Geological literature key information extraction method based on large model

Abstract

The invention discloses a geological literature key information extraction method based on a large model, and the method comprises the steps: collecting geological literature data of a preset region, and carrying out the preprocessing of the geological literature data; performing joint information extraction on the geological multi-source data through a geological knowledge graph to obtain literature information data, and performing corresponding hyponymy on the literature information data by adopting a multi-granularity semantic alignment mechanism to obtain a semantic alignment relationship; setting an initial weight of the literature information data according to a geological rule base and the semantic alignment relationship, dynamically adjusting the initial weight based on decision routing of reinforcement learning to obtain an optimized weight, and performing heterogeneous knowledge graph fusion according to the literature information data and the optimized weight to obtain fused data; and constructing a geological literature key information extraction model according to the fused data, performing man-machine collaborative optimization on the geological literature key information extraction model according to a result deviation, inputting to-be-extracted data into the geological literature key information extraction model, and outputting an extraction result.

Description

Technical Field

[0001] This invention relates to the field of geological information extraction, and more particularly to a method for extracting key information from geological documents based on large models. Background Technology

[0002] In the field of geological research, geological literature contains various types of data, including text, maps, and tables, carrying core information such as regional geological characteristics, entity attributes, and relational relationships. It serves as crucial data support for geological exploration and resource assessment. However, the sheer volume and diverse formats of current geological literature present significant limitations to traditional information extraction methods: insufficient adaptability to multimodal data, making it difficult to simultaneously process technical terms in text, spatial coordinates in maps, and structured data in tables; a lack of domain-specific semantic understanding, easily confusing obscure geological terms with entities of the same name, and failing to effectively establish hierarchical relationships between entities of different granularities; and rigid data weighting, making dynamic adjustments based on data source credibility and information importance difficult, resulting in insufficient accuracy and completeness of extraction results. Furthermore, manual extraction of key information is time-consuming, labor-intensive, and inefficient, failing to meet the needs of geological research for rapid data acquisition and integration.

[0003] Therefore, there is an urgent need for a method that integrates domain knowledge and advanced technologies to solve problems such as collaborative extraction, semantic alignment, and dynamic weight optimization of multimodal geological data, so as to achieve efficient and accurate extraction of key information from geological literature and provide strong data support for geological research. Summary of the Invention

[0004] The purpose of this invention is to provide a method for extracting key information from geological literature based on large models.

[0005] To achieve the above objectives, the present invention is implemented according to the following technical solution: This invention includes the following steps: Collect geological literature data from a preset area and preprocess the geological literature data; the geological literature data includes text data, map data, and tabular data. Literature information data is obtained by jointly extracting information from the geological multi-source data using a geological knowledge graph, and semantic alignment relationship is obtained by performing corresponding hierarchical relationship on the literature information data using a multi-granularity semantic alignment mechanism. The initial weights of the literature information data are set according to the geological rule base and the semantic alignment relationship. The initial weights are dynamically adjusted based on the decision routing of reinforcement learning to obtain optimized weights. The heterogeneous knowledge graph is fused according to the literature information data and the optimized weights to obtain fused data. A key information extraction model for geological documents is constructed based on the fused data. The key information extraction model for geological documents is then optimized by human-computer collaboration based on the result deviation. The data to be extracted is input into the key information extraction model for geological documents, and the extraction results are output.

[0006] Furthermore, a method for obtaining literature information data by jointly extracting information from the aforementioned multi-source geological data using a geological knowledge graph includes: For the text data: pre-trained word vectors based on geological word vectors, and the word vector space that is easily confused is corrected by a geological terminology dictionary; character-level BPE segmentation is used for obscure terms, and after the input text is encoded by BERT, the contextual dependencies are captured by a bidirectional long short-term memory network, and the entity label sequence is obtained by combining the conditional random layer with the geological entity transition probability matrix; the entity disambiguation algorithm is used to match geological knowledge graph nodes, and for entities with the same name, subject labels are assigned based on contextual similarity to obtain the text entity set; For map data: DBNet algorithm is used to detect text regions in the map, CRNN model is used to identify text content, spatial coordinates of text detection boxes are extracted, and combined with map scale to convert to actual geographic coordinates or relative positions, establish spatial association between text and entities, and output map entity set; For tabular data: Based on text density and table line features, the table is divided into high-density data area and low-density annotation area. For borehole tables, the three-dimensional data of depth, lithology and fossils are automatically extracted and arranged. The numerical data is standardized by unit, the qualitative description is mapped to quantitative attribute values, matched with the knowledge graph attribute dictionary, and the output is a table entity set. The method combines rule base and statistical learning to extract entity pair relationships, calls the geological knowledge graph reasoning engine to verify the relationships, automatically marks contradictory relationships as pending confirmation, and outputs non-contradictory relationships as a set of verified relationships; rule-driven: when an entity is followed by ± value, an entity plus value attribute relationship is automatically established; statistical-driven: relationships are identified by combining co-occurrence frequency with BERT relationship classifier. Output the text entity set, figure entity set, table entity set, and verification relation set as document information data.

[0007] Furthermore, a method for obtaining semantic alignment relationships by performing corresponding hierarchical relationships on the document information data using a multi-granularity semantic alignment mechanism includes: The entities in the literature information data are labeled with granularity, and the rationality of the granularity is verified by the geological ontology database. Entities with mixed granularity are split or merged, and the entities in the literature information data are layered according to a 5-level granularity system, where the 5 levels are domain level, discipline level, type level, instance level, and attribute level. Semantic association of entities at different granularities is performed based on dual verification of word vector similarity and domain rules; Word vector similarity verification: The cosine similarity between entities is calculated using optimized geological word vectors. The similarity thresholds between domain-level and subject-level entities, and between subject-level and type-level entities are greater than or equal to 0.85. The similarity thresholds between type-level and instance-level entities, and between instance-level and attribute-level entities are greater than or equal to 0.78. Entities with hierarchical relationships but below the similarity threshold are subject to rule verification. Domain rule validation: If the definition of entity A contains the core features of entity B, then establish a hierarchical relationship between A and B; if entity B inherits all the core attributes of entity A, then confirm the hierarchical relationship; entities with semantic conflicts are marked as parallel relationships. By constructing complete hierarchical relationship chains for entities of different granularities through hierarchical transmission and conflict resolution, the structured semantically aligned relationship table is output. Hierarchical transmission automatically completes the relationship chain based on transitivity rules. Conflict resolution: when an entity has multiple potential superior concepts, subject tags are selected first, and intermediate nodes are automatically inserted for relationships that skip intermediate levels.

[0008] Furthermore, the method for setting the initial weights of the document information data based on the geological rule base and the semantic alignment relationship includes: The rule base includes entity type rules, data source rules, and attribute importance rules, and the weight benchmark values ​​are determined through expert consensus. The baseline weights are corrected by multi-granular semantic alignment. For hierarchical relationships, the subordinate entity inherits the weight of the superior entity and increases the weight. For synonym relationships, the weight of authoritative domain terms is given priority. Based on the rule base, entity type matching and data source matching are performed on entities in the literature information data to obtain entity type value and data source weight. Semantic alignment correction value is obtained by correcting the semantic alignment baseline value. The entity type value and data source weight are weighted and summed, and the semantic alignment correction value is added to obtain the entity weight. Attribute type matching is performed on the attributes associated with entities, and attribute weights are obtained by correcting the reliability of attribute values ​​based on quantitative attributes, qualitative data, precise numerical values, and fuzzy descriptions. The system performs relationship type matching between entities, and obtains relationship weights by adjusting the evidence strength based on multi-source evidence, single-source evidence, knowledge graph verification passed, and relationships to be verified; the entity weights, attribute weights, and relationship weights are output as the initial weights.

[0009] Furthermore, the method for dynamically adjusting the initial weights to obtain optimized weights based on reinforcement learning-based decision routing includes: The weight adjustment task is modeled as a Markov decision process, and dynamic decision-making is carried out through state, action, and reward loops. The geological entity type and literature topic to be extracted are the task context, the identified entity set and attributes are the extracted information, and the credibility score of each data source is the multi-source data reliability label. The adjustment of the weights of multi-source data includes three basic actions and magnitude control. The three basic actions are to enhance text weight, increase map weight, and integrate multi-source weights. The weight adjustment range for enhancing text weight is +0.1 to +0.3, the weight adjustment range for increasing map weight is +0.1 to +0.3, and the weight adjustment range for integrating multi-source weights is ±0.05 for each source weight. The reward value is calculated based on the consistency between the extraction results and the geological knowledge map, data reliability, and task objectives. The expression is: ; in The degree of matching between entity attributes and geological knowledge graphs. To ensure the weight allocation aligns with the credibility of the data source, To ensure the completeness of entity attribute extraction under the current weight, For matching degree weight, As a weight for fit, Integrity weight; An evaluation network and a target network strategy are introduced, and the deep Q-network in reinforcement learning is optimized using the evaluation network and the target network. The deep Q-network structure consists of a 3-layer fully connected network and an activation function. The evaluation network outputs the Q-value of each action in real time. The input is the current state S, and the output is the Q-value of each action in the action space. The target network freezes and copies the parameters from the evaluation network periodically to calculate the target Q-value. The 3-layer fully connected network consists of an input layer, a hidden layer, and an output layer. Using initial weights based on a geological rule base as initial actions, the network is evaluated by inputting the current state S and outputting the Q-value of each action. The greedy strategy selects actions, adjusts weights, extracts entity attributes, calculates reward values, observes new states, and updates the Q-value. The expression is: ; in This is the current state. For the new state, As a discount factor, For the current action, For new actions, The evaluation network value under the new state and new action. The target network value under the current state and current action. To maximize the value of all possible new actions in the action space; Every 100 training steps, the parameters of the evaluation network are copied to the target network. The temporal difference error is calculated based on the absolute value of the difference between the evaluation network and the target network. Sample priorities are assigned according to the importance of the samples, and the sample sampling probability is calculated. Importance sampling weights are introduced to correct the loss function. This process is iterated until the policy network converges, and the optimized weight set for the current task is output as the optimized weights.

[0010] Furthermore, the method for obtaining fused data by fusing heterogeneous knowledge graphs based on the literature information data and the optimized weights includes: The system extracts text subgraphs, image subgraphs, structured subgraphs, and external data subgraphs from entities and relationships within the literature information data. It then constructs a multimodal subgraph based on these subgraphs and assigns fusion priorities to each subgraph based on optimized weights. / / Text subgraphs contain literature text entities and attribute relationships; weights are initialized according to entity extraction confidence. Image subgraphs contain spatial entities and coordinate attributes; weights are dynamically adjusted based on the recognition accuracy of deep bounding box networks and convolutional recurrent neural networks. Structured subgraphs; External data subgraphs. The BERT-base-cased model is used to encode the entity attribute text. A 128-dimensional feature vector is obtained through max pooling. The embedding space is optimized by fusing it with geological word vectors. The spatial topological relationship of map entities is learned by using hyperbolic graph convolutional network. The entity features are trained by adaptive gradient algorithm, and the depth, lithology and fossil triples are transformed into structured embedding vectors. By comparing the similarity of the embedding vectors of entities in different subgraphs, the geological knowledge graph reasoning engine is called to verify the rationality of the alignment, and subject labels are assigned to entities with the same name by combining optimized weights and context keywords. For multiple sources of evidence for the same relationship, the confidence level is calculated by weighting the evidence with optimized weights. A relationship confidence threshold of 0.6 is set. Relationships with a confidence level below the threshold are marked as requiring manual verification. Using verified relationships in the knowledge graph to infer missing relationships, and combining geophysical data subgraphs to complete relationships between entities across subgraphs; Output the entity ID, name, type, multi-source attributes, weight value, relationship ID, head / tail entity ID, relationship type, confidence level, and weight source as fused data.

[0011] Furthermore, the method for constructing a key information extraction model for geological documents based on the fused data includes: The TransE algorithm is used to map knowledge graph entities to vector space, so that the vector distance between semantically similar entities is less than the similarity threshold, and the distance between dissimilar entities is greater than the dissimilarity threshold. For multi-source entities, feature vectors are fused by weighted averaging, with weights dynamically allocated proportionally to vectorized entity features, and geological relationship path features are introduced to enhance relationship features. A framework for a key information extraction model from geological documents is constructed based on a multimodal feature encoder, a dynamic weight fusion layer, and a relational reasoning and information extraction layer. Multimodal feature encoder: Employing a geologically fine-tuned BERT input text fragment to output contextual feature vectors, text feature editing is obtained; based on the spatial coordinates of map entities and text content, visual features are extracted via a convolutional neural network, concatenated with text features, and weighted through an attention mechanism to obtain map feature encoding; for structured tabular data, Transformer-XL is used to capture long sequence dependencies, outputting temporal feature vectors to obtain table feature encoding; weights for text feature editing, map feature encoding, and table feature encoding are dynamically allocated through a multimodal attention layer. Dynamic Weight Fusion Layer: Based on reinforcement learning decision routing, dynamic weight adjustment of multi-source features is performed. The current task to be extracted, the fused feature vector, and the reliability label of multi-source data are defined as the state. The action space is defined by adjusting the weights of each modality feature and the given weight value range. The weights are adjusted by a given reward function, which includes positive reward and negative penalty. The positive reward is the matching of the extraction result with the geological knowledge graph, and the negative penalty is the entity type error. Relational Reasoning and Information Extraction Layer: BiLSTM-CRF is used to identify entity boundaries, and a geological entity dictionary is combined for entity boundary detection; attribute values ​​are bound to entities through an entity-attribute attention matrix to obtain attribute value associations; entity pair relation probabilities are calculated based on a knowledge graph embedding model, and relation reasoning is performed based on entity pair relation probabilities; the entity set, attribute set, and relation are output as extraction results; Training data is constructed, and multi-task joint training is performed using a main task and an auxiliary task. The main task includes entity recognition and relation extraction, while the auxiliary task is multi-source data weight prediction. Geological constraints are added during the training process to correct prediction results that violate the rules through penalty terms.

[0012] Furthermore, a method for human-machine collaborative optimization of the geological literature key information extraction model based on result deviations includes: Set confidence thresholds for entity recognition and relation extraction, and mark low-confidence samples as samples to be verified; call the geological knowledge graph inference engine to identify conflicting data that contradicts the domain rules, and the results include entity type conflicts, relational logic contradictions, and attribute value anomalies; Compare the extraction results of text, images, and tables: if the same entity has conflicting attributes in different data sources, mark it as a cross-modal conflict sample; classify and statistically analyze high-frequency error types by entity, relationship, and attribute, and push them to obtain an error sample pool; The system performs entity-level, relation-level, and attribute-level corrections on the erroneous sample pool. When experts correct entities, it automatically recommends standard attributes of similar entities in the knowledge graph and displays historical amendment examples of this type of error to assist experts in making quick decisions and obtaining corrected samples. Entity-level correction supports entity type adjustment and boundary calibration; relation-level correction provides a drop-down selection of relation types and supports relation direction adjustment; attribute-level correction supports unit calibration and error range supplementation for numerical attributes, and allows free editing of text attributes. The format of the corrected samples is standardized, and model parameters are extracted based on the corrected samples and key information from geological literature for domain adaptation fine-tuning and small sample learning optimization.

[0013] The beneficial effects of this invention are: This invention is a method for extracting key information from geological literature based on large-scale models. Compared with existing technologies, this invention has the following technical advantages: This invention addresses the challenge of heterogeneous data extraction by employing preprocessing, joint information extraction, corresponding hierarchical relationships, setting initial weights, dynamically adjusting initial weights, heterogeneous knowledge graph fusion, model construction, and model optimization steps. It adapts to multimodal data, simultaneously processing text, image, and tabular data. Furthermore, it improves the accuracy of entity recognition and relationship extraction through multi-granular semantic alignment and dynamic weight optimization. A human-machine collaborative optimization mechanism continuously corrects model biases, enhancing robustness. Automated extraction replaces manual work, significantly improving efficiency and adapting to various geological scenarios, demonstrating strong universality. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the steps of the method for extracting key information from geological documents based on a large model, as described in this invention. Detailed Implementation

[0015] The present invention will be further described below through specific embodiments. The illustrative embodiments and descriptions herein are used to explain the present invention, but are not intended to limit the present invention.

[0016] The method for extracting key information from geological literature based on a large model, as described in this invention, includes the following steps: like Figure 1 As shown, this embodiment includes the following steps: Collect geological literature data from a preset area and preprocess the geological literature data; the geological literature data includes text data, map data, and tabular data. In the actual assessment, the geological literature on Mesozoic magmatic rocks in a certain region was used as the extraction object, covering 300 text documents, 80 geological maps and 50 borehole table data; The data collected included geological survey reports, journal articles, and other literature data for the region. The text data included descriptions of igneous rock types and isotopic ages, the map data included geological profiles and lithological distribution maps, and the tabular data mainly consisted of borehole core analysis tables. During the preprocessing stage, 23 duplicate documents were removed, 17 blurry areas in the maps were corrected, and the units of numerical values ​​in the tables were standardized. Literature information data is obtained by jointly extracting information from the geological multi-source data using a geological knowledge graph, and semantic alignment relationship is obtained by performing corresponding hierarchical relationship on the literature information data using a multi-granularity semantic alignment mechanism. The initial weights of the literature information data are set according to the geological rule base and the semantic alignment relationship. The initial weights are dynamically adjusted based on the decision routing of reinforcement learning to obtain optimized weights. The heterogeneous knowledge graph is fused according to the literature information data and the optimized weights to obtain fused data. A key information extraction model for geological documents is constructed based on the fused data. The key information extraction model for geological documents is then optimized by human-computer collaboration based on the result deviation. The data to be extracted is input into the key information extraction model for geological documents, and the extraction results are output.

[0017] In this embodiment, the method for obtaining literature information data by jointly extracting information from the multi-source geological data using a geological knowledge graph includes: For the text data: pre-trained word vectors based on geological word vectors, and the word vector space that is easily confused is corrected by a geological terminology dictionary; character-level BPE segmentation is used for obscure terms, and after the input text is encoded by BERT, the contextual dependencies are captured by a bidirectional long short-term memory network, and the entity label sequence is obtained by combining the conditional random layer with the geological entity transition probability matrix; the entity disambiguation algorithm is used to match geological knowledge graph nodes, and for entities with the same name, subject labels are assigned based on contextual similarity to obtain the text entity set; For map data: DBNet algorithm is used to detect text regions in the map, CRNN model is used to identify text content, spatial coordinates of text detection boxes are extracted, and combined with map scale to convert to actual geographic coordinates or relative positions, establish spatial association between text and entities, and output map entity set; For tabular data: Based on text density and table line features, the table is divided into high-density data area and low-density annotation area. For borehole tables, the three-dimensional data of depth, lithology and fossils are automatically extracted and arranged. The numerical data is standardized by unit, the qualitative description is mapped to quantitative attribute values, matched with the knowledge graph attribute dictionary, and the output is a table entity set. The method combines rule base and statistical learning to extract entity pair relationships, calls the geological knowledge graph reasoning engine to verify the relationships, automatically marks contradictory relationships as pending confirmation, and outputs non-contradictory relationships as a set of verified relationships; rule-driven: when an entity is followed by ± value, an entity plus value attribute relationship is automatically established; statistical-driven: relationships are identified by combining co-occurrence frequency with BERT relationship classifier. Output the text entity set, figure entity set, table entity set, and verification relation set as document information data; In the actual assessment, the text data processing involved: pre-training word vectors based on geological word vectors, correcting the easily confused word vector space of granite and granodiorite using the "Geological Terminology" dictionary; character-level BPE segmentation was used for rare zircon U-Pb dating terms, and after BERT encoding, entity label sequences were obtained by combining BiLSTM with the geological entity transition probability matrix; an entity disambiguation algorithm was used to match the geological knowledge graph, and for quartz entities with the same name, subject labels of igneous rock minerals and sedimentary rock minerals were assigned according to the context, ultimately obtaining more than 1,200 text entities; Map data processing: The DBNet algorithm is used to detect text regions in the map, and the CRNN model is used to identify the text content of granite bodies and fault zones. The spatial coordinates of the text detection boxes are extracted, and combined with the 1:50000 scale of the map, they are converted into actual geographic coordinates, spatial associations are established, and more than 850 map entities are output. Table data processing: Based on text density and table line features, the table region is split, and the three-dimensional data of depth, lithology and fossils are automatically extracted from the borehole table. The qualitative description of dense and hard is mapped to the quantitative attribute value of 0.85, matched with the knowledge graph attribute dictionary, and more than 600 table entities are output. The system identifies the relationship between the ±20 Ma entity and numerical attributes of granite through rule-driven identification, and identifies intrusive contact and phase transition relationships through statistical-driven identification combined with the BERT relationship classifier. The system calls the geological knowledge graph inference engine for verification, marks 32 contradictory relationships as pending confirmation, and outputs a set of more than 1,500 verification relationships.

[0018] In this embodiment, the method for obtaining semantic alignment relationships by performing corresponding hierarchical relationships on the document information data using a multi-granularity semantic alignment mechanism includes: The entities in the literature information data are labeled with granularity, and the rationality of the granularity is verified by the geological ontology database. Entities with mixed granularity are split or merged, and the entities in the literature information data are layered according to a 5-level granularity system, where the 5 levels are domain level, discipline level, type level, instance level, and attribute level. Semantic association of entities at different granularities is performed based on dual verification of word vector similarity and domain rules; Word vector similarity verification: The cosine similarity between entities is calculated using optimized geological word vectors. The similarity thresholds between domain-level and subject-level entities, and between subject-level and type-level entities are greater than or equal to 0.85. The similarity thresholds between type-level and instance-level entities, and between instance-level and attribute-level entities are greater than or equal to 0.78. Entities with hierarchical relationships but below the similarity threshold are subject to rule verification. Domain rule validation: If the definition of entity A contains the core features of entity B, then establish a hierarchical relationship between A and B; if entity B inherits all the core attributes of entity A, then confirm the hierarchical relationship; entities with semantic conflicts are marked as parallel relationships. Complete hierarchical relationship chains of entities at different granularities are constructed through hierarchical transmission and conflict resolution, and the output is a structured semantic alignment relationship table; hierarchical transmission automatically completes the relationship chain based on transitivity rules; conflict resolution: when an entity has multiple potential superior concepts, subject tags are selected first, and intermediate nodes are automatically inserted for relationships that skip intermediate levels. In the actual evaluation, each level corresponds to the hierarchical structure of the geological knowledge graph, and entities are layered according to a five-level granularity system: domain level, discipline level, type level, instance level, and attribute level. For example, the geological domain (domain level) - igneous petrology (discipline level) - intrusive rocks (type level) - granite bodies (instance level) - age (attribute level). Cosine similarity is calculated using optimized geological word vectors. The similarity between discipline level and type level entities is ≥0.85, and the similarity between type level and instance level entities is ≥0.78. For entities of basalt and basic volcanic rocks that are below the threshold but have a hierarchical relationship, alignment relationships are established through domain rule verification, and finally a structured semantic alignment relationship table is formed.

[0019] In this embodiment, the method for setting the initial weight of the document information data based on the geological rule base and the semantic alignment relationship includes: The rule base includes entity type rules, data source rules, and attribute importance rules, and the weight benchmark values ​​are determined through expert consensus. The baseline weights are corrected by multi-granular semantic alignment. For hierarchical relationships, the subordinate entity inherits the weight of the superior entity and increases the weight. For synonym relationships, the weight of authoritative domain terms is given priority. Based on the rule base, entity type matching and data source matching are performed on entities in the literature information data to obtain entity type value and data source weight. Semantic alignment correction value is obtained by correcting the semantic alignment baseline value. The entity type value and data source weight are weighted and summed, and the semantic alignment correction value is added to obtain the entity weight. Attribute type matching is performed on the attributes associated with entities, and attribute weights are obtained by correcting the reliability of attribute values ​​based on quantitative attributes, qualitative data, precise numerical values, and fuzzy descriptions. The relationship between entities is matched by relationship type, and the relationship weight is obtained by correcting the evidence strength based on multi-source evidence, single-source evidence, knowledge graph verification passed, and relationship to be verified. Output the entity weight, attribute weight, and relationship weight as the initial weights; In actual assessments, the following rules apply: Entity type: igneous rocks > sedimentary rocks > metamorphic rocks; tectonic entities > mineral entities, with a weight range of 0.6-0.9; Data source: laboratory analysis data > field measurement data > literature citation data, with a weight range of 0.5-0.95; Attribute importance: spatiotemporal attributes > physical attributes > descriptive attributes, with a weight range of 0.4-0.85. Entity types include igneous rocks, sedimentary rocks, and metamorphic rocks; attribute types include spatiotemporal attributes, physical attributes, and descriptive attributes, with physical attributes including tectonic attributes and structural attributes; relationship types include intrusive contact and phase transition. The baseline weight for igneous rocks is 0.85, the baseline weight for isotopic age is 0.85, the baseline weight for tectonic attributes is 0.7, the baseline weight for structural attributes is 0.5, the baseline weight for intrusive contact relationships in the rule base is 0.8, and the baseline weight for phase transition relationships is 0.75. Based on the rule base, the baseline weight of igneous rock in the entity type rule is 0.85, the weight of laboratory analysis data in the data source rule is 0.95, and the weight of isotopic age in the attribute importance rule is 0.85. Combining the semantic alignment relationship, the lower entity granite inherits the weight of the upper entity intrusive rock and increases it by 0.05 to obtain the initial weight set.

[0020] In this embodiment, the method for dynamically adjusting the initial weights to obtain optimized weights based on reinforcement learning-based decision routing includes: The weight adjustment task is modeled as a Markov decision process, and dynamic decision-making is carried out through state, action, and reward loops. The geological entity type and literature topic to be extracted are the task context, the identified entity set and attributes are the extracted information, and the credibility score of each data source is the multi-source data reliability label. The adjustment of the weights of multi-source data includes three basic actions and magnitude control. The three basic actions are to enhance text weight, increase map weight, and integrate multi-source weights. The weight adjustment range for enhancing text weight is +0.1 to +0.3, the weight adjustment range for increasing map weight is +0.1 to +0.3, and the weight adjustment range for integrating multi-source weights is ±0.05 for each source weight. The reward value is calculated based on the consistency between the extraction results and the geological knowledge map, data reliability, and task objectives. The expression is: ; in The degree of matching between entity attributes and geological knowledge graphs. To ensure the weight allocation aligns with the credibility of the data source, To ensure the completeness of entity attribute extraction under the current weight, For matching degree weight, As a weight for fit, Integrity weight; An evaluation network and a target network strategy are introduced, and the deep Q-network in reinforcement learning is optimized using the evaluation network and the target network. The deep Q-network structure consists of a 3-layer fully connected network and an activation function. The evaluation network outputs the Q-value of each action in real time. The input is the current state S, and the output is the Q-value of each action in the action space. The target network freezes and copies the parameters from the evaluation network periodically to calculate the target Q-value. The 3-layer fully connected network consists of an input layer, a hidden layer, and an output layer. Using initial weights based on a geological rule base as initial actions, the network is evaluated by inputting the current state S and outputting the Q-value of each action. The greedy strategy selects actions, adjusts weights, extracts entity attributes, calculates reward values, observes new states, and updates the Q-value. The expression is: ; in This is the current state. For the new state, As a discount factor, For the current action, For new actions, The evaluation network value under the new state and new action. The target network value under the current state and current action. To maximize the value of all possible new actions in the action space; Every 100 training steps, the parameters of the evaluation network are copied to the target network. The temporal difference error is calculated based on the absolute value of the difference between the evaluation network and the target network. Sample priorities are assigned according to the importance of the samples, and the sample sampling probability is calculated. Importance sampling weights are introduced to correct the loss function. The process is iterated until the policy network converges. The optimized weight set for the current task is then output as the optimized weights. In actual assessment, Greedy strategy selects actions: There is a 10% chance of random exploration and a 90% chance of choosing the action with the highest Q value. The sample sampling probability is calculated by dividing the sample priority by the sum of sample priorities. Weight adjustment is modeled as a Markov decision process, using the extraction of magmatic rock age attributes as the task context. - Greedy strategy ( The selected action increases the weight of the text data by 0.2; the R value is calculated based on the reward function, where the matching degree weight is 0.4, the fit degree weight is 0.3, and the integrity weight is 0.3. After 1000 steps of training, the policy network converges and outputs the optimized weight set.

[0021] In this embodiment, the method for obtaining fused data by fusing heterogeneous knowledge graphs based on the literature information data and the optimized weights includes: The system extracts text subgraphs, image subgraphs, structured subgraphs, and external data subgraphs from entities and relationships within the literature information data. It then constructs a multimodal subgraph based on these subgraphs and assigns fusion priorities to each subgraph based on optimized weights. / / Text subgraphs contain literature text entities and attribute relationships; weights are initialized according to entity extraction confidence. Image subgraphs contain spatial entities and coordinate attributes; weights are dynamically adjusted based on the recognition accuracy of deep bounding box networks and convolutional recurrent neural networks. Structured subgraphs; External data subgraphs. The BERT-base-cased model is used to encode the entity attribute text. A 128-dimensional feature vector is obtained through max pooling. The embedding space is optimized by fusing it with geological word vectors. The spatial topological relationship of map entities is learned by using hyperbolic graph convolutional network. The entity features are trained by adaptive gradient algorithm, and the depth, lithology and fossil triples are transformed into structured embedding vectors. By comparing the similarity of the embedding vectors of entities in different subgraphs, the geological knowledge graph reasoning engine is called to verify the rationality of the alignment, and subject labels are assigned to entities with the same name by combining optimized weights and context keywords. For multiple sources of evidence for the same relationship, the confidence level is calculated by weighting the evidence with optimized weights. A relationship confidence threshold of 0.6 is set. Relationships with a confidence level below the threshold are marked as requiring manual verification. Using verified relationships in the knowledge graph to infer missing relationships, and combining geophysical data subgraphs to complete relationships between entities across subgraphs; Output the entity ID, name, type, multi-source attributes, weight value, relationship ID, head / tail entity ID, relationship type, confidence level, and weight source as fused data; In the actual evaluation, the optimization weights are ranked as follows: text subgraph weights > structured subgraphs > map subgraphs > external data subgraphs; the spatial topological relationships of map entities are learned using hyperbolic graph convolutional networks, with an embedding dimension of 128. A multimodal subgraph is constructed from text, maps, structured data, and external geophysical data. Text subgraphs are assigned the highest priority based on optimized weights. An BERT-base-cased model is used to encode entity attribute text, obtaining 128-dimensional feature vectors. A hyperbolic graph convolutional network is used to learn the spatial topological relationships of map entities. The similarity of subgraph entity embedding vectors is compared, and subject labels are assigned to entities with the same fault name. Relationship confidence is calculated using optimized weights, and 28 relationships with confidence scores below 0.6 are marked for manual verification. 35 missing relationships across subgraphs are completed, and the fused data is output.

[0022] In this embodiment, the method for constructing a key information extraction model from geological literature based on the fused data includes: The TransE algorithm is used to map knowledge graph entities to a 128-dimensional vector space, so that the vector distance between semantically similar entities is less than the similarity threshold, and the distance between dissimilar entities is greater than the dissimilarity threshold. For multi-source entities, feature vectors are fused by weighted averaging, with weights dynamically allocated proportionally to vectorized entity features, and geological relationship path features are introduced to enhance relationship features. A framework for a key information extraction model from geological documents is constructed based on a multimodal feature encoder, a dynamic weight fusion layer, and a relational reasoning and information extraction layer. Multimodal feature encoder: Employing a geologically fine-tuned BERT input text fragment to output contextual feature vectors, text feature editing is obtained; based on the spatial coordinates of map entities and text content, visual features are extracted via a convolutional neural network, concatenated with text features, and weighted through an attention mechanism to obtain map feature encoding; for structured tabular data, Transformer-XL is used to capture long sequence dependencies, outputting temporal feature vectors to obtain table feature encoding; weights for text feature editing, map feature encoding, and table feature encoding are dynamically allocated through a multimodal attention layer. Dynamic Weight Fusion Layer: Based on reinforcement learning decision routing, dynamic weight adjustment of multi-source features is performed. The current task to be extracted, the fused feature vector, and the reliability label of multi-source data are defined as the state. The action space is defined by adjusting the weights of each modality feature and the given weight value range. The weights are adjusted by a given reward function, which includes positive reward and negative penalty. The positive reward is the matching of the extraction result with the geological knowledge graph, and the negative penalty is the entity type error. Relational Reasoning and Information Extraction Layer: BiLSTM-CRF is used to identify entity boundaries, and a geological entity dictionary is combined for entity boundary detection; attribute values ​​are bound to entities through an entity-attribute attention matrix to obtain attribute value associations; entity pair relation probabilities are calculated based on a knowledge graph embedding model, and relation reasoning is performed based on entity pair relation probabilities; the entity set, attribute set, and relation are output as extraction results; Training data is constructed, and multi-task joint training is performed using a main task and an auxiliary task. The main task includes entity recognition and relation extraction, while the auxiliary task is multi-source data weight prediction. Geological constraints are added during the training process to correct prediction results that violate the rules through penalty terms. In actual evaluation, entity attributes are standardized, relationships with confidence scores less than 0.5 are filtered out, and core relationship pairs are retained; the similarity threshold is 0.3, the heterogeneity threshold is 0.7, and the vector space is 128-dimensional. Training data: 5,000 geological documents were selected, and 120,000+ entities and 80,000+ relationships were manually annotated to form a geological annotation corpus; pseudo-label data was automatically generated from the "Geological Dictionary" and "Regional Geological Records" to expand the training sample to 300,000+. The TransE algorithm is used to map knowledge graph entities to a 128-dimensional vector space, constructing a framework consisting of a multimodal feature encoder, a dynamic weight fusion layer, and a relation reasoning and information extraction layer. The multimodal feature encoder uses a BERT fine-tuned approach from the geological domain to extract text features, a convolutional neural network to extract map features, and Transformer-XL to extract table features. The dynamic weight fusion layer adjusts modal weights based on reinforcement learning. The relation reasoning layer uses BiLSTM-CRF to identify entity boundaries and combines this with a knowledge graph embedding model to reason about relationships.

[0023] In this embodiment, the method for human-machine collaborative optimization of the geological literature key information extraction model based on the result deviation includes: Set confidence thresholds for entity recognition and relation extraction, and mark low-confidence samples as samples to be verified; call the geological knowledge graph inference engine to identify conflicting data that contradicts the domain rules, and the results include entity type conflicts, relational logic contradictions, and attribute value anomalies; Compare the extraction results of text, images, and tables: if the same entity has conflicting attributes in different data sources, mark it as a cross-modal conflict sample; classify and statistically analyze high-frequency error types by entity, relationship, and attribute, and push them to obtain an error sample pool; The system performs entity-level, relation-level, and attribute-level corrections on the erroneous sample pool. When experts correct entities, it automatically recommends standard attributes of similar entities in the knowledge graph and displays historical amendment examples of this type of error to assist experts in making quick decisions and obtaining corrected samples. Entity-level correction supports entity type adjustment and boundary calibration; relation-level correction provides a drop-down selection of relation types and supports relation direction adjustment; attribute-level correction supports unit calibration and error range supplementation for numerical attributes, and allows free editing of text attributes. The format of the corrected samples is standardized, and model parameters are extracted based on the corrected samples and key information from geological literature for domain adaptation fine-tuning and small sample learning optimization. In actual evaluation, the confidence threshold for entity recognition was 0.75, and the confidence threshold for relation extraction was 0.65. Domain adaptation fine-tuning: The hierarchical parameter update freezes the first 8 layers of the BERT pre-trained layer, and only fine-tunes the last 4 layers and the BiLSTM-CRF layer; a domain penalty term is added to the loss function to add extra loss for predictions that violate geological rules; Minimal learning optimization: SMOTE oversampling is used for low-frequency error types, expanding the sample size to 500+; the model converges after 30 iterations using a cosine annealing learning rate. 120 low-confidence samples were labeled, identifying 25 sets of entity type conflicts and 18 sets of contradictory relational logic. Experts corrected these conflicts by adjusting entity types, resolving relational directions, and calibrating numerical units. Based on the corrected samples, the model underwent domain-adaptation fine-tuning and few-sample learning optimization, converging after 30 iterations. The 100 new geological documents to be extracted were input into the optimized model, and more than 2,800 entities, more than 1,500 attributes, and more than 1,200 sets of relationships were successfully extracted. The entity recognition accuracy rate reached 92.5%, and the relationship extraction accuracy rate reached 89.3%. Key information such as the type, distribution, age, and contact relationship with the surrounding rocks of the Mesozoic magmatic rocks in the region were effectively extracted.

[0024] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for extracting key information from geological literature based on a large model, characterized by, The method comprises the following steps: Collecting geological literature data of a preset area, and preprocessing the geological literature data; the geological literature data comprises text data, drawing data, and table data; Performing joint information extraction on the geological multi-source data through a geological knowledge graph to obtain literature information data, and performing corresponding hyponym-hypernym relationship on the literature information data through a multi-granularity semantic alignment mechanism to obtain semantic alignment relationship; Setting an initial weight of the literature information data according to a geological rule base and the semantic alignment relationship, dynamically adjusting the initial weight based on reinforcement learning decision routing to obtain an optimized weight, and performing heterogeneous knowledge graph fusion on the literature information data and the optimized weight to obtain fusion data; Constructing a geological literature key information extraction model according to the fusion data, performing human-machine collaborative optimization on the geological literature key information extraction model according to result deviation, inputting to-be-extracted data into the geological literature key information extraction model, and outputting an extraction result. 2.The method of claim 1, wherein, The method for performing joint information extraction on the geological multi-source data through a geological knowledge graph to obtain literature information data comprises the following steps: For the text data: pre-training word vectors based on geological word vectors, and correcting the word vector space of easily confused words through a geological term dictionary; performing character-level BPE segmentation on rare terms, inputting the text into a BERT encoder, capturing context dependency through a bidirectional long short-term memory network, and obtaining an entity label sequence through a conditional random layer combined with a geological entity transition probability matrix; and matching the nodes of the geological knowledge graph through an entity disambiguation algorithm, assigning a subject label to the same-named entities through context similarity, and obtaining a text entity set; For the drawing data: detecting the text area of the drawing through a DBNet algorithm, recognizing the text content through a CRNN model, extracting the spatial coordinates of the text detection box, converting the spatial coordinates into actual geographic coordinates or relative positions combined with the scale of the drawing, establishing the spatial association between the text and the entity, and outputting a drawing entity set; For the table data: dividing the table into a high-density data area and a low-density labeling area based on text density and table line features, automatically extracting the three-dimensional data of depth, lithology, and fossils for a drilling table, standardizing the units of numerical data, mapping qualitative descriptions into quantitative attribute values, matching the attribute dictionary of the knowledge graph, and outputting a table entity set; Extracting entity pairs through a rule base combined with statistical learning, calling a geological knowledge graph reasoning engine to verify the relationship, automatically marking the contradictory relationship as to-be-confirmed, and outputting the non-contradictory relationship as a verified relationship set; wherein the rule-driven method: automatically establishing an entity plus numerical attribute relationship when an entity is followed by a numerical value; and the statistical-driven method: identifying the relationship through co-occurrence frequency combined with a BERT relationship classifier; Outputting the text entity set, the drawing entity set, the table entity set, and the verified relationship set as the literature information data. 3.The method of claim 1, wherein, The method for performing corresponding hyponym-hypernym relationship on the literature information data through a multi-granularity semantic alignment mechanism to obtain semantic alignment relationship comprises the following steps: The entities in the literature information data are granularly labeled, the granular rationality is verified through a geological ontology library, mixed granular entities are split or combined, and the entities in the literature information data are layered according to a 5-level granular system; the 5 levels are respectively a domain level, a discipline level, a type level, an instance level and an attribute level; The semantic association of different granular entities is verified based on word vector similarity and domain rules; Word vector similarity verification: the cosine similarity between entities is calculated by using an optimized geological word vector; the similarity threshold between entities at the domain level and the discipline level, between entities at the discipline level and the type level is greater than or equal to 0.85; the similarity threshold between entities at the type level and the instance level, between entities at the instance level and the attribute level is greater than or equal to 0.78; the entities with a similarity lower than the threshold but having a hierarchical relationship are verified by rules; Domain rule verification: if the definition of entity A contains the core features of entity B, the hierarchical relationship A→B is established; if entity B inherits all the core attributes of entity A, the hierarchical relationship is confirmed; the entities with semantic conflicts are marked as parallel relationships; A complete hierarchical relationship chain of different granular entities is constructed through hierarchical transmission and conflict resolution, and is output as a structured semantic alignment relationship table; hierarchical transmission automatically completes the relationship chain based on the transitivity rule; conflict resolution: when an entity has multiple potential superior concepts, the intermediate node is automatically inserted by selecting the intermediate level based on the discipline label. 4.The method of claim 1, wherein, A method for setting the initial weight of the literature information data according to a geological rule library and the semantic alignment relationship, comprising: The rule library includes entity type rules, data source rules and attribute importance rules, and the weight benchmark value is determined by expert consensus method; The benchmark weight is corrected by multi-granularity semantic alignment; the lower entity inherits the weight of the superior entity and improves the weight for the hierarchical relationship; the weight of the domain authority term is preferred for the synonymous relationship; According to the rule library, the entity type matching and data source matching of the entity in the literature information data are performed to obtain the entity type value and the data source weight; the benchmark value is corrected by semantic alignment to obtain a semantic alignment correction value; the entity type value, the data source weight and the semantic alignment correction value are weighted and added to obtain the entity weight; The attribute type matching of the attribute associated with the entity is performed, and the attribute weight is obtained by attribute value reliability correction according to quantitative attributes, qualitative data, accurate numerical values and fuzzy descriptions; The relationship type matching of the relationship between entities is performed, and the relationship weight is obtained by evidence strength correction according to multi-source evidence, single-source evidence, knowledge graph verification and to-be-verified relationship; the entity weight, the attribute weight and the relationship weight are output as the initial weight. 5.The method of claim 1, wherein, A method for dynamically adjusting the initial weight to obtain an optimized weight based on reinforcement learning decision routing, comprising: The weight adjustment task is modeled as a Markov decision process, and dynamic decision is made through a state, action and reward cycle; the current geological entity type to be extracted and the literature theme are the task context; the identified entity set and attribute are the extracted information; and the credibility score of each data source is the multi-source data reliability label. The adjustment of the weight of the multi-source data includes three types of basic actions and amplitude control. The three types of basic actions are enhancing the weight of the text, improving the weight of the map, and fusing the weight of the multi-source. The weight adjustment amplitude of enhancing the weight of the text is +0.1~+0.

3. The weight adjustment amplitude of improving the weight of the map is +0.1~+0.

3. The weight adjustment amplitude of fusing the weight of the multi-source is ±0.05 of each source weight; The reward value is calculated based on consistency of the extraction result with the geological knowledge graph, data reliability, and a task target , and the expression is: ； wherein is a matching degree of the entity attribute and the geological knowledge graph, is a fit degree of the weight distribution and the data source credibility, is an integrity of the entity attribute extraction under the current weight, is a matching degree weight, is a fit degree weight, is an integrity weight; The evaluation network and the target network strategy are introduced, and the deep Q network in the evaluation network and the target network optimization reinforcement learning is adopted. The deep Q network structure includes three layers of full connection network and activation function. The evaluation network: the Q value of each action is output in real time, and the input is the current state S, and the output is the Q value of each action in the action space. The target network: periodically frozen and copied from the evaluation network, used to calculate the target Q value. The three layers of full connection network are input layer, hidden layer and output layer respectively. The initial weight based on the geological rule base is taken as an initial action, the network input current state S is evaluated, the Q values of each action are output, and the The greedy strategy selects an action, extracts entity attributes after adjusting the weight, and calculates a reward value. The new state is observed to update the Q value, and the expression is as follows: ； wherein is the current state, is the new state, is the discount factor, is the current action, is the new action, is the evaluation network value for the new state new action, is the target network value for the current state current action, is the maximum over all possible new actions in the action space; Every 100 steps of training, the evaluation network parameters are copied to the target network. The absolute value of the difference between the evaluation network and the target network is used to calculate the time difference error. The sample priority is allocated according to the importance of the sample, and the sample sampling probability is calculated. The importance sampling weight is introduced to correct the loss function, and the optimization weight set of the current task is output as the optimization weight. 6.The method of claim 1, wherein, According to the literature information data and the optimization weight, a method for fusing heterogeneous knowledge graphs is provided to obtain fused data, which includes: The entity and relationship in the literature information data are used to obtain a text subgraph, a map subgraph, a structured subgraph and an external data subgraph. The text subgraph, the map subgraph, the structured subgraph and the external data subgraph are used to construct a multi-modal subgraph. The optimization weight is used to give a fusion priority to each subgraph. The text subgraph includes literature text entities and attribute relationships. The weight is initialized according to the entity extraction confidence. The map subgraph includes spatial entities and coordinate attributes. The weight is dynamically adjusted according to the deep bounding box network and the convolutional recurrent neural network recognition accuracy. The structured subgraph and the external data subgraph are obtained. The BERT-base-cased model is used to encode the entity attribute text. The 128-dimensional feature vector is obtained by maximum pooling. The embedding space is optimized by fusing the geological word vector. The spatial topological relationship of the map entity is learned by using the hyperbolic graph convolutional network. The entity features are trained by using the adaptive gradient algorithm. The depth, lithology and fossil triplets are converted into structured embedding vectors. The embedding vector similarity of different subgraph entities is compared. The geological knowledge graph reasoning engine is called to verify the alignment rationality. The same named entity is combined with the optimization weight and the context keyword to give a discipline label. The confidence of the multi-source evidence of the same relationship is calculated by weighting according to the optimization weight. The relationship confidence threshold is set to 0.

6. The relationship below the relationship confidence threshold is marked for manual verification. The verified relationship in the knowledge graph is used to infer the missing relationship. The cross-subgraph relationship between entities is completed in combination with the geophysical data subgraph. The entity ID, name, type, multi-source attribute, weight value, relationship ID, head and tail entity ID, relationship type, confidence, and weight source are output as the fused data. 7.The method of claim 1, wherein, According to the fused data, a method for constructing a geological literature key information extraction model is provided, which includes: The TransE algorithm is used to map the entities of the knowledge graph to the vector space, so that the vector distance of semantically similar entities is less than the similarity threshold, and the distance of dissimilar entities is greater than the dissimilarity threshold; The feature vectors of multiple source entities are fused by weighted average, the weights are dynamically allocated to the vectorized entity features in proportion, and the geological relationship path features are introduced to enhance the relationship features; According to the multi-modal feature encoder, the dynamic weight fusion layer, and the relationship reasoning and information extraction layer, a framework of the geological literature key information extraction model is constructed. The multi-modal feature encoder: the BERT input text segment is fine-tuned in the geological field to output the context feature vector, and the text feature editing is obtained; based on the spatial coordinates of the map entity and the text content, the visual features and the text features are extracted by the convolutional neural network, and then the map feature coding is obtained by weighting through the attention mechanism; the long sequence dependence of the structured table data is captured by the Transformer-XL, and the time sequence feature vector is output to obtain the table feature coding; the weights of the text feature editing, the map feature coding, and the table feature coding are dynamically allocated through the multi-modal attention layer; The dynamic weight fusion layer: based on the reinforcement learning decision routing, the dynamic weight adjustment of the multi-source features is performed, the current extraction task to be extracted, the fused feature vector, and the reliability label of the multi-source data are defined as the state; the adjustment of the feature weight of each mode and the given weight value range are regarded as the action space; the weight is adjusted by the given reward function, and the reward function includes positive reward and negative punishment, the positive reward is that the extraction result matches the geological knowledge graph, and the negative punishment is that the entity type is wrong; The relationship reasoning and information extraction layer: the BiLSTM-CRF is used to identify the entity boundary, and the entity boundary detection is performed in combination with the geological entity dictionary; the attribute value is bound to the entity through the entity-attribute attention matrix to obtain the attribute value association; the entity pair relationship probability is calculated based on the knowledge graph embedding model, and the relationship reasoning is performed according to the entity pair relationship probability; the entity set, the attribute set, and the relationship are output as the extraction result; The training data is constructed, and the main task and the auxiliary task are used for multi-task joint training; in the training process, the geological constraint is added to correct the prediction results that violate the rules through the penalty term. 8.The method of claim 1, wherein, According to the result deviation, the method for human-computer collaborative optimization of the geological literature key information extraction model comprises the following steps: Set the entity recognition confidence threshold and the relationship extraction confidence threshold, and mark the low confidence as a sample to be checked; call the geological knowledge graph reasoning engine to identify the conflict data that is inconsistent with the field rules, and the results include entity type conflict, relationship logic contradiction, and attribute value anomaly; Compare the text, map, and table three-mode extraction results: if the attributes of the same entity conflict in different data sources, mark it as a cross-modal conflict sample; classify and count the high-frequency error types according to the entity, relationship, and attribute, and push them to obtain an error sample pool; The error sample pool is corrected at an entity layer, a relation layer and an attribute layer. When an expert corrects an entity, standard attributes of a same type of entity in a knowledge graph are automatically recommended, historical correction cases of the same type of error are displayed to assist the expert in making a quick decision, and a corrected sample is obtained. The entity layer correction supports entity type adjustment and boundary calibration. The relation layer correction provides a relation type drop-down selection and supports relation direction adjustment. The attribute layer correction supports unit calibration and error range supplementation for numerical attributes, and supports free editing for text attributes. The corrected samples are uniformly formatted, and model parameters are extracted according to the corrected samples and geological literature key information to perform domain adaptation fine-tuning and small sample learning optimization.

Images