainstream architecture, such as a convolutional neural network of a VGG (Visual Geometry Group) or ResNet (Residual Neural Network) architecture.

[0120] Specifically, in the process of extracting feature vectors from different geological sample images, the formula (1) The geological sample images are standardized to eliminate the brightness and contrast differences between images and improve the stability of subsequent processing. Then, the standardized geological sample images are subjected to convolution feature processing and feature dimensionality reduction processing respectively through preset convolution function and preset dimensionality reduction function, and the effective image feature vector corresponding to each geological sample image is quantified to construct a geological structure feature vector set. While effectively capturing the characteristics of the geological sample images, the high-dimensional features are reduced to low dimensions and redundant features are removed.

[0121] Through this solution, mathematical analysis methods are used to perform standardized vector dimensionality reduction processing on the image information of different geological examples, and mathematical formulas are used to automatically and accurately extract the effective image feature vectors in the geological example images. While effectively capturing the features within the geological example images, high-dimensional features are reduced to low dimensions, redundant features are removed, the burden of subsequent computational analysis is reduced, and the generalization ability of the subsequent analysis process is improved.

[0122] In some embodiments, the key text segment information set is analyzed to extract the keyword set of each key text segment; based on the keyword set and the key text segment information set, an aggregate vector analysis is performed on each key text segment to extract the aggregate text vector corresponding to each key text segment, specifically the following formula (2):

[0123]

[0124] Among them, G m is the aggregate text vector corresponding to the mth key text segment, n is the total number of keywords in the keyword set corresponding to the current key text segment, V j is the jth keyword in the current keyword set, and Embed() is the preset word embedding function; a text feature vector set is constructed based on the aggregated text vector corresponding to each key text segment.

[0125] The keyword set may be a set of keywords in a key text segment that are highly relevant to the mineralization genes.

[0126] The aggregated text vector may be vector information used to describe overall features of a key text segment.

[0127] The preset word embedding function may be a preset function for converting keywords into feature vectors. The preset word embedding function may adopt a word embedding model, such as Word2Vec (Word to Vector) or BERT (Bidirectional Encoder Representations from Transformers).

[0128] Specifically, the keyword extraction algorithm in the natural language processing technology is used to extract keywords from each key text segment in the key text segment information set, and the order of keywords in the key text segment is used as the index to construct the keyword set of each key text segment. The keyword set of each key text segment is constructed by the Embed(V j ) extracts the high-dimensional feature vectors of each keyword, which reflect the semantic features of the keyword. Then, the high-dimensional feature vectors corresponding to each key text segment information are averaged by formula (2), and the aggregated text vector corresponding to each key text segment information is quantified to reflect the overall features corresponding to each key text segment. Based on the aggregated text vector corresponding to each key text segment, a text feature vector set is constructed.

[0129] Through this scheme, the keywords corresponding to each key text segment in the key text segment information are extracted to obtain the keyword set corresponding to each key text segment. On this basis, through mathematical analysis means, an aggregation vector analysis is performed on each key text segment to quantify the aggregation text vector corresponding to each key text segment. According to the aggregation text vector corresponding to each key text segment, a text feature vector set is constructed, so that the text feature vector set can comprehensively reflect the different semantic features in the theoretical text information, and provide a scientific data basis for the subsequent mineralization gene analysis process.

[0130] In some embodiments, the metallogenic gene feature information set is analyzed to determine the feature complexity of each metallogenic gene feature; the metallogenic gene feature information set is analyzed to determine the feature space center point vector; based on the feature complexity and the feature space center point vector, the metallogenic gene feature information set is subjected to binary screening to determine the pointing node set and the target node set, respectively, as specifically expressed in the following mathematical expression (3):

[0131]

[0132] Among them, S T is the target node set, C p is the pth metallogenic gene feature in the metallogenic gene feature information set, β k is the preset mapping coefficient corresponding to the kth feature power, d is the preset target bias coefficient, w p is the characteristic complexity of the pth metallogenic gene feature, θ T is the preset target evaluation threshold, γ is the preset adjustment parameter, μ is the feature space center point vector, b is the preset pointing bias coefficient, θ I The preset pointing evaluation threshold.

[0133] Feature complexity can be the complexity of the feature vector corresponding to each metallogenic gene feature in the metallogenic gene feature information set.

[0134] The feature space center point vector may be the center point vector information within the spatial coordinates of the feature vector in the metallogenic gene feature information set. The feature space center point vector may be obtained by performing cluster center calculation on all feature vectors in the metallogenic gene feature information set.

[0135] Binary screening can be a screening process to determine whether different metallogenic gene features in the metallogenic gene feature information set are target nodes or pointing nodes.

[0136] The preset mapping coefficients may be coefficients used to adjust the degree of influence of different features on the overall node expression.

[0137] The preset target bias coefficient may be a coefficient used to control the range of the target evaluation result.

[0138] The preset target evaluation threshold may be threshold information used to determine whether the current metallogenic gene feature is a target node, and the preset target evaluation threshold may be obtained by fitting experimental data.

[0139] The preset adjustment parameter may be a parameter value used to adjust the similarity measurement strength between the current feature vector and the feature space center point vector during the pointing evaluation process.

[0140] The preset pointing bias coefficient may be a coefficient used to control the range of the pointing evaluation result.

[0141] The preset pointing evaluation threshold may be threshold information used to determine whether the current metallogenic gene feature is a pointing node, and the preset pointing evaluation threshold may be obtained by fitting experimental data.

[0142] Specifically, after obtaining the metallogenic gene feature information set, it is necessary to determine whether the different metallogenic gene features in the metallogenic gene feature information set are target nodes or pointing nodes, so as to clarify the correlation between the metallogenic gene features and provide a data basis for the subsequent structured processing of the metallogenic gene information. The expression of metallogenic gene characteristics in different dimensions is weightedly calculated to comprehensively consider the importance of characteristics of different dimensions in metallogenic potential. The result is centralized to a reasonable range by presetting the target bias coefficient to ensure that the evaluation result does not only depend on the original value of the characteristic. Then, combined with the characteristic complexity, the target node evaluation result of the current metallogenic gene characteristic is obtained. By comparing the target node evaluation result with the preset target evaluation threshold, it is accurately judged whether the current metallogenic gene characteristic is the target node.

[0143] At the same time, through formula (3) The deviation distance between different metallogenic gene features and the spatial center point vector is quantified. When the deviation distance between the metallogenic gene feature and the spatial center point vector is larger, it means that the metallogenic gene feature is at the edge of space. At this time, the probability that the metallogenic gene feature is pointing to a node is greater, and vice versa. Combined with the regulation of the preset pointing bias coefficient and feature complexity, the pointing evaluation result of the metallogenic gene feature is accurately quantified, and based on the comparison result between the pointing evaluation result and the preset pointing evaluation threshold, it is accurately judged whether the current metallogenic gene feature is pointing to a node.

[0144] Through this scheme, mathematical analysis methods are used, based on the feature space center point vector and the feature complexity of each metallogenic gene feature, and through clear mathematical expressions, binary screening of the metallogenic gene features in the metallogenic gene feature information set is performed to determine the target nodes and pointing nodes in the metallogenic gene feature information set, improve the accuracy and scientificity of node judgment, and provide an important data foundation for the subsequent construction of a structured information database.

[0145] In some embodiments, the pointing node set and the target node set are analyzed to determine the pointing correlation between each pointing node and each target node; based on the pointing correlation, a feature-weighted directed graph is constructed according to the pointing node set and the target node set; and the feature-weighted directed graph is used as a mineralization gene structure information library.

[0146] The pointing association degree may be a mathematically quantified value used to describe the closeness of the relationship between the pointing node and the corresponding target node.

[0147] The feature-weighted directed graph can be a computer structured data constructed based on the data structure of the computer directed graph, according to the pointing node set and the target node set, combined with the pointing association between different nodes.

[0148] Specifically, after determining the pointing node set and the target node set respectively, it is necessary to analyze the specific correlation degree between the pointing node and the corresponding target node to clarify the influence relationship between the pointing node and the target node, so that in the subsequent analysis of the mineralization gene, the target minerals pointed to by different mineralization genes can be quickly determined according to the changes in the genes, thereby improving the accuracy and speed of the mineralization gene analysis. In order to clarify the complex network relationship between the nodes, a weighted directed graph is selected as the data structure of the mineralization gene structure information library, so as to facilitate the computer to realize the rapid and automated analysis and processing of mineralization genes. Based on the pointing correlation degree, according to the pointing node set and the target node set, different features in the pointing node set are used as the pointing nodes in the weighted directed graph, different features in the target node set are used as the target nodes in the weighted directed graph, and the pointing correlation degree is used as the weighted path between the pointing node and the corresponding target node to construct a feature weighted directed graph, and the feature weighted directed graph is used as the mineralization gene structure information library.

[0149] Through this scheme, on the basis of the pointing node set and the target node set, according to the pointing correlation between different pointing nodes and target nodes, a feature-weighted finite graph is constructed with the weighted directed graph as the basic data structure, and the feature-weighted directed graph is used as the metallogenic gene structure information library, so that the computer can quickly parse the node information in the metallogenic gene structure information library, providing structured data support for subsequent metallogenic analysis.

[0150] In some embodiments, the feature relationship adjacency matrix is ​​determined based on the directional association degree, specifically as the following expression (4):

[0151]

[0152] Among them, A xy is the feature relationship adjacency matrix, Con(x,y) is the pointing correlation between the xth pointing node and the yth target node, T x is the eigenvector corresponding to the x-th pointing node, R y is the feature vector corresponding to the y-th target node, and δ is the preset relationship threshold; the feature vectors corresponding to each node in the pointed node set and the target node set are used as directed graph nodes, and each directed correlation degree in the feature relationship adjacency matrix is ​​used as the directed graph path to construct a feature weighted directed graph.

[0153] The feature relationship adjacency matrix may be a data matrix containing association relationships between different pointing nodes and corresponding target nodes.

[0154] The preset relationship threshold may be a correlation threshold indicating a strong correlation between two nodes.

[0155] A directed graph node can be an independent entity or object in a directed graph.

[0156] A directed graph path can be a directed edge that reaches from one node to another in a directed graph.

[0157] Specifically, through the expression (4) The pointing correlation between the pointing node and the target node is quantified and normalized. If the pointing correlation is greater than the preset relationship threshold, it means that there is a strong correlation between the current pointing node and the current target node. At this time, it is considered that there is a valid path between the two nodes, and the path weight is the pointing correlation between the two nodes. If the pointing correlation is less than the preset relationship threshold, it means that there is no significant correlation between the current pointing node and the current target node. At this time, it is considered that there is no valid path between the two nodes, and then a feature adjacency matrix reflecting the relationship structure between all pointing nodes and target nodes is constructed, and then a feature weighted directed graph is constructed according to the corresponding relationship between different nodes in the feature adjacency matrix.

[0158] Through this scheme, mathematical analysis methods are used to quantify the pointing correlation between different nodes, and through mathematical expressions, a characteristic adjacency matrix is ​​constructed based on the pointing correlation to reflect the relationship structure between all pointing nodes and target nodes. Based on the characteristic adjacency matrix, a characteristic weighted directed graph is constructed according to the pointing node set and the target node set, so that the characteristic weighted directed graph can clearly reflect the degree of correlation between different nodes, providing a scientific data basis for subsequent mineralization analysis.

[0159] In some embodiments, the mining area exploration data set is analyzed to extract exploration gene feature information; the exploration gene feature information is matched with the feature weighted directed graph, and the associated node set and the associated weight set are determined based on the matching results; based on the associated node set, the mining area exploration data set is analyzed to extract the gene indicator set; based on the associated weight set and the gene indicator set, the associated node set is screened for effectiveness to determine the effective node information set.

[0160] The exploration gene characteristic information may be characteristic vector information corresponding to the metallogenic gene information in the mining area exploration data.

[0161] The associated node set may be a set of nodes in the feature-weighted directed graph corresponding to the metallogenic genes in the exploration gene feature information.

[0162] The association weight set may be a set of path weights between nodes in the association node set.

[0163] The gene indicator set may be a set of specific indicators of the mineralization genes corresponding to each node.

[0164] Specifically, different types of exploration data in the mining area exploration data set are analyzed through image feature vector conversion or text feature vector conversion technology to obtain feature vector information of corresponding mineralization genes, so as to obtain exploration gene feature information. Then, through a node matching algorithm, such as a cosine similarity algorithm, the feature vectors in the exploration gene feature information are matched with the feature vectors corresponding to the nodes in the feature weighted directed graph, and the corresponding nodes and the path weights between the nodes are extracted to construct an associated node set and an associated weight set. On the basis of the associated node set, the specific content of the mineralization genes corresponding to the associated nodes in the mining area exploration data set is queried to obtain a set of gene indicator quantities. The higher the path weight and specific content of the node corresponding to the mineralization gene, the greater the probability that the mineralization gene points to the target mineral, and this part of the nodes is determined as a valid node.

[0165] Through this solution, the information in the mining exploration data set is converted into corresponding feature vector information, and the exploration gene feature information is matched with the feature weighted directed graph to obtain a highly correlated association node information set and a corresponding association weight set. On this basis, the effectiveness of the associated node information set is screened through the gene indicator set, and the effective node information set is determined to achieve rapid and automated matching of mining exploration data, further enhancing data accuracy and reducing data redundancy.

[0166] In some embodiments, the mining area exploration data set is analyzed to determine the regional distribution information of the metallogenic genes; based on the valid node information set, the metallogenic gene structure information library is analyzed to determine the corresponding mining area target node set in the metallogenic gene structure information library; based on the regional distribution information of the metallogenic genes, the mining area target node set is analyzed to determine the metallogenic regional distribution information; based on the metallogenic regional distribution information, the metallogenic status analysis report is determined and output.

[0167] The regional distribution information of metallogenic genes may be the distribution of each metallogenic gene in the mining area exploration data set in different regions within the mining area.

[0168] The mining area target node set may be a set of target nodes corresponding to each target mineral in the mining area in the metallogenic gene structure information database.

[0169] The mineralization area distribution information may be the mineralization status of each target mineral in each area of ​​the mining area.

[0170] Specifically, since the mining area exploration dataset reflects the exploration situation of the entire mining area, if the mineralization status of the entire mining area is analyzed directly on the basis of the mining area exploration dataset, it is easy to cause the analysis results to lack details, and it is difficult to accurately analyze the changes in the mineralization status of different areas in the mining area, and thus it is impossible to accurately locate the mineralization-intensive areas in the mining area, resulting in a lack of data support for the formulation of subsequent development plans. Therefore, by analyzing the data in the mining area exploration dataset, the distribution of the corresponding mineralization genes in the exploration data in different areas of the mining area is determined, and at the same time, the mineralization gene structure information database is searched according to the valid node information set. , determine several target nodes corresponding to the valid nodes to form a target node set of the mining area. The target minerals corresponding to these target nodes are the minerals that need to be paid attention to in the current mineralization analysis process of the mining area. Through the regional distribution information of the mineralization genes that reflects the distribution of the mineralization genes in various regions within the mining area, the mineralization analysis of different target minerals corresponding to the target nodes is carried out. Specifically, the more types and the greater the content of the mineralization genes pointing to the target nodes in a region, the more considerable the reserves of the target minerals corresponding to the target nodes in the region. The mineralization status in different regions is integrated, and the mineralization status analysis report is constructed and output through data visualization technology.

[0171] Through this solution, based on the distribution of different mineralization genes in different areas of the mining area, the mineralization status of the target minerals corresponding to the target nodes in the mining area within the effective nodes are analyzed in different areas of the mining area to construct the mineralization area distribution information corresponding to the current mining area, and according to the distribution of different target minerals in different areas within the mineralization area distribution information, the mineralization status analysis report is determined and output, so that the mineralization status analysis report clearly reflects the changes in the distribution of different target minerals between different areas of the mining area, thereby enabling users to accurately locate the mineralization-intensive areas within the mining area and provide data support for the formulation of subsequent mineral development plans.

[0172] Figure 3 A schematic diagram of the structure of a mineralization gene analysis system based on an expert knowledge base provided in one embodiment of the present application is shown in FIG. Figure 3 As shown, a mineralization gene analysis method 300 based on an expert knowledge base in this embodiment includes: a feature analysis module 301, a node analysis module 302, a node construction module 303, a node screening module 304 and a state output module 305.

[0173] The feature analysis module 301 is used to obtain authoritative theoretical information, analyze the authoritative theoretical information, and determine the metallogenic gene feature information set;

[0174] The node analysis module 302 is used to perform a directional association analysis on the metallogenic genes according to the metallogenic gene feature information set, and determine a directional node set and a target node set;

[0175] A node construction module 303 is used to construct a metallogenic gene structure information library according to the pointing node set and the target node set;

[0176] The node screening module 304 is used to obtain a mining area exploration data set, analyze the mining area exploration data set based on the metallogenic gene structure information database, and determine a valid node information set;

[0177] The status output module 305 is used to analyze the valid node information set and the mining area exploration data set based on the metallogenic gene structure information library, and determine and output a metallogenic status analysis report.

[0178] Optionally, the feature analysis module 301 is specifically configured to:

[0179] According to the geological example image information, performing a standardized vector dimensionality reduction process on the geological example image information to determine a geological structure feature vector set;

[0180] Analyzing the theoretical text information, performing entity recognition screening on the theoretical text information, and extracting a key text segment information set;

[0181] According to the key text segment information set, performing text feature vector analysis on the key text segment information set to determine a text feature vector set;

[0182] The metallogenic gene feature information set is constructed based on the geological structure feature vector set and the text feature vector set.

[0183] Optionally, when the feature analysis module 301 performs a normalized vector dimensionality reduction process on the geological example image information to determine the geological structure feature vector set, the feature analysis module 301 is specifically configured to:

[0184] Analyzing the geological example image information and extracting the image mean and image standard deviation of each geological example image;

[0185] Based on the image mean and the image standard deviation, a feature vector dimensionality reduction extraction process is performed on each geological example image to determine the effective image feature vector corresponding to each geological example image, specifically the following formula:

[0186]

[0187] Among them, E i is the effective image feature vector corresponding to the i-th geological example image, PCA() is a preset dimensionality reduction function, CNN() is a preset convolution function, I i is the ith geological example image, mean(I i is the image mean of the i-th geological example image, std(Ii ) is the image standard deviation of the i-th geological example image;

[0188] The geological structure feature vector set is constructed according to the effective image feature vector corresponding to each geological example image.

[0189] Optionally, when the feature analysis module 301 performs text feature vector analysis on the key text segment information set to determine the text feature vector set, the feature analysis module 301 is specifically configured to:

[0190] Analyzing the key text segment information set to extract a keyword set for each key text segment;

[0191] Based on the keyword set and the key text segment information set, an aggregate vector analysis is performed on each key text segment to extract the aggregate text vector corresponding to each key text segment, specifically the following formula:

[0192]

[0193] Among them, G m is the aggregated text vector corresponding to the mth key text segment, n is the total number of keywords in the keyword set corresponding to the current key text segment, V j is the jth keyword in the current keyword set, and Embed() is a preset word embedding function;

[0194] The text feature vector set is constructed according to the aggregated text vector corresponding to each key text segment.

[0195] Optionally, the node analysis module 302 is specifically configured to:

[0196] Analyzing the metallogenic gene feature information set to determine the feature complexity of each metallogenic gene feature;

[0197] Analyzing the metallogenic gene feature information set to determine the feature space center point vector;

[0198] Based on the feature complexity and the feature space center point vector, the mineralization gene feature information set is subjected to binary screening to determine the pointing node set and the target node set, respectively. Specifically, the mathematical expression is as follows:

[0199]

[0200] Among them, S T is the target node set, C p is the pth metallogenic gene feature in the metallogenic gene feature information set, β kis the preset mapping coefficient corresponding to the kth feature power, d is the preset target bias coefficient, w p is the characteristic complexity of the pth metallogenic gene feature, θ T is the preset target evaluation threshold, γ is the preset adjustment parameter, μ is the center point vector of the feature space, b is the preset pointing bias coefficient, θ I The preset pointing evaluation threshold.

[0201] Optionally, the node construction module 303 is specifically configured to:

[0202] Analyzing the pointing node set and the target node set to determine the pointing association degree between each pointing node and each target node;

[0203] Based on the pointing association degree, a feature-weighted directed graph is constructed according to the pointing node set and the target node set;

[0204] The feature-weighted directed graph is used as the mineralization gene structure information database.

[0205] Optionally, when constructing the feature-weighted directed graph based on the pointing association degree and according to the pointing node set and the target node set, the node construction module 303 is specifically configured to:

[0206] According to the directional correlation degree, the feature relationship adjacency matrix is ​​determined, which is specifically expressed as follows:

[0207]

[0208] Among them, A xy is the feature relationship adjacency matrix, Con(x,y) is the directional correlation between the xth pointing node and the yth target node, T x is the eigenvector corresponding to the x-th pointing node, R y is the feature vector corresponding to the y-th target node, and δ is the preset relationship threshold;

[0209] The feature weighted directed graph is constructed by taking the feature vectors corresponding to each node in the pointing node set and the target node set as directed graph nodes and taking each pointing association degree in the feature relationship adjacency matrix as a directed graph path.

[0210] Optionally, the node screening module 304 is specifically configured to:

[0211] Analyzing the mining area exploration data set to extract exploration gene feature information;

[0212] Matching the exploration gene feature information with the feature weighted directed graph, and determining an associated node set and an associated weight set based on the matching result;

[0213] Analyzing the mining area exploration data set based on the associated node set to extract a set of gene indicators;

[0214] According to the association weight set and the gene indicator set, the association node set is screened for effectiveness to determine the effective node information set.

[0215] Optionally, the status output module 305 is specifically configured to:

[0216] Analyze the mining area exploration data set to determine the regional distribution information of metallogenic genes;

[0217] Analyzing the metallogenic gene structure information database according to the valid node information set to determine the corresponding mining area target node set in the metallogenic gene structure information database;

[0218] Based on the regional distribution information of the metallogenic genes, analyzing the target node set of the mining area to determine the distribution information of the metallogenic area;

[0219] According to the mineralization area distribution information, the mineralization status analysis report is determined and output.

[0220] The system of this embodiment can be used to execute the method of any of the above embodiments. Its implementation principles and technical effects are similar and will not be described in detail here.

Claims

1. A mineralization gene analysis method based on an expert knowledge base, characterized in that: include: Obtaining authoritative theoretical information, the authoritative theoretical information including geological example image information and theoretical text information, analyzing the authoritative theoretical information, and determining a metallogenic gene feature information set; According to the metallogenic gene characteristic information set, a directional association analysis is performed on the metallogenic genes to determine a directional node set and a target node set; The pointing node set is the node set corresponding to the mineralization gene pointing to the target mineral; The target node set is the node set corresponding to the target mineral; Constructing a mineralization gene structure information library according to the pointing node set and the target node set; Acquire a mining area exploration data set, analyze the mining area exploration data set based on the metallogenic gene structure information database, and determine a valid node information set; The valid node information set is the set of nodes with analytical value in the current user's target analysis area; Based on the metallogenic gene structure information database, analyzing the valid node information set and the mining area exploration data set, determining and outputting a metallogenic status analysis report; The method of performing a directional association analysis on the metallogenic genes according to the metallogenic gene feature information set to determine a directional node set and a target node set includes: Analyzing the metallogenic gene feature information set to determine the feature complexity of each metallogenic gene feature; Analyzing the metallogenic gene feature information set to determine the feature space center point vector; Based on the feature complexity and the feature space center point vector, the mineralization gene feature information set is subjected to binary screening to determine the pointing node set and the target node set, respectively. Specifically, the mathematical expression is as follows: Among them, S T is the target node set, S I is the set of pointing nodes, C p is the pth metallogenic gene feature in the metallogenic gene feature information set, β k is the preset mapping coefficient corresponding to the kth feature power, d is the preset target bias coefficient, w p is the characteristic complexity of the pth metallogenic gene feature, θ T is the preset target evaluation threshold, γ is the preset adjustment parameter, μ is the center point vector of the feature space, b is the preset pointing bias coefficient, θ I The preset pointing evaluation threshold.

2. The method according to claim 1, characterized in that The analyzing the authoritative theoretical information to determine the metallogenic gene characteristic information set includes: According to the geological example image information, performing a standardized vector dimensionality reduction process on the geological example image information to determine a geological structure feature vector set; Analyzing the theoretical text information, performing entity recognition screening on the theoretical text information, and extracting a key text segment information set; According to the key text segment information set, performing text feature vector analysis on the key text segment information set to determine a text feature vector set; The metallogenic gene feature information set is constructed based on the geological structure feature vector set and the text feature vector set.

3. The method according to claim 2, characterized in that The step of performing standardized vector dimensionality reduction processing on the geological example image information to determine a geological structure feature vector set includes: Analyzing the geological example image information and extracting the image mean and image standard deviation of each geological example image; Based on the image mean and the image standard deviation, a feature vector dimensionality reduction extraction process is performed on each geological example image to determine the effective image feature vector corresponding to each geological example image, specifically the following formula: Among them, E i is the effective image feature vector corresponding to the i-th geological example image, PCA() is a preset dimensionality reduction function, CNN() is a preset convolution function, I i is the ith geological example image, mean(I i ) is the image mean of the i-th geological example image, std(I i ) is the image standard deviation of the i-th geological example image; The geological structure feature vector set is constructed according to the effective image feature vector corresponding to each geological example image.

4. The method according to claim 2, characterized in that The performing text feature vector analysis on the key text segment information set according to the key text segment information set to determine the text feature vector set includes: Analyzing the key text segment information set to extract a keyword set for each key text segment; Based on the keyword set and the key text segment information set, an aggregate vector analysis is performed on each key text segment to extract the aggregate text vector corresponding to each key text segment, specifically the following formula: Among them, G m is the aggregated text vector corresponding to the mth key text segment, n is the total number of keywords in the keyword set corresponding to the current key text segment, V j is the jth keyword in the current keyword set, and Embed() is a preset word embedding function; The text feature vector set is constructed according to the aggregated text vector corresponding to each key text segment.

5. The method according to claim 1, wherein The step of constructing a metallogenic gene structure information database based on the pointing node set and the target node set includes: Analyzing the pointing node set and the target node set to determine the pointing correlation between each pointing node and each target node; Based on the directional association degree, constructing a feature-weighted directed graph according to the directional node set and the target node set; The feature-weighted directed graph is used as the mineralization gene structure information database.

6. The method according to claim 5, characterized in that The step of constructing a feature-weighted directed graph based on the pointing association degree and according to the pointing node set and the target node set includes: According to the directional correlation degree, the feature relationship adjacency matrix is ​​determined, which is specifically expressed as follows: Among them, A xy is the feature relationship adjacency matrix, Con(x,y) is the directional correlation between the xth pointing node and the yth target node, T x is the eigenvector corresponding to the x-th pointing node, R y is the feature vector corresponding to the y-th target node, and δ is the preset relationship threshold; The feature weighted directed graph is constructed by taking the feature vectors corresponding to each node in the pointing node set and the target node set as directed graph nodes and taking each pointing association degree in the feature relationship adjacency matrix as a directed graph path.

7. The method according to claim 5, characterized in that The step of analyzing the mining area exploration data set based on the metallogenic gene structure information database to determine the valid node information set includes: Analyzing the mining area exploration data set to extract exploration gene feature information; Matching the exploration gene feature information with the feature weighted directed graph, and determining an associated node set and an associated weight set based on the matching result; Analyzing the mining area exploration data set based on the associated node set to extract a set of gene indicators; According to the association weight set and the gene indicator set, the association node set is screened for effectiveness to determine the effective node information set.

8. The method according to claim 6, characterized in that The method of analyzing the effective node information set and the mining area exploration data set based on the metallogenic gene structure information database to determine and output a metallogenic status analysis report includes: Analyze the mining area exploration data set to determine the regional distribution information of metallogenic genes; Analyzing the metallogenic gene structure information database according to the valid node information set to determine the corresponding mining area target node set in the metallogenic gene structure information database; Based on the regional distribution information of the metallogenic genes, analyzing the target node set of the mining area to determine the distribution information of the metallogenic area; According to the mineralization area distribution information, the mineralization status analysis report is determined and output.

9. A mineralization gene analysis system based on expert knowledge base, characterized in that: The method according to any one of claims 1 to 7 comprises: A feature analysis module is used to obtain authoritative theoretical information, analyze the authoritative theoretical information, and determine the metallogenic gene feature information set; A node analysis module is used to perform a directional association analysis on the metallogenic genes according to the metallogenic gene feature information set, and determine a directional node set and a target node set; A node construction module, configured to construct a metallogenic gene structure information library according to the pointing node set and the target node set; A node screening module is used to obtain a mining area exploration data set, analyze the mining area exploration data set based on the metallogenic gene structure information database, and determine a valid node information set; The status output module is used to analyze the valid node information set and the mining area exploration data set based on the metallogenic gene structure information library, and determine and output a metallogenic status analysis report.

Images