Trap feature identification method and device

Abstract

The invention provides a trap feature recognition method and device, and the method comprises the steps: carrying out the word segmentation, coding and vectorization processing of obtained geology and exploration data; establishing an underground trap feature recognition model based on a pre-constructed trap knowledge graph and the vectorized geological and exploration data; extracting target geology and exploration data thereof from the obtained target document; performing word segmentation, coding and vectorization processing on the target geology and exploration data thereof; and inputting the vectorized target geology and exploration data thereof into the underground trap feature recognition model to obtain trap features corresponding to the target document. An underground trap feature recognition model is established by constructing a trap knowledge spectrogram and cooperating with geology and exploration data, and then trap feature information such as stratum, geological attributes and trap types in a target document is obtained through the underground trap feature recognition model.

Description

Technical Field

[0001] The present invention relates to the field of information processing technologies, and particularly to a method and apparatus for identifying trap features. Background Art

[0002] Subsurface traps are an important part of geological structures and usually contain important resources such as oil, natural gas, water, and minerals. Traditional methods for identifying trap features rely on expert experience and manual analysis, suffering from problems such as low efficiency, strong subjectivity, and high costs. Summary of the Invention

[0003] The present invention provides a method and apparatus for identifying trap features, which analyze trap feature information such as the formation, geological attributes, and trap types in a document based on the identification results provided by machine learning.

[0004] In a first aspect, the present invention provides a method for identifying trap features, including:

[0005] Performing word segmentation, encoding, and vectorization processing on the obtained geology and its exploration data;

[0006] Based on a pre-constructed trap knowledge graph and the geology and its exploration data after vectorization processing, establishing a subsurface trap feature identification model;

[0007] Extracting target geology and its exploration data from the obtained target document;

[0008] Performing word segmentation, encoding, and vectorization processing on the target geology and its exploration data;

[0009] Inputting the target geology and its exploration data after vectorization processing into the subsurface trap feature identification model to obtain the trap features corresponding to the target document.

[0010] Optionally, the subsurface trap feature identification model includes: an unsupervised trap feature identification model; inputting the target geology and its exploration data after vectorization processing into the subsurface trap feature identification model to obtain the trap features corresponding to the target document, including:

[0011] Inputting the target geology and its exploration data after vectorization processing into the trained unsupervised trap feature identification model to determine the similarity of the word segments in the target geology and its exploration data after vectorization processing in their respective sentences;

[0012] According to the similarity, determining candidate keywords in the target geology and its exploration data;

[0013] Using a clustering algorithm and similarity calculation to select the trap features from the candidate keywords.

[0014] Optionally, the underground trap feature recognition model includes: a supervised trap feature recognition model; inputting the vectorized target geology and its exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document, including:

[0015] Inputting the vectorized target geology and its exploration data into the trained supervised trap feature recognition model to obtain a corresponding output vector;

[0016] Feeding the output vector into an activation function to obtain a corresponding predicted category; the predicted category is the trap feature.

[0017] Optionally, based on a pre-constructed trap knowledge graph and the vectorized geology and its exploration data, an underground trap feature recognition model is established, including:

[0018] Fusing and comparing the geological knowledge in the trap knowledge graph with the vectorized geology and its exploration data to obtain training corpus;

[0019] Training the constructed unsupervised trap feature recognition model with the training corpus to obtain the trained unsupervised trap feature recognition model;

[0020] Using the annotation information in the training corpus as labels to train the constructed supervised trap feature recognition model with the training corpus to obtain the trained supervised trap feature recognition model.

[0021] In a second aspect, the present invention provides a trap feature recognition device, including:

[0022] A data preprocessing module for performing word segmentation, encoding, and vectorization processing on the obtained geology and its exploration data;

[0023] A model establishment module for establishing an underground trap feature recognition model based on a pre-constructed trap knowledge graph and the vectorized geology and its exploration data;

[0024] A data extraction module for extracting target geology and its exploration data from the obtained target document;

[0025] A target data preprocessing module for performing word segmentation, encoding, and vectorization processing on the target geology and its exploration data;

[0026] A recognition module for inputting the vectorized target geology and its exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document.

[0027] Optionally, the underground trap feature recognition model includes: an unsupervised trap feature recognition model; the recognition module includes:

[0028] An input sub-module, configured to input the vectorized target geology and its exploration data into the trained unsupervised trap feature recognition model, and determine the similarity of the word segments in the vectorized target geology and its exploration data in their respective sentences;

[0029] A candidate keyword determination sub-module, configured to determine candidate keywords in the target geology and its exploration data according to the similarity;

[0030] A trap feature determination sub-module, configured to select the trap features from the candidate keywords by using a clustering algorithm and similarity calculation.

[0031] Optionally, the underground trap feature recognition model includes: a supervised trap feature recognition model; the input sub-module includes:

[0032] A first input unit, configured to input the vectorized target geology and its exploration data into the trained supervised trap feature recognition model to obtain a corresponding output vector;

[0033] A second input unit, configured to input the output vector into an activation function to obtain a corresponding predicted category; the predicted category is the trap feature.

[0034] Optionally, the model establishment module includes:

[0035] Fuse and compare the geological knowledge in the trap knowledge graph with the vectorized geology and its exploration data to obtain training corpus;

[0036] Use the training corpus to train the constructed unsupervised trap feature recognition model to obtain the trained unsupervised trap feature recognition model;

[0037] Use the training corpus with the annotation information in the training corpus as labels to train the constructed supervised trap feature recognition model to obtain the trained supervised trap feature recognition model.

[0038] In a third aspect, the present application provides an electronic device, including a processor and a memory, where the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps in the method provided in the first aspect above are run, including:

[0039] Perform word segmentation, encoding, and vectorization processing on the obtained geology and its exploration data;

[0040] Based on the pre-constructed trap knowledge graph and the geological and exploration data after vectorization processing, an underground trap feature recognition model is established;

[0041] Extract the target geological and exploration data from the obtained target document;

[0042] Perform word segmentation, encoding, and vectorization processing on the target geological and exploration data;

[0043] Input the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document.

[0044] Optionally, the underground trap feature recognition model includes: an unsupervised trap feature recognition model; inputting the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document includes:

[0045] Input the vectorized target geological and exploration data into the trained unsupervised trap feature recognition model to determine the similarity of the word segmentation in the vectorized target geological and exploration data in its corresponding sentence;

[0046] Determine the candidate keywords in the target geological and exploration data according to the similarity;

[0047] Use clustering algorithms and similarity calculations to select the trap features from the candidate keywords.

[0048] Optionally, the underground trap feature recognition model includes: a supervised trap feature recognition model; inputting the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document includes:

[0049] Input the vectorized target geological and exploration data into the trained supervised trap feature recognition model to obtain the corresponding output vector;

[0050] Send the output vector into an activation function to obtain the corresponding predicted category; the predicted category is the trap feature.

[0051] Optionally, based on the pre-constructed trap knowledge graph and the geological and exploration data after vectorization processing, establishing an underground trap feature recognition model includes:

[0052] Fuse and compare the geological knowledge in the trap knowledge graph with the vectorized geological and exploration data to obtain training corpus;

[0053] Train the constructed unsupervised trap feature recognition model using the training corpus to obtain the trained unsupervised trap feature recognition model;

[0054] Using the annotation information in the training corpus as labels, train the constructed supervised trap feature recognition model using the training corpus to obtain the trained supervised trap feature recognition model.

[0055] Fourthly, the present application provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, it runs the steps in the method provided in the first aspect above, including:

[0056] Perform word segmentation, encoding, and vectorization processing on the obtained geological and exploration data;

[0057] Based on the pre-constructed trap knowledge graph and the vectorized geological and exploration data, establish an underground trap feature recognition model;

[0058] Extract target geological and exploration data from the obtained target document;

[0059] Perform word segmentation, encoding, and vectorization processing on the target geological and exploration data;

[0060] Input the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document.

[0061] Optionally, the underground trap feature recognition model includes: an unsupervised trap feature recognition model; inputting the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document includes:

[0062] Input the vectorized target geological and exploration data into the trained unsupervised trap feature recognition model to determine the similarity of the word segmentation in the vectorized target geological and exploration data in its corresponding sentence;

[0063] According to the similarity, determine the candidate keywords in the target geological and exploration data;

[0064] Use clustering algorithms and similarity calculations to select the trap features from the candidate keywords.

[0065] Optionally, the underground trap feature recognition model includes: a supervised trap feature recognition model; inputting the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document includes:

[0066] Input the target geology and its exploration data after vectorization into the trained supervised trap feature recognition model to obtain a corresponding output vector;

[0067] Send the output vector into an activation function to obtain a corresponding predicted category; the predicted category is the trap feature.

[0068] Optionally, based on a pre-constructed trap knowledge graph and the geology and its exploration data after vectorization, establish an underground trap feature recognition model, including:

[0069] Fuse and compare the geological knowledge in the trap knowledge graph with the geology and its exploration data after vectorization to obtain training corpus;

[0070] Use the training corpus to train the constructed unsupervised trap feature recognition model to obtain the trained unsupervised trap feature recognition model;

[0071] Use the annotation information in the training corpus as labels to train the constructed supervised trap feature recognition model to obtain the trained supervised trap feature recognition model.

[0072] It can be seen from the above technical solutions that the present invention has the following advantages:

[0073] The present invention provides a method and device for identifying trap features. The method includes: performing word segmentation, encoding, and vectorization processing on the obtained geology and its exploration data; establishing an underground trap feature recognition model based on a pre-constructed trap knowledge graph and the geology and its exploration data after vectorization; extracting the target geology and its exploration data from the obtained target document; performing word segmentation, encoding, and vectorization processing on the target geology and its exploration data; inputting the target geology and its exploration data after vectorization into the underground trap feature recognition model to obtain the trap features corresponding to the target document. By constructing a trap knowledge spectrum diagram and jointly establishing an underground trap feature recognition model with the geology and its exploration data, trap feature information such as the formation, geological attributes, and trap types belonging to the target document can then be obtained through the underground trap feature recognition model. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0075] Figure 1 Flow chart of the first embodiment of the method for identifying trap features of the present invention;

[0076] Figure 2 Flow chart of the second embodiment of the method for identifying trap features of the present invention;

[0077] Figure 3 Relevant feature map of the second embodiment of the method for identifying trap features of the present invention;

[0078] Figure 4 Schematic diagram of the vectorization process of the Bert model in the second embodiment of the method for identifying trap features of the present invention;

[0079] Figure 5 Schematic diagram of identifying feature information by the clustering algorithm and similarity judgment algorithm in the second embodiment of the method for identifying trap features of the present invention;

[0080] Figure 6 Structure block diagram of the embodiment of the device for identifying trap features of the present invention. Detailed implementation manners

[0081] The embodiments of the present invention provide a method and device for identifying trap features, and analyze trap feature information such as the formation, geological attributes, and trap types in the document through the recognition results provided by machine learning.

[0082] To make the objectives, features, and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the embodiments described below are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0083] Embodiment 1. Please refer to Figure 1 , Figure 1 which is a flow chart of the first embodiment of the method for identifying trap features of the present invention, including:

[0084] S101, perform word segmentation, encoding, and vectorization processing on the obtained geological and exploration data;

[0085] S102, establish an underground trap feature recognition model based on the pre-constructed trap knowledge graph and the vectorized geological and exploration data;

[0086] S103, extract the target geological and exploration data from the obtained target document;

[0087] S104. Perform word segmentation, encoding, and vectorization processing on the target geology and its exploration data;

[0088] S105. Input the vectorized target geology and its exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document;

[0089] In an optional embodiment, the underground trap feature recognition model includes: an unsupervised trap feature recognition model; inputting the vectorized target geology and its exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document includes:

[0090] Input the vectorized target geology and its exploration data into the trained unsupervised trap feature recognition model to determine the similarity of the word segments in the vectorized target geology and its exploration data in their respective sentences;

[0091] Determine the candidate keywords in the target geology and its exploration data according to the similarity;

[0092] Use clustering algorithms and similarity calculations to select the trap features from the candidate keywords.

[0093] It should be noted that the trap feature refers to the feature used to describe the potential underground oil and gas reservoir conditions in the fields of geology and oil and gas exploration. The trap feature conducts research on the characteristics of reservoir rock properties, structures, faults, trap covers, etc., and is mainly used to predict and identify the existence and potential of underground oil and gas traps.

[0094] The clustering algorithm is an unsupervised learning method used to divide a set of objects into subsets with similarity, called clusters. The clustering algorithm aims to identify the inherent patterns and structures in the data, so that the objects within the same cluster have high similarity, while the objects between different clusters have low similarity.

[0095] In a trap feature recognition method provided by an embodiment of the present invention, it includes: performing word segmentation, encoding, and vectorization processing on the obtained geology and its exploration data; establishing an underground trap feature recognition model based on a pre-constructed trap knowledge graph and the vectorized geology and its exploration data; extracting the target geology and its exploration data from the obtained target document; performing word segmentation, encoding, and vectorization processing on the target geology and its exploration data; inputting the vectorized target geology and its exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document. By constructing a trap knowledge spectrum diagram and jointly establishing an underground trap feature recognition model with geology and its exploration data, and then obtaining trap feature information such as the formation, geological attributes, and trap types in the target document through the underground trap feature recognition model.

[0096] Example 2. Please refer to Figure 2 , Figure 2 which is a flowchart of the steps of Example 2 of a method for identifying trap characteristics of the present invention. The method includes:

[0097] Step S201: Perform word segmentation, encoding, and vectorization processing on the obtained geological and exploration data.

[0098] In the embodiment of the present invention, geological and exploration data including seismic data, geological profiles, and gold data are obtained.

[0099] In a specific implementation, in order to ensure the consistency and availability of geological and exploration data, quality control and preprocessing are performed on the geological and exploration data.

[0100] Step S202: Fuse and compare the geological knowledge in the trap knowledge graph with the vectorized geological and exploration data to obtain training corpus.

[0101] In the embodiment of the present invention, a trap knowledge graph including multi-dimensional information such as geology, stratigraphy, and rock properties is established in advance. The trap knowledge graph will include key information such as stratigraphic distribution, geological properties, and trap types, as well as the complex relationships between them. Then, the geological knowledge obtained from the trap knowledge graph is fused and compared with the data characteristics to obtain keywords corresponding to the geological and exploration data, thereby forming a training corpus.

[0102] In a specific implementation, the continuous update and expansion of the knowledge graph will make it a powerful reference tool.

[0103] Step S203: Use the training corpus to train the constructed unsupervised trap feature recognition model to obtain the trained unsupervised trap feature recognition model.

[0104] It should be noted that the unsupervised training of the unsupervised trap feature recognition model is a machine learning method that does not rely on labeled training data, and the evaluation process usually quantifies the ability of the model to learn the structure and pattern in the data.

[0105] Step S204: Use the training corpus to train the constructed supervised trap feature recognition model with the annotation information in the training corpus as labels to obtain the trained supervised trap feature recognition model.

[0106] It should be noted that the supervised training of the supervised trap feature recognition model is a machine learning method in which the model is trained using labeled training data.

[0107] In an embodiment of the present invention, geological and exploration data are used as training corpus, and keywords are used as annotation information to train the constructed supervised training model, so as to obtain a trained supervised trap feature recognition model.

[0108] The embodiment of the present invention uses natural language processing technology to automatically extract keywords, which may include geological physical properties such as formation thickness, density, acoustic velocity, and resistivity.

[0109] In specific implementation, on the one hand, the importance and relevance of features need to be considered; on the other hand, an artificial supervision method is required to train the model with keywords, and then keyword extraction is performed on the document that needs to extract keywords based on the model. Therefore, the model training process of the embodiment of the present invention includes unsupervised training and supervised training.

[0110] Unsupervised training needs to extract keywords of the document by using the statistical information of words in the training corpus. These statistical information mainly include word weights, word positions in the document, and word association information, such as Figure 3 shown. This method mainly depends on feature selection, and the quality of feature selection has a great impact on the effect.

[0111] Supervised training needs to extract keywords of the document that needs to extract keywords according to the model according to the already annotated training corpus.

[0112] Step S205: Extract target geological and exploration data from the obtained target document;

[0113] Step S206: Perform word segmentation, encoding, and vectorization processing on the target geological and exploration data;

[0114] In an embodiment of the present invention, the target document needs to be processed first: first, word segmentation is performed on the target document, and a language network graph of the target document is constructed, and then the language network graph is analyzed to find important words or phrases on this graph, that is, initial keywords. After obtaining the initial keywords of the target document, the initial keywords are encoded, and the Bert model is used to perform vectorization processing on the words and the sentences where the words are located, as Figure 4 shown.

[0115] Among them, the word segmentation process is as follows:

[0116] For example, for the following passage: "The research results of this project can support the exploration and development of future petrochemical shale oil and gas and reserve management work", the analysis results are as follows:

[0117]

[0118] Step S207: Input the vectorized target geology and its exploration data into the trained unsupervised trap feature recognition model to determine the similarity of the segmented words in the vectorized target geology and its exploration data within their respective sentences.

[0119] Step S208: Determine the candidate keywords in the target geology and its exploration data based on the similarity.

[0120] Step S209: Select the trap features from the candidate keywords by using clustering algorithms and similarity calculations.

[0121] In the embodiment of the present invention, after obtaining the vectorized representations of each sentence and the words in the sentence, input them into the trained unsupervised trap feature recognition model, convert the vectorized target geology and its exploration data into latent variable vectors, which can represent the semantic features of the text, and then use cosine similarity to calculate the similarity between each word and its sentence. Set a threshold during the calculation process, and retain the words with similarity greater than the threshold as candidate keywords. Finally, use clustering algorithms and similarity judgment algorithms to identify feature information such as the formation, geological attributes, and trap types to which they belong. The specific results of trap feature selection are as Figure 5 shown.

[0122] Step S210: Input the vectorized target geology and its exploration data into the trained supervised trap feature recognition model to obtain the corresponding output vector.

[0123] Step S211: Feed the output vector into an activation function to obtain the corresponding predicted class; the predicted class is the trap feature.

[0124] In the embodiment of the present invention, input the vectorized target geology and its exploration data into the trained supervised trap feature recognition model, and feed the obtained output vector into an activation function, such as the Softmax function, to obtain the corresponding predicted class, that is, the trap feature.

[0125] Whether for long texts or short texts, the main theme of the entire text can often be glimpsed through some keywords. At the same time, whether it is text-based recommendation or text-based search, there is also a great dependence on text keywords. The accuracy of keyword extraction is directly related to the final effect of the recommendation system or the search system. Therefore, keyword extraction is a very important part of the text mining field. A method for identifying trap features in an embodiment of the present invention performs word segmentation, encoding, and vectorization processing on the obtained geological and exploration data; based on a pre-constructed trap knowledge graph and the vectorized geological and exploration data, an underground trap feature recognition model is established; from the obtained target document, target geological and exploration data are extracted; the target geological and exploration data are subjected to word segmentation, encoding, and vectorization processing; the vectorized target geological and exploration data are input into the underground trap feature recognition model to obtain the trap features corresponding to the target document. By constructing a trap knowledge spectrum diagram and collaborating with geological and exploration data to establish an underground trap feature recognition model including an unsupervised trap feature recognition model and a supervised trap feature recognition model, and then obtaining the keyword recognition result of the target document through the underground trap feature recognition model, so as to determine trap feature information such as the formation, geological attributes, and trap types belonging to the target document.

[0126] Embodiment 3, please refer to Figure 6 , Figure 6 is a structural block diagram of an embodiment of an apparatus for identifying trap features of the present invention, including:

[0127] A data preprocessing module 301 is configured to perform word segmentation, encoding, and vectorization processing on the obtained geological and exploration data;

[0128] A model establishment module 302 is configured to establish an underground trap feature recognition model based on a pre-constructed trap knowledge graph and the vectorized geological and exploration data;

[0129] A data extraction module 303 is configured to extract target geological and exploration data from the obtained target document;

[0130] A target data preprocessing module 304 is configured to perform word segmentation, encoding, and vectorization processing on the target geological and exploration data;

[0131] An identification module 305 is configured to input the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document.

[0132] In an alternative embodiment, the underground trap feature recognition model includes: an unsupervised trap feature recognition model; the identification module 305 includes:

[0133] An input sub-module, configured to input the target geology and its exploration data after vectorization processing into the trained unsupervised trap feature recognition model, and determine the similarity of the word segmentation in the target geology and its exploration data after vectorization processing in their respective sentences;

[0134] A candidate keyword determination sub-module, configured to determine candidate keywords in the target geology and its exploration data according to the similarity;

[0135] A trap feature determination sub-module, configured to select the trap features from the candidate keywords by using a clustering algorithm and similarity calculation.

[0136] In an optional embodiment, the underground trap feature recognition model includes: a supervised trap feature recognition model; the input sub-module includes:

[0137] A first input unit, configured to input the target geology and its exploration data after vectorization processing into the trained supervised trap feature recognition model to obtain a corresponding output vector;

[0138] A second input unit, configured to send the output vector into an activation function to obtain a corresponding predicted class; the predicted class is the trap feature.

[0139] In an optional embodiment, the model establishment module 302 includes:

[0140] Fusing and comparing the geological knowledge in the trap knowledge graph with the geology and its exploration data after vectorization processing to obtain training corpus;

[0141] Training the constructed unsupervised trap feature recognition model by using the training corpus to obtain the trained unsupervised trap feature recognition model;

[0142] Using the annotation information in the training corpus as labels, and training the constructed supervised trap feature recognition model by using the training corpus to obtain the trained supervised trap feature recognition model.

[0143] Embodiment 4, The embodiment of the present invention further provides an electronic device, including a memory and a processor, where a computer program is stored in the memory, and when the computer program is executed by the processor, the processor is caused to execute the steps of a method for recognizing a trap feature as described in any one of the above embodiments, including:

[0144] Performing word segmentation, encoding and vectorization processing on the obtained geology and its exploration data;

[0145] Based on the pre - constructed trap knowledge graph and the geological and exploration data after vectorization processing, an underground trap feature recognition model is established;

[0146] Extract the target geological and exploration data from the obtained target document;

[0147] Perform word segmentation, encoding, and vectorization processing on the target geological and exploration data;

[0148] Input the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document.

[0149] In an alternative embodiment, the underground trap feature recognition model includes: an unsupervised trap feature recognition model; inputting the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document includes:

[0150] Input the vectorized target geological and exploration data into the trained unsupervised trap feature recognition model to determine the similarity of the word segments in the vectorized target geological and exploration data in their respective sentences;

[0151] Determine the candidate keywords in the target geological and exploration data according to the similarity;

[0152] Use a clustering algorithm and similarity calculation to select the trap features from the candidate keywords.

[0153] In an alternative embodiment, the underground trap feature recognition model includes: a supervised trap feature recognition model; inputting the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document includes:

[0154] Input the vectorized target geological and exploration data into the trained supervised trap feature recognition model to obtain the corresponding output vector;

[0155] Send the output vector into an activation function to obtain the corresponding predicted category; the predicted category is the trap feature.

[0156] In an alternative embodiment, based on the pre - constructed trap knowledge graph and the geological and exploration data after vectorization processing, establishing an underground trap feature recognition model includes:

[0157] Fuse and compare the geological knowledge in the trap knowledge graph with the vectorized geological and exploration data to obtain training corpus;

[0158] Training the constructed unsupervised trap feature recognition model with the training corpus to obtain the trained unsupervised trap feature recognition model;

[0159] Using the annotation information in the training corpus as labels, training the constructed supervised trap feature recognition model with the training corpus to obtain the trained supervised trap feature recognition model.

[0160] Embodiment 5, The embodiment of the present invention also provides a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of a method for identifying trap features as described in any of the above embodiments, including:

[0161] Performing word segmentation, encoding, and vectorization processing on the obtained geological and exploration data;

[0162] Based on a pre-constructed trap knowledge graph and the vectorized geological and exploration data, establishing an underground trap feature recognition model;

[0163] Extracting target geological and exploration data from the obtained target document;

[0164] Performing word segmentation, encoding, and vectorization processing on the target geological and exploration data;

[0165] Inputting the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document.

[0166] In an alternative embodiment, the underground trap feature recognition model includes: an unsupervised trap feature recognition model; Inputting the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document, including:

[0167] Inputting the vectorized target geological and exploration data into the trained unsupervised trap feature recognition model to determine the similarity of the word segmentation in the vectorized target geological and exploration data in its corresponding sentence;

[0168] Based on the similarity, determining candidate keywords in the target geological and exploration data;

[0169] Using a clustering algorithm and similarity calculation to select the trap features from the candidate keywords.

[0170] In an alternative embodiment, the underground trap feature recognition model includes: a supervised trap feature recognition model; inputting the vectorized target geology and its exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document, including:

[0171] Inputting the vectorized target geology and its exploration data into the trained supervised trap feature recognition model to obtain a corresponding output vector;

[0172] Feeding the output vector into an activation function to obtain a corresponding predicted category; the predicted category is the trap feature.

[0173] In an alternative embodiment, based on a pre-constructed trap knowledge graph and the vectorized geology and its exploration data, an underground trap feature recognition model is established, including:

[0174] Fusing and comparing the geological knowledge in the trap knowledge graph with the vectorized geology and its exploration data to obtain training corpus;

[0175] Training the constructed unsupervised trap feature recognition model with the training corpus to obtain the trained unsupervised trap feature recognition model;

[0176] Using the annotation information in the training corpus as labels, training the constructed supervised trap feature recognition model with the training corpus to obtain the trained supervised trap feature recognition model.

[0177] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0178] In several embodiments provided in the present application, it should be understood that the methods, devices, electronic devices, and storage media disclosed by the present invention can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, and the indirect coupling or communication connection of the devices or units may be in an electrical, mechanical, or other form.

[0179] The unit described as a separation component may or may not be physically separated. The component displayed as a unit may or may not be a physical unit, that is, it may be located in one place or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0180] In addition, each functional unit in various embodiments of the present invention may be integrated in a processing unit, or each unit may exist physically alone, or two or more units may be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0181] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned readable storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs, etc., which can store program codes.

[0182] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of various embodiments of the present invention.

Claims

1. A method for identifying trap characteristics, characterized in that, Including: Performing word segmentation, encoding, and vectorization processing on the obtained geological and exploration data; Based on the pre-constructed trap knowledge graph and the vectorized geological and exploration data, establishing an underground trap feature recognition model; Extracting target geological and exploration data from the obtained target document; Performing word segmentation, encoding, and vectorization processing on the target geological and exploration data; Inputting the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document.

2. The method for identifying trap characteristics according to claim 1, characterized in that, The underground trap feature recognition model includes: an unsupervised trap feature recognition model; inputting the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document, including: Inputting the vectorized target geological and exploration data into the trained unsupervised trap feature recognition model to determine the similarity of the word segments in the vectorized target geological and exploration data in their respective sentences; Determining candidate keywords in the target geological and exploration data according to the similarity; Selecting the trap features from the candidate keywords by using a clustering algorithm and similarity calculation.

3. The method for identifying trap characteristics according to claim 2, characterized in that, The underground trap feature recognition model includes: a supervised trap feature recognition model; inputting the vectorized target geological and exploration data into the underground trap feature recognition model to obtain the trap features corresponding to the target document, including: Inputting the vectorized target geological and exploration data into the trained supervised trap feature recognition model to obtain a corresponding output vector; Feeding the output vector into an activation function to obtain a corresponding predicted category; the predicted category is the trap feature.

4. The method for identifying trap characteristics according to claim 3, characterized in that, Based on the pre-constructed trap knowledge graph and the vectorized geological and exploration data, establishing an underground trap feature recognition model, including: Fusing and comparing the geological knowledge in the trap knowledge graph with the vectorized geological and exploration data to obtain training corpus; Using the training corpus to train the constructed unsupervised trap feature recognition model to obtain the trained unsupervised trap feature recognition model; Using the annotation information in the training corpus as labels and using the training corpus to train the constructed supervised trap feature recognition model to obtain the trained supervised trap feature recognition model.

5. An apparatus for identifying trap characteristics, characterized in that, Including: A data preprocessing module for performing word segmentation, encoding, and vectorization processing on the obtained geological and exploration data; A model establishment module for establishing an underground trap feature recognition model based on the pre-constructed trap knowledge graph and the vectorized geological and exploration data; A data extraction module for extracting target geological and exploration data from the obtained target document; A target data preprocessing module for performing word segmentation, encoding, and vectorization processing on the target geological and exploration data; An identification module, configured to input the vectorized target geology and its exploration data into the underground trap feature identification model, and obtain the trap features corresponding to the target document.

6. The apparatus for identifying trap characteristics according to claim 5, characterized in that, The underground trap feature identification model includes: an unsupervised trap feature identification model; the identification module includes: An input sub-module, configured to input the vectorized target geology and its exploration data into the trained unsupervised trap feature identification model, and determine the similarity of the word segments in the vectorized target geology and its exploration data in their respective sentences. A candidate keyword determination sub-module, configured to determine the candidate keywords in the target geology and its exploration data according to the similarity. A trap feature determination sub-module, configured to select the trap features from the candidate keywords by using a clustering algorithm and similarity calculation.

7. The apparatus for identifying trap characteristics according to claim 6, characterized in that, The underground trap feature identification model includes: a supervised trap feature identification model; the input sub-module includes: A first input unit, configured to input the vectorized target geology and its exploration data into the trained supervised trap feature identification model to obtain a corresponding output vector. A second input unit, configured to input the output vector into an activation function to obtain a corresponding predicted category; the predicted category is the trap feature.

8. The apparatus for identifying trap characteristics according to claim 7, characterized in that, The model establishment module includes: Fusing and comparing the geological knowledge in the trap knowledge graph with the vectorized geology and its exploration data to obtain training corpus. Training the constructed unsupervised trap feature identification model by using the training corpus to obtain the trained unsupervised trap feature identification model. Using the annotation information in the training corpus as labels, and training the constructed supervised trap feature identification model by using the training corpus to obtain the trained supervised trap feature identification model.

9. An electronic device, characterized in that, It includes a processor and a memory, and the memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the method according to any one of claims 1-4 is run.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the method according to any one of claims 1-4 is run.

Images