An AI-based secondary analysis method and system for geological exploration data

Abstract

A secondary analysis method and system for geological exploration data based on artificial intelligence provided by this application clusters the integrated features corresponding to the object with each geological exploration description feature in the geological exploration description feature queue to obtain each clustering result of the target to be processed, and then uses each clustering result of the target to be processed for semantic recognition of thematic geological exploration data, so that the semantic relevance between the geological exploration data semantics and the object's geological exploration data semantics can be extracted in the semantic recognition of thematic geological exploration data, improving the accuracy of the obtained geological exploration data semantic types. Then, according to the geological exploration data semantic types, the geological exploration data semantic analysis result corresponding to the object is obtained, that is, the accuracy of the obtained geological exploration data semantics is improved, thereby improving the analysis accuracy of geological exploration data.

Description

Technical Field

[0001] The present application relates to the technical field of secondary data analysis, and more specifically, to a method and system for secondary analysis of geological survey data based on artificial intelligence. Background Art

[0002] "Geological exploration" is the investigation and research activities of exploring and detecting geology through various means and methods, determining the appropriate bearing layer, determining the foundation type according to the bearing capacity of the bearing layer, and calculating the foundation parameters. It is to investigate and study the geological conditions of rocks, strata, structures, minerals, hydrology, landforms, etc. in a certain area in order to find out the quality and quantity of minerals and the technical conditions for mining and utilization, and provide the mineral reserves and geological data required for mine construction design.

[0003] At present, there may be omissions in exploration during the survey, so it is necessary to conduct a secondary analysis of the exploration report to ensure whether the mineral resources are fully discovered. However, how to conduct the accuracy of the analysis is a problem that is difficult to solve at present. Summary of the invention

[0004] In order to improve the technical problems existing in the relevant technologies, the present application provides a secondary analysis method and system for geological survey data based on artificial intelligence.

[0005] In a first aspect, a method for secondary analysis of geological survey data based on artificial intelligence is provided, including:

[0006] Acquire the secondary geological survey data to be analyzed, perform feature extraction on the secondary geological survey data to be analyzed, and obtain a geological survey description feature queue;

[0007] Combining the geological survey description feature queue to perform the start subject description content recognition to obtain the geological survey description feature of the start object, and combining the geological survey description feature queue to perform the final subject description content recognition to obtain the geological survey description feature of the final object;

[0008] Integrate the geological survey description features of each object to obtain the integrated features corresponding to the object;

[0009] Clustering each geological exploration description feature in the geological exploration description feature queue with the integrated feature to obtain a clustering result of each target to be processed corresponding to the object;

[0010] Perform thematic geological exploration data semantic recognition on each of the target clustering results to be processed to obtain the geological exploration data semantic type corresponding to each of the target clustering results to be processed, and obtain the geological exploration data semantic analysis result corresponding to the object in combination with the geological exploration data semantic type corresponding to each of the target clustering results to be processed.

[0011] In this application, integrating the geological exploration description features of each object to obtain the integrated feature corresponding to the object includes:

[0012] Performing weighted sum calculation on the geological exploration description features of the start object and the geological exploration description features of the end object to obtain the target integrated feature corresponding to the object.

[0013] In this application, performing weighted sum calculation on the geological exploration description features of the start object and the geological exploration description features of the end object to obtain the target integrated feature corresponding to the object includes:

[0014] Calculating the start weighted vector corresponding to the geological exploration description features of the start object based on a preset start confidence level, and calculating the end weighted vector corresponding to the geological exploration description features of the end object based on a preset end confidence level;

[0015] Calculating the vector sum of the start weighted vector and the end weighted vector to obtain the target integrated feature corresponding to the object.

[0016] In this application, there are no less than two geological exploration description features of the start object, and no less than two geological exploration description features of the end object; determining the object from the secondary geological exploration data to be analyzed by combining the geological exploration description features of the start object and the geological exploration description features of the end object includes:

[0017] Determining the geological exploration description features of the real-time start object from the geological exploration description features of each start object;

[0018] Obtaining the real-time start queue position of the geological exploration description features of the real-time start object in the geological exploration description feature queue, and obtaining each end queue position of the geological exploration description features of each end object in the geological exploration description feature queue;

[0019] Determining the geological exploration description features of the real-time end object corresponding to the geological exploration description features of the real-time start object by combining the position relationship between the real-time start queue position and each end queue position;

[0020] Determining the real-time object from the secondary geological exploration data to be analyzed according to the geological exploration description features of the real-time start object and the geological exploration description features of the real-time end object.

[0021] In this application, determining the geological exploration description features of the real-time end object corresponding to the geological exploration description features of the real-time start object by combining the position relationship between the real-time start queue position and each end queue position includes:

[0022] Select each of the final queue positions in combination with the real-time start queue position to obtain each target final queue position;

[0023] Calculate the position differences between the real-time start queue position and each of the target final queue positions, and determine the minimum position difference from each position difference;

[0024] Use the geological exploration description features of the final object corresponding to the minimum position difference as the geological exploration description features of the real-time final object.

[0025] In this application, the integration based on the geological exploration description features of each object to obtain the integration features corresponding to the object includes:

[0026] Obtain the preset confidence levels and the total number of vectors corresponding to the geological exploration description features of each object;

[0027] Perform weighted average calculation on the geological exploration description features of each object in combination with the preset confidence levels and the total number of vectors corresponding to the geological exploration description features of each object to obtain the integration features corresponding to the object.

[0028] In this application, the method further includes:

[0029] Input the secondary geological exploration data to be analyzed into the geological exploration data topic extraction thread for topic extraction to obtain the output object and the geological exploration data semantic analysis result corresponding to the object;

[0030] The geological exploration data topic extraction thread is pre-trained using an artificial intelligence thread based on training examples, and the training examples include training geological exploration data, the object directory corresponding to the training geological exploration data, and the geological exploration data semantic directory corresponding to the training geological exploration data.

[0031] In this application, the geological exploration data topic extraction thread includes an object extraction thread and a geological exploration data semantic extraction thread; the input of the secondary geological exploration data to be analyzed into the geological exploration data topic extraction thread for topic extraction to obtain the output object and the geological exploration data semantic analysis result corresponding to the object includes:

[0032] Perform feature extraction processing on the secondary geological exploration data to be analyzed to obtain the geological exploration description feature queue;

[0033] Input the geological exploration description feature queue into the object extraction thread for topic description content recognition to obtain the geological exploration description features of each object, and determine the object in combination with the geological exploration description features of each object;

[0034] Integrate the geological exploration description features of each of the objects to obtain the integrated features corresponding to the objects, and cluster each geological exploration description feature in the geological exploration description feature queue with the integrated features respectively to obtain each to-be-processed target clustering result corresponding to the object;

[0035] Input each of the to-be-processed target clustering results into the geological exploration data semantic extraction thread for geological exploration data semantic recognition, obtain the geological exploration data semantic categories corresponding to each of the to-be-processed target clustering results, and combine the geological exploration data semantic categories corresponding to each of the to-be-processed target clustering results to obtain the geological exploration data semantic analysis result corresponding to the object.

[0036] In this application, the object extraction thread includes a start object extraction thread and a final object extraction thread; the input of the geological exploration description feature queue into the object extraction thread for topic description content recognition to obtain the geological exploration description features of each object, and the determination of the object by combining the geological exploration description features of each object includes:

[0037] Input the geological exploration description feature queue into the start object extraction thread for start description content recognition to obtain the geological exploration description features of the start object, and input the geological exploration description feature queue into the final object extraction thread for final description content recognition to obtain the geological exploration description features of the final object;

[0038] Determine the object from the to-be-analyzed secondary geological exploration data by combining the geological exploration description features of the start object and the geological exploration description features of the final object.

[0039] In this application, the training of the geological exploration data topic extraction thread includes the following steps:

[0040] Perform feature extraction processing on the training geological exploration data in the training examples to obtain a training geological exploration description feature queue;

[0041] Input the training geological exploration description feature queue into the original object extraction thread for topic description content recognition to obtain the geological exploration description features of each training object, and determine the original object by combining the geological exploration description features of each training object;

[0042] Integrate the geological exploration description features of each training object to obtain the original integrated features corresponding to the original object, and cluster each training geological exploration description feature in the training geological exploration description feature queue with the original integrated features respectively to obtain each original to-be-processed target clustering result corresponding to the original object;

[0043] Input each of the original target clustering results to be processed into the original geological exploration data semantic extraction thread for geological exploration data semantic recognition, obtain the original geological exploration data semantic types corresponding to each of the original target clustering results to be processed, and combine the original geological exploration data semantic types corresponding to each of the original target clustering results to obtain the original geological exploration data semantics corresponding to the original object;

[0044] Calculate object evaluation index information by combining the original object and the object directory corresponding to the training geological exploration data using the object evaluation index algorithm, and calculate geological exploration data semantic evaluation index information by combining the original geological exploration data semantics and the geological exploration data semantic directory corresponding to the training geological exploration data using the geological exploration data semantic evaluation index algorithm;

[0045] Determine thread evaluation index information by combining the object evaluation index information and the geological exploration data semantic evaluation index information, and optimize the original object extraction thread and the original geological exploration data semantic extraction thread by combining the thread evaluation index information. When the training is completed, obtain the geological exploration data topic extraction thread.

[0046] In this application, the feature extraction process of the training geological exploration data in the training examples to obtain a training geological exploration description feature queue includes:

[0047] Input the training geological exploration data in the training examples into the geological exploration data feature distribution thread for feature distribution to obtain the training geological exploration description feature queue, and the geological exploration data feature distribution thread is pre-trained using an artificial intelligence thread based on feature distribution training examples.

[0048] In this application, the original object extraction thread includes an original start object extraction thread and an original final object extraction thread; the process of inputting the training geological exploration description feature queue into the original object extraction thread for topic description content recognition to obtain the geological exploration description features of each training object, and determining the original object by combining the geological exploration description features of each training object includes:

[0049] Input the training geological exploration description feature queue into the original start object extraction thread for start description content recognition to obtain the geological exploration description features of the original start object;

[0050] Input the training geological exploration description feature queue into the original final object extraction thread for final description content recognition to obtain the geological exploration description features of the original final object;

[0051] Determine the original object from the training geological exploration data by combining the geological exploration description features of the original start object and the geological exploration description features of the original final object;

[0052] Calculating object evaluation index information by using an object evaluation index algorithm in combination with the object directory corresponding to the training geological exploration data and the original object, including:

[0053] Calculating start object evaluation index information by using a start object evaluation index algorithm in combination with the geological exploration description features of the original start object and the start object directory in the object directory;

[0054] Calculating final object evaluation index information by using a final object evaluation index algorithm in combination with the geological exploration description features of the original final object and the final object directory in the object directory, and obtaining the object evaluation index information by combining the start object evaluation index information and the final object evaluation index information.

[0055] In this application, determining the object from the secondary geological exploration data to be analyzed by combining the geological exploration description features of the start object and the geological exploration description features of the final object includes: obtaining an object start description attribute corresponding to the geological exploration description features of the start object, an object final description attribute corresponding to the geological exploration description features of the final object, and a description attribute between the object start description attribute and the object final description attribute from the secondary geological exploration data to be analyzed, so as to obtain the object.

[0056] In a second aspect, a secondary analysis system for geological exploration data based on artificial intelligence is provided, including a processor and a memory that communicate with each other, and the processor is configured to read and execute a computer program from the memory to implement the above method.

[0057] A method and system for secondary analysis of geological exploration data based on artificial intelligence provided by an embodiment of the present application obtain secondary geological exploration data to be analyzed, perform feature extraction processing on the secondary geological exploration data to be analyzed to obtain a geological exploration description feature queue. Based on the geological exploration description feature queue, perform topic description content recognition to obtain the geological exploration description features of each object, and determine the object based on the geological exploration description features of each object. Integrate the geological exploration description features of each object to obtain the integrated features corresponding to the object. Cluster each geological exploration description feature in the geological exploration description feature queue with the integrated features respectively to obtain each to-be-processed target clustering result corresponding to the object. Perform topic geological exploration data semantic recognition on each to-be-processed target clustering result to obtain the geological exploration data semantic types corresponding to each to-be-processed target clustering result, and obtain the geological exploration data semantic analysis result corresponding to the object based on the geological exploration data semantic types corresponding to each to-be-processed target clustering result. Since the integrated features corresponding to the object are clustered with each geological exploration description feature in the geological exploration description feature queue to obtain each to-be-processed target clustering result, and then each to-be-processed target clustering result is used for topic geological exploration data semantic recognition, the correlation between the geological exploration data semantics and the geological exploration data semantics of the object can be extracted in the topic geological exploration data semantic recognition, the accuracy of the obtained geological exploration data semantic types is improved, and then the geological exploration data semantic analysis result corresponding to the object is obtained based on the geological exploration data semantic types, that is, the accuracy of the obtained geological exploration data semantics is improved, thereby improving the analysis accuracy of the geological exploration data. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present application and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0059] Figure 1 It is a flowchart of a method for secondary analysis of geological exploration data based on artificial intelligence provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] In order to better understand the above technical solutions, the technical solutions of the present application will be described in detail below through the drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solutions of the present application, rather than limitations on the technical solutions of the present application. Without conflict, the technical features in the embodiments of the present application and the embodiments can be combined with each other.

[0061] Please refer to Figure 1, which shows an artificial intelligence-based secondary analysis method for geological exploration data. This method may include the technical solutions described in the following steps 202-step 210.

[0062] Step 202, obtain the secondary geological exploration data to be analyzed, perform feature extraction processing on the secondary geological exploration data to be analyzed, and obtain a geological exploration description feature queue.

[0063] Exemplarily, the secondary geological exploration data is obtained by integrating the data after the first exploration and conducting the second exploration to determine whether there is any missing ore information underground. For example: during the first exploration, generally the ore sources with ore reserves reaching a preset value are explored, and resources with low ore content are not explored and analyzed. Another example: when the mining of ore resources is completed, since the equipment has not been transferred yet, at this time, the exploration data can also be analyzed secondarily to determine whether there are small-scale mineral resources nearby, which can just continue to use the equipment and by the way, mine the small-scale mineral resources to reduce the equipment usage cost, etc.

[0064] Specifically, obtain the secondary geological exploration data to be analyzed. Among them, the secondary geological exploration data to be analyzed can be obtained from the terminal, or from the database, or collected from the Internet, or obtained from the provided geological exploration data set. Then, use the feature distribution algorithm to perform feature extraction processing on the secondary geological exploration data to be analyzed to obtain a geological exploration description feature queue.

[0065] Step 204, combine the geological exploration description feature queue to start the recognition of the theme description content, obtain the geological exploration description features of the starting object, and combine the geological exploration description feature queue to perform the recognition of the final theme description content, and obtain the geological exploration description features of the final object.

[0066] Exemplarily, the recognition of the theme description content is the different exploration report titles corresponding to the theme in this application. Recognizing this title can identify the content in the report.

[0067] Specifically, use the geological exploration description feature queue for theme extraction. First, perform the recognition of the theme description content on the geological exploration description feature queue, that is, identify whether each geological exploration description feature in the geological exploration description feature queue is the geological exploration description feature corresponding to the theme object, and obtain the geological exploration description features of each object. Then, determine the object vector according to the geological exploration description features of each object, and determine the corresponding object from the secondary geological exploration data to be analyzed according to the object vector. There can be multiple such objects. Among them, the object can be determined according to the context-free information in the geological exploration description features of each object. It is also possible to determine the object according to the positional relationship of the geological exploration description features of each object in the queue. For example, the geological exploration description features of objects with adjacent positions can be used as the geological exploration description features of the same object.

[0068] Step 206: Integrate based on the geological exploration description features of each object to obtain the integrated features corresponding to the object.

[0069] Exemplarily, the object can be understood as a geographical area, and each area has separate geological exploration data.

[0070] Among them, integration refers to performing vector operations on the geological exploration description features of each object to obtain the integrated features after integration. The integrated features refer to the vectors used to represent the semantic analysis results of the geological exploration data corresponding to the object. Different objects have different integrated features, and each object has corresponding integrated features.

[0071] Specifically, integrate the geological exploration description features of each object corresponding to the object to obtain the integrated features corresponding to the object. Among them, integrating the geological exploration description features of each object may refer to performing weighted sum calculation on the geological exploration description features of each object, or performing weighted average calculation on the geological exploration description features of the object, or performing vector dot product calculation on the geological exploration description features of each object, or performing vector product calculation on the geological exploration description features of each object, etc.

[0072] In a possible embodiment, when there are multiple objects, respectively integrate the geological exploration description features of each object corresponding to each object to obtain the integrated features corresponding to each object.

[0073] Step 208: Cluster each geological exploration description feature in the geological exploration description feature queue with the integrated features respectively to obtain each to-be-processed target clustering result corresponding to the object.

[0074] Among them, the to-be-processed target clustering result refers to the vector obtained by clustering the geological exploration description features and the integrated features, and is used to identify the semantic of the thematic geological exploration data.

[0075] Specifically, cluster each geological exploration description feature in the geological exploration description feature queue with the integrated features respectively to obtain each to-be-processed target clustering result corresponding to the object. Among them, clustering can be performed in the order that the geological exploration description feature is in the front and the integrated feature is in the back, or in the order that the geological exploration description feature is in the back and the integrated feature is in the front.

[0076] In a possible embodiment, when there are integrated features corresponding to multiple objects, respectively cluster the integrated features corresponding to each object with each geological exploration description feature in the geological exploration description feature queue to obtain each to-be-processed target clustering result corresponding to each object.

[0077] Step 210: Perform semantic recognition of thematic geological exploration data on each target clustering result to be processed, obtain the semantic types of geological exploration data corresponding to each target clustering result to be processed, and obtain the semantic analysis result of the geological exploration data corresponding to the object based on the semantic types of geological exploration data corresponding to each target clustering result to be processed.

[0078] Exemplarily, the semantics appearing in this application can be understood as the key features in geological exploration data. For example, the actual report of minerals appearing in geological exploration data.

[0079] Specifically, perform semantic recognition of thematic geological exploration data on each target clustering result to be processed respectively, and obtain the semantic types of geological exploration data corresponding to each target clustering result to be processed. Among them, artificial intelligence threads can be used to perform semantic recognition of thematic geological exploration data on each target clustering result to be processed, or algorithms based on pattern matching can be used to perform semantic recognition of thematic geological exploration data on each target clustering result to be processed. Then, obtain the semantic analysis result of the geological exploration data corresponding to the object according to the semantic types of geological exploration data corresponding to each target clustering result to be processed from the secondary geological exploration data to be analyzed. Among them, the target clustering results to be processed corresponding to the same semantic type of geological exploration data with adjacent positions are used as the target clustering results to be processed with the same semantic of geological exploration data, and then the geological exploration data corresponding to the target clustering results to be processed with the same semantic of geological exploration data in the secondary geological exploration data to be analyzed is extracted to obtain the semantic analysis result of the geological exploration data corresponding to the object.

[0080] The above artificial intelligence-based secondary analysis method for geological exploration data obtains the secondary geological exploration data to be analyzed, extracts and processes the features of the secondary geological exploration data to be analyzed to obtain a geological exploration description feature queue. Based on the geological exploration description feature queue, the theme description content is identified to obtain the geological exploration description features of each object, and the object is determined based on the geological exploration description features of each object. Based on the geological exploration description features of each object, integration is performed to obtain the integrated features corresponding to the object. Each geological exploration description feature in the geological exploration description feature queue is clustered with the integrated features to obtain each target clustering result to be processed corresponding to the object. The theme geological exploration data semantic recognition is performed on each target clustering result to be processed to obtain the geological exploration data semantic types corresponding to each target clustering result to be processed, and the geological exploration data semantic analysis result corresponding to the object is obtained based on the geological exploration data semantic types corresponding to each target clustering result to be processed. Since the integrated features corresponding to the object are clustered with each geological exploration description feature in the geological exploration description feature queue to obtain each target clustering result to be processed, and then each target clustering result to be processed is used for theme geological exploration data semantic recognition, the correlation between the geological exploration data semantics and the geological exploration data semantics of the object can be extracted in the theme geological exploration data semantic recognition, the accuracy of the obtained geological exploration data semantic types is improved, and then the geological exploration data semantic analysis result corresponding to the object is obtained based on the geological exploration data semantic types, that is, the accuracy of the obtained geological exploration data semantics is improved, thereby improving the analysis accuracy of the geological exploration data

[0081] In a possible embodiment, step 204, that is, based on the geological exploration description feature queue, the theme description content is identified to obtain the geological exploration description features of each object, and the object is determined based on the geological exploration description features of each object, includes:

[0082] Step 302, based on the geological exploration description feature queue, start the theme description content identification to obtain the geological exploration description features of the start object, and based on the geological exploration description feature queue, perform the final theme description content identification to obtain the geological exploration description features of the final object.

[0083] Among them, the start theme description content identification refers to the identification of the start description attribute in the object. The geological exploration description features of the start object refer to the vector corresponding to the start description attribute in the object. The final theme description content identification refers to the identification of the final description attribute in the object, and the geological exploration description features of the final object refer to the vector corresponding to the final description attribute in the object.

[0084] Specifically, first, start topic description content recognition is performed on each geological exploration description feature in the geological exploration description feature queue, that is, it is determined whether each geological exploration description feature is a vector corresponding to the start description attribute in the object. When a geological exploration description feature corresponding to the start description attribute in the object is recognized, the corresponding geological exploration description feature is used as the geological exploration description feature of the start object. Among them, multiple geological exploration description features of the start object can be recognized. Then, final topic description content recognition is performed on each geological exploration description feature in the geological exploration description feature queue, that is, it is determined whether each geological exploration description feature is a vector corresponding to the final description attribute in the object. When a geological exploration description feature corresponding to the final description attribute in the object is recognized in each geological exploration description feature, the corresponding geological exploration description feature is used as the geological exploration description feature of the final object. Among them, multiple geological exploration description features of the final object can also be recognized.

[0085] Step 302: Determine the object from the secondary geological exploration data to be analyzed based on the geological exploration description features of the start object and the geological exploration description features of the final object.

[0086] Specifically, the corresponding geological exploration description features of the final object are obtained according to the geological exploration description features of the start object. Then, the object start description attribute corresponding to the geological exploration description features of the start object, the object final description attribute corresponding to the geological exploration description features of the final object, and the description attribute between the object start description attribute and the object final description attribute are obtained from the secondary geological exploration data to be analyzed, so as to obtain the object.

[0087] Step 206, that is, based on the geological exploration description features of each object, integration is performed to obtain the integrated features corresponding to the object, including:

[0088] Step 304: Perform weighted sum calculation on the geological exploration description features of the start object and the geological exploration description features of the final object to obtain the target integrated features corresponding to the object.

[0089] Specifically, the target integrated features refer to the integrated features obtained after performing weighted sum on the geological exploration description features. The confidence levels corresponding to the geological exploration description features of the start object and the confidence levels corresponding to the geological exploration description features of the final object that are set in advance are obtained. Then, the geological exploration description features of the start object and the geological exploration description features of the final object are weighted to obtain the weighted geological exploration description features of the start object and the weighted geological exploration description features of the final object. Then, the vector sum of the weighted geological exploration description features of the start object and the weighted geological exploration description features of the final object is calculated, and this vector sum is used as the target integrated features corresponding to the object.

[0090] In a possible embodiment, step 306: Perform weighted sum calculation on the geological exploration description features of the start object and the geological exploration description features of the final object to obtain the target integrated features corresponding to the object, including:

[0091] Calculate the start weighted vector corresponding to the geological exploration description features of the start object based on a preset start confidence level, and calculate the end weighted vector corresponding to the geological exploration description features of the end object based on a preset end confidence level. Calculate the vector sum of the start weighted vector and the end weighted vector to obtain the target integrated feature corresponding to the object.

[0092] Among them, the start weighted vector refers to the vector obtained by weighting the geological exploration description features of the start object. The preset start confidence level refers to the start confidence level corresponding to the geological exploration description features of the start object set in advance. Different geological exploration description features of the start object can be set with different start confidence levels, or the same start confidence level can be set. The end weighted vector refers to the vector obtained by weighting the geological exploration description features of the end object. The preset end confidence level refers to the end confidence level corresponding to the geological exploration description features of the end object set in advance. Different geological exploration description features of the end object can be set with different end confidence levels, or the same end confidence level can be set.

[0093] Specifically, calculate the start weighted vector corresponding to the geological exploration description features of the start object based on a preset start confidence level, and calculate the end weighted vector corresponding to the geological exploration description features of the end object based on a preset end confidence level. Calculate the vector sum of the start weighted vector and the end weighted vector to obtain the target integrated feature corresponding to the object. By calculating the weighted sum to integrate the vectors, the accuracy of the obtained integrated feature is improved, and thus the semantics of the identified geological exploration data is made more accurate.

[0094] In a possible embodiment, the geological exploration description features of the start object include no less than two, and the geological exploration description features of the end object include no less than two;

[0095] The step of determining the object from the secondary geological exploration data to be analyzed based on the geological exploration description features of the start object and the geological exploration description features of the end object includes:

[0096] Step 402, determine the geological exploration description features of the real-time start object from the geological exploration description features of each start object.

[0097] Step 404, obtain the real-time start queue position of the geological exploration description features of the real-time start object in the geological exploration description feature queue, and obtain each end queue position of the geological exploration description features of each end object in the geological exploration description feature queue.

[0098] Among them, the geological exploration description features of the real-time start object refer to the geological exploration description features of the start object for which the corresponding geological exploration description features of the end object are to be determined. The real-time start queue position refers to the position of the real-time start geological exploration description features in the geological exploration description feature queue, which can be a label, etc. The end queue position refers to the position of the geological exploration description features of the end object in the geological exploration description feature queue.

[0099] Specifically, when the geological exploration description features of multiple start objects and the geological exploration description features of multiple end objects are recognized, it is necessary to determine the correspondence between the geological exploration description features of the start objects and the geological exploration description features of the end objects. At this time, select the geological exploration description features of the start objects from the geological exploration description features of each start object as the geological exploration description features of the real-time start objects. Among them, it can be randomly selected without replacement, or it can be sequentially selected according to the position of the geological exploration description features of the start objects in the geological exploration description feature queue. Then, find the real-time start queue position corresponding to the geological exploration description features of the real-time start objects from the geological exploration description feature queue. And find the final queue position corresponding to the geological exploration description features of each end object from the geological exploration description feature queue.

[0100] Step 406: Determine the geological exploration description features of the real-time end objects corresponding to the geological exploration description features of the real-time start objects based on the positional relationship between the real-time start queue position and each final queue position.

[0101] Among them, the geological exploration description features of the real-time end objects refer to the geological exploration description features of the end objects corresponding to the geological exploration description features of the real-time start objects.

[0102] Specifically, find the nearest final queue position from each final queue position according to the real-time start queue position, and then use the geological exploration description features of the end object corresponding to the nearest final queue position as the geological exploration description features of the real-time end objects corresponding to the geological exploration description features of the real-time start objects.

[0103] Step 408: Determine the real-time objects from the secondary geological exploration data to be analyzed according to the geological exploration description features of the real-time start objects and the geological exploration description features of the real-time end objects.

[0104] Specifically, the real-time objects refer to the objects that need to be determined in real time. Determine the real-time start object description attributes corresponding to the geological exploration description features of the real-time start objects and the real-time end object description attributes corresponding to the geological exploration description features of the real-time end objects from the secondary geological exploration data to be analyzed. Then, use the real-time start object description attributes as the starting point and the real-time end object description attributes as the ending point to extract the real-time objects from the secondary geological exploration data to be analyzed. In a possible embodiment, determine the geological exploration description features of no less than two end objects corresponding to the geological exploration description features of no less than two start objects, and then obtain no less than two objects from the secondary geological exploration data to be analyzed.

[0105] In the above embodiments, by determining the geological exploration description features of the real-time final object corresponding to the geological exploration description features of the real-time start object based on the positional relationship between the real-time start queue position and each final queue position, and then determining the real-time object from the secondary geological exploration data to be analyzed according to the geological exploration description features of the real-time start object and the geological exploration description features of the real-time final object, the real-time object can be obtained more quickly, and the efficiency of obtaining the real-time object is improved.

[0106] In a possible embodiment, step 406, determining the geological exploration description features of the real-time final object corresponding to the geological exploration description features of the real-time start object based on the positional relationship between the real-time start queue position and each final queue position, includes:

[0107] Step 502, selecting each final queue position based on the real-time start queue position to obtain each target final queue position.

[0108] Specifically, according to the real-time start queue position, the final queue positions before the real-time start queue position in each final queue position are selected and removed to obtain each target final queue position, that is, each target final queue position is after the real-time start queue position.

[0109] Step 504, calculating the position differences between the real-time start queue position and each target final queue position respectively, and determining the minimum position difference from each position difference.

[0110] Specifically, calculate the position differences between the real-time start queue position and each target final queue position respectively. Then compare the magnitudes of each position difference, and determine the minimum position difference from each position difference.

[0111] Step 506, taking the geological exploration description features of the final object corresponding to the minimum position difference as the geological exploration description features of the real-time final object.

[0112] Specifically, directly take the geological exploration description features of the final object corresponding to the minimum position difference as the geological exploration description features of the real-time final object.

[0113] In the above embodiments, by calculating the position difference, and then taking the geological exploration description features of the final object corresponding to the minimum position difference as the geological exploration description features of the real-time final object, the geological exploration description features of the real-time final object can be obtained quickly, improving the efficiency.

[0114] In a possible embodiment, step 206, that is, integrating based on the geological exploration description features of each object to obtain the integrated features corresponding to the object, includes the steps:

[0115] Obtain the preset confidence levels corresponding to the geological exploration description features of each object and the total number of vectors; based on the preset confidence levels corresponding to the geological exploration description features of each object and the total number of vectors, perform weighted average calculation on the geological exploration description features of each object to obtain the integrated features corresponding to the object.

[0116] Among them, the preset confidence level refers to the confidence level corresponding to the geological exploration description features of the object set in advance. The total number of vectors refers to the total number of geological exploration description features of each object.

[0117] Specifically, obtain the preset confidence levels corresponding to the geological exploration description features of each object and count the total number of vectors corresponding to the geological exploration description features of each object. Then perform weighted calculation on the geological exploration description features of each object according to the corresponding preset confidence levels to obtain the weighted geological exploration description features of each object, and then calculate the ratio of the weighted geological exploration description features of each object to the total number of vectors to obtain the integrated features corresponding to the object.

[0118] In a possible embodiment, vector product calculation can be performed on the weighted geological exploration description features of each object, and the obtained calculation result is used as the integrated features corresponding to the object. In a possible embodiment, vector sum calculation can be performed on the weighted geological exploration description features of each object, and the obtained calculation result is used as the integrated features corresponding to the object. In a possible embodiment, the weighted geological exploration description features of each object can be directly clustered, and the vectors after clustering are used as the integrated features corresponding to the object.

[0119] In the above embodiment, by performing weighted average calculation on the geological exploration description features of each object, the integrated features corresponding to the object are obtained, improving the accuracy of the obtained integrated features.

[0120] In a possible embodiment, the secondary analysis method of geological exploration data based on artificial intelligence further includes the steps of:

[0121] Input the secondary geological exploration data to be analyzed into the geological exploration data theme extraction thread for theme extraction to obtain the output object and the corresponding geological exploration data semantic analysis result; the geological exploration data theme extraction thread is obtained by training in advance using an artificial intelligence thread based on training examples, and the training examples include training geological exploration data, the object directory corresponding to the training geological exploration data, and the geological exploration data semantic directory corresponding to the training geological exploration data.

[0122] Among them, the geological exploration data theme extraction thread is used to extract the objects and the semantics of the geological exploration data from the input geological exploration data, and obtain structured data. The object directory refers to the directory of whether each description attribute in the training geological exploration data is an object, including the non-object description attribute directory and the object description attribute directory. The geological exploration data semantics directory refers to the directory of the types of geological exploration data semantics corresponding to each description attribute in the training geological exploration data, and this directory includes other directories and the directories corresponding to the types of geological exploration data semantics.

[0123] The geological exploration data theme extraction thread performs feature extraction processing on the secondary geological exploration data to be analyzed, and obtains a geological exploration description feature queue. The geological exploration data theme extraction thread performs theme description content recognition based on the geological exploration description feature queue, obtains the geological exploration description features of each object, and determines the object based on the geological exploration description features of each object. The geological exploration data theme extraction thread integrates based on the geological exploration description features of each object to obtain the integrated features corresponding to the object. The geological exploration data theme extraction thread clusters each geological exploration description feature in the geological exploration description feature queue with the integrated features respectively, and obtains each to-be-processed target clustering result corresponding to the object. The geological exploration data theme extraction thread performs theme geological exploration data semantics recognition on each to-be-processed target clustering result, obtains the types of geological exploration data semantics corresponding to each to-be-processed target clustering result, and obtains the geological exploration data semantics analysis result corresponding to the object based on the types of geological exploration data semantics corresponding to each to-be-processed target clustering result.

[0124] In the above embodiment, by inputting the secondary geological exploration data to be analyzed into the geological exploration data theme extraction thread for theme extraction, the output objects and the geological exploration data semantics analysis results corresponding to the objects are directly obtained. Since the geological exploration data theme extraction thread is a theme extraction thread from the input end to the output end, it can better learn the geological exploration data semantics correlation in the geological exploration data, thereby effectively improving the thread performance of the geological exploration data theme extraction thread. Furthermore, using the geological exploration data theme extraction thread for theme extraction improves the accuracy and efficiency of theme extraction.

[0125] In a possible embodiment, the geological exploration data theme extraction thread includes an object extraction thread and a geological exploration data semantics extraction thread;

[0126] Inputting the secondary geological exploration data to be analyzed into the geological exploration data theme extraction thread for theme extraction, and obtaining the output objects and the geological exploration data semantics analysis results corresponding to the objects, includes:

[0127] Step 602, perform feature extraction processing on the secondary geological exploration data to be analyzed, and obtain a geological exploration description feature queue.

[0128] Specifically, the geological exploration data theme extraction thread inputs the geological exploration description feature queue into the object extraction thread for theme description content recognition, that is, the object extraction thread outputs the geological exploration description feature with the highest object probability to obtain the geological exploration description feature of each object. Then, the object is determined according to the geological exploration description feature of each object.

[0129] Step 606: Integrate based on the geological exploration description feature of each object to obtain the integrated feature corresponding to the object. Cluster each geological exploration description feature in the geological exploration description feature queue with the integrated feature respectively to obtain each to-be-processed target clustering result corresponding to the object.

[0130] Specifically, the geological exploration data theme extraction thread performs weighted average calculation on the geological exploration description feature of each object to obtain the integrated feature corresponding to the object. Then, the integrated feature corresponding to the object clusters each geological exploration description feature in the geological exploration description feature queue with the integrated feature respectively to obtain each to-be-processed target clustering result corresponding to the object.

[0131] Step 608: Input each to-be-processed target clustering result into the geological exploration data semantic extraction thread for geological exploration data semantic recognition to obtain the geological exploration data semantic type corresponding to each to-be-processed target clustering result, and obtain the geological exploration data semantic analysis result corresponding to the object based on the geological exploration data semantic type corresponding to each to-be-processed target clustering result.

[0132] Among them, the geological exploration data semantic extraction thread is used to extract the geological exploration data semantics in the secondary geological exploration data to be analyzed. The geological exploration data semantic extraction thread is a multi-classification recognition thread and a queue annotation thread. Different application scenarios can set different numbers of geological exploration data semantic types.

[0133] Specifically, the geological exploration data theme extraction thread inputs each to-be-processed target clustering result into the geological exploration data semantic extraction thread for geological exploration data semantic recognition to obtain the geological exploration data semantic type corresponding to each to-be-processed target clustering result, and obtain the geological exploration data semantic analysis result corresponding to the object based on the geological exploration data semantic type corresponding to each to-be-processed target clustering result.

[0134] In the above embodiments, by performing theme description content recognition in the object extraction thread and geological exploration data semantic recognition in the geological exploration data semantic extraction thread, the accuracy of the obtained object and geological exploration data semantics is improved.

[0135] In a possible embodiment, the object extraction thread includes a start object extraction thread and a final object extraction thread;

[0136] Step 604, input the geological exploration description feature queue into the object extraction thread for topic description content recognition to obtain the geological exploration description features of each object, and determine the object based on the geological exploration description features of each object, including:

[0137] Step 702, input the geological exploration description feature queue into the start object extraction thread for start description content recognition to obtain the geological exploration description features of the start object, and input the geological exploration description feature queue into the final object extraction thread for final description content recognition to obtain the geological exploration description features of the final object.

[0138] Step 704, determine the object from the secondary geological exploration data to be analyzed based on the geological exploration description features of the start object and the geological exploration description features of the final object.

[0139] Among them, the start object extraction thread is used to identify the description attribute corresponding to the start position in the object, that is, to identify whether the description attribute corresponding to this position is the description attribute of the object start position. The final object extraction thread is used to identify the description attribute corresponding to the final position in the object, that is, to identify whether the description attribute corresponding to this position is the description attribute of the object final position.

[0140] Then, obtain the start object description attribute from the secondary geological exploration data to be analyzed according to the geological exploration description features of the start object, and obtain the final object description attribute from the secondary geological exploration data to be analyzed according to the geological exploration description features of the final object. Then, extract the transitional description attribute according to the start object description attribute and the final object description attribute, so as to obtain the object.

[0141] In the above embodiment, the start geological exploration description feature is recognized through the start object extraction thread, and then the final geological exploration description feature is recognized through the final object extraction thread. Finally, the object is determined from the secondary geological exploration data to be analyzed according to the geological exploration description features of the start object and the geological exploration description features of the final object, which improves the accuracy and efficiency of obtaining the object.

[0142] In a possible embodiment, the training of the geological exploration data topic extraction thread includes the following steps:

[0143] Step 802, perform feature extraction processing on the training geological exploration data in the training example to obtain a training geological exploration description feature queue.

[0144] Among them, the training geological exploration description feature queue refers to the geological exploration description feature queue obtained according to the training geological exploration data. The training example includes the training geological exploration data, the corresponding object catalog in the training geological exploration data, and the corresponding geological exploration data semantic catalog in the training geological exploration data. In a possible embodiment, the training example may also include the queue position catalog corresponding to each description attribute in the training geological exploration data.

[0145] Step 804: Input the training geological exploration description feature queue into the original object extraction thread for topic description content recognition to obtain the geological exploration description features of each training object, and determine the original object based on the geological exploration description features of each training object.

[0146] Among them, the original geological exploration data topic extraction thread includes an original object extraction thread and an original geological exploration data semantic extraction thread. The original geological exploration data topic extraction thread refers to the geological exploration data topic extraction thread with thread parameters initialized. The original object extraction thread refers to the object extraction thread with thread parameters initialized. The geological exploration description feature of a training object refers to the geological exploration description feature of the object obtained by performing topic description content recognition using the object extraction thread during training. The original object refers to the object extracted using the original object extraction thread.

[0147] Specifically, initializing the thread parameters in the geological exploration data topic extraction thread obtains the original object extraction thread. Then input the training geological exploration description feature queue into the original object extraction thread for topic description content recognition to obtain the geological exploration description features of each output training object, and then determine the original object according to the geological exploration description features of each training object.

[0148] Step 806: Integrate based on the geological exploration description features of each training object to obtain the original integration feature corresponding to the original object, and cluster each training geological exploration description feature in the training geological exploration description feature queue with the original integration feature respectively to obtain each original to-be-processed target clustering result corresponding to the original object.

[0149] Among them, the original integration feature refers to the vector used to represent the original object. The original to-be-processed target clustering result refers to the vector that needs to perform geological exploration data semantic recognition, and this original to-be-processed target clustering result is obtained by clustering the training geological exploration description feature and the original integration feature.

[0150] Specifically, calculate the weighted average of the geological exploration description features of each training object, and use the weighted average vector as the original integration feature corresponding to the original object. Then cluster each training geological exploration description feature in the training geological exploration description feature queue with the original integration feature to obtain each original to-be-processed target clustering result corresponding to the original object.

[0151] Step 808: Input each original to-be-processed target clustering result into the original geological exploration data semantic extraction thread for geological exploration data semantic recognition to obtain the original geological exploration data semantic types corresponding to each original to-be-processed target clustering result, and obtain the original geological exploration data semantics corresponding to the original object based on the original geological exploration data semantic types corresponding to each original to-be-processed target clustering result.

[0152] Among them, the original geological exploration data semantic extraction thread refers to the geological exploration data semantic extraction thread with initialized thread parameters, that is, this original geological exploration data semantic extraction thread is a thread that has not been trained yet. The original geological exploration data semantic type refers to the geological exploration data semantic type corresponding to the training examples identified by using the original geological exploration data semantic extraction thread. The original geological exploration data semantics refers to the geological exploration data semantic analysis result corresponding to the original object in the training example.

[0153] Specifically, each original target clustering result to be processed is input into the original geological exploration data semantic extraction thread for geological exploration data semantic recognition, and the original geological exploration data semantic type corresponding to each original recognition target clustering result is obtained. Based on each original geological exploration data semantic type, then the adjacent same geological exploration data semantic types in each original geological exploration data semantic type are taken as the same original geological exploration data semantics to determine the corresponding original geological exploration data semantics.

[0154] Step 810 calculates the object evaluation index information based on the object directory corresponding to the original object and the training geological exploration data using the object evaluation index algorithm, and calculates the geological exploration data semantic evaluation index information based on the geological exploration data semantic directory corresponding to the original geological exploration data semantics and the training geological exploration data using the geological exploration data semantic evaluation index algorithm.

[0155] Among them, the object evaluation index algorithm refers to the evaluation index algorithm used to calculate the object evaluation index, and this object evaluation index algorithm can be the binary classification cross-entropy evaluation index algorithm. The geological exploration data semantic evaluation index algorithm refers to the evaluation index algorithm used to calculate the geological exploration data semantic evaluation index, and this geological exploration data semantic evaluation index algorithm can be the multi-classification cross-entropy evaluation index algorithm. The object evaluation index information is used to represent the error information between the trained object and the object directory. The geological exploration data semantic evaluation index information is used to represent the error information between the trained geological exploration data semantics and the geological exploration data semantic directory.

[0156] In step 812, the thread evaluation index information is determined based on the object evaluation index information and the geological exploration data semantic evaluation index information, and the original object extraction thread and the original geological exploration data semantic extraction thread are optimized based on the thread evaluation index information. When the training is completed, the geological exploration data theme extraction thread is obtained.

[0157] Among them, the thread evaluation index information refers to the sum of the object evaluation index information and the geological exploration data semantic evaluation index information, and is used to represent the error information between the result of the training thread and the true result.

[0158] Then, use the thread evaluation metric information to perform backpropagation to optimize the original object extraction thread and the original geological exploration data semantic extraction thread until the trained object extraction thread and the geological exploration data semantic extraction thread are obtained when the thread evaluation metric information is less than the pre-set threshold, that is, obtain the geological exploration data topic extraction thread according to the object extraction thread and the geological exploration data semantic extraction thread. Or when the number of training times reaches the maximum number of iterations, the training is completed to obtain the geological exploration data topic extraction thread. Or when the thread parameters no longer change, the training is completed to obtain the geological exploration data topic extraction thread. When the training is not completed, return to step 802 for loop iteration until the training is completed to obtain the geological exploration data topic extraction thread.

[0159] In a possible embodiment, the original geological exploration data topic extraction thread further includes an original encoding thread, which is an encoding thread for initializing the thread parameters. This encoding thread is used to encode the input geological exploration data to obtain the vector representation of each description attribute in the geological exploration data. Then, input the training geological exploration data into the original encoding thread for encoding to obtain the training geological exploration description feature queue. Then, obtain the vector directory corresponding to the training geological exploration data, calculate the vector error information using the vector directory and the training geological exploration description feature queue, obtain the target thread evaluation metric information based on the vector error information, the object evaluation metric information, and the geological exploration data semantic evaluation metric information. Then, use the target thread evaluation metric information to inversely optimize the original geological exploration data topic extraction thread to obtain the optimized geological exploration data topic extraction thread, that is, obtain the optimized encoding thread, the optimized topic object extraction thread, and the topic geological exploration data semantic extraction thread, and continuously perform loop iteration until the training completion condition is reached to obtain the target geological exploration data topic extraction thread. Thereby, the accuracy of the obtained target geological exploration data topic extraction thread can be improved.

[0160] In the above embodiment, by jointly training the object extraction thread and the geological exploration data semantic extraction thread, the geological exploration data topic extraction thread at the end of training is obtained, improving the accuracy of the obtained geological exploration data topic extraction thread.

[0161] In a possible embodiment, step 802, that is, performing feature extraction processing on the training geological exploration data in the training example to obtain the training geological exploration description feature queue, includes the steps of:

[0162] Input the training geological exploration data in the training example into the geological exploration data feature distribution thread for feature distribution to obtain the training geological exploration description feature queue. The geological exploration data feature distribution thread is obtained by training in advance using an artificial intelligence thread based on the feature distribution training example.

[0163] Specifically, obtain the feature distribution training examples, input the feature distribution training geological exploration data into the original geological exploration data feature distribution thread for feature distribution to obtain the original vector, calculate the vector error information between the original vector and the feature distribution directory corresponding to the feature distribution training geological exploration data, and use this vector error information to optimize the original geological exploration data feature distribution thread until the vector error information is less than the preset threshold or reaches the maximum number of iterations or the thread parameters of the geological exploration data feature distribution thread no longer change. At this time, it indicates that the training completion condition is reached, and the thread obtained from the last training is used as the geological exploration data feature distribution thread. Then, deploy and use the geological exploration data feature distribution thread. At this time, when obtaining the training geological exploration data to be feature-distributed, input the training geological exploration data into the geological exploration data feature distribution thread for feature distribution to obtain the training geological exploration description feature queue.

[0164] In the above embodiment, by using the pre-trained geological exploration data feature distribution thread for feature distribution and then performing topic extraction, the efficiency of topic extraction is improved.

[0165] In a possible embodiment, the original object extraction thread includes an original start object extraction thread and an original end object extraction thread;

[0166] Step 804, input the training geological exploration description feature queue into the original object extraction thread for topic description content recognition to obtain the geological exploration description features of each training object, and determine the original object based on the geological exploration description features of each training object, including:

[0167] Step 902, input the training geological exploration description feature queue into the original start object extraction thread for start description content recognition to obtain the geological exploration description features of the original start object;

[0168] Step 902, input the training geological exploration description feature queue into the original end object extraction thread for end description content recognition to obtain the geological exploration description features of the original end object.

[0169] Among them, the original start object extraction thread refers to the start object extraction thread with the thread parameters initialized in the original geological exploration data topic extraction thread. The original end object extraction thread refers to the end object extraction thread with the thread parameters initialized in the original geological exploration data topic extraction thread. The geological exploration description features of the original start object refer to the geological exploration description features of the start object obtained by using the initialized thread parameters, and the geological exploration description features of the original end object refer to the geological exploration description features of the end object obtained by using the initialized thread parameters.

[0170] Specifically, the original object extraction thread can include an original start object extraction thread and an original end object extraction thread. When the training geological exploration description feature queue is obtained, the training geological exploration description feature queue can be input into the original start object extraction thread for start description content recognition to obtain the geological exploration description features of the original start object, and then the training geological exploration description feature queue can be input into the original end object extraction thread for end description content recognition to obtain the geological exploration description features of the original end object. Among them, both the original start object extraction thread and the original start object extraction thread can be fully connected neural threads.

[0171] Step 904: Determine the original object from the training geological exploration data based on the geological exploration description features of the original start object and the geological exploration description features of the original end object.

[0172] Specifically, determine the corresponding start object description attribute according to the geological exploration description features of the original start object, determine the corresponding end object description attribute according to the geological exploration description features of the original end object, and then determine the original object from the secondary geological exploration data to be analyzed according to the start object description attribute and the end object description attribute.

[0173] Then, in step 910, calculate the object evaluation index information based on the object directory corresponding to the original object and the training geological exploration data, including:

[0174] Step 906: Calculate the start object evaluation index information based on the geological exploration description features of the original start object and the start object directory in the object directory using the start object evaluation index algorithm.

[0175] Among them, the start object directory refers to the directory in the object directory used to label the start position object. The start object evaluation index information represents the error information between the start object directory and the geological exploration description features of the original start object. The start object evaluation index algorithm refers to the evaluation index algorithm for binary classification tasks, which can be the cross-entropy evaluation index algorithm.

[0176] Step 908: Calculate the end object evaluation index information based on the geological exploration description features of the original end object and the end object directory in the object directory using the end object evaluation index algorithm, and obtain the object evaluation index information based on the start object evaluation index information and the end object evaluation index information.

[0177] Among them, the end object directory refers to the directory in the object directory used to label the end position object. The end object evaluation index information represents the error information between the end object directory and the geological exploration description features of the original end object. The end object evaluation index algorithm refers to the evaluation index algorithm for binary classification tasks, which can be the cross-entropy evaluation index algorithm.

[0178] In the above embodiments, by calculating the final object evaluation index information and the initial object evaluation index information, the object evaluation index information is obtained, thereby improving the efficiency of calculating the object evaluation index information.

[0179] Step 210, that is, obtaining the semantic analysis result of the geological exploration data corresponding to the object based on the semantic types of the geological exploration data corresponding to each target clustering result to be processed, includes:

[0180] In a specific embodiment, a method for secondary analysis of geological exploration data based on artificial intelligence is provided, specifically including the following steps:

[0181] Step 1002, obtaining the secondary geological exploration data to be analyzed, inputting the secondary geological exploration data to be analyzed into the geological exploration data theme extraction thread for theme extraction, and the geological exploration data theme extraction thread performs feature extraction processing on the secondary geological exploration data to be analyzed to obtain a geological exploration description feature queue.

[0182] Step 1004, the geological exploration data theme extraction thread inputs the geological exploration description feature queue into the initial object extraction thread for initial description content recognition to obtain the geological exploration description features of the initial object, and inputs the geological exploration description feature queue into the final object extraction thread for final description content recognition to obtain the geological exploration description features of the final object.

[0183] Step 1006, the geological exploration data theme extraction thread determines the geological exploration description features of the real-time initial object from the geological exploration description features of each initial object, obtains the real-time start queue position of the geological exploration description features of the real-time initial object in the geological exploration description feature queue, and obtains each final queue position of the geological exploration description features of each final object in the geological exploration description feature queue.

[0184] Step 1008, the geological exploration data theme extraction thread determines the geological exploration description features of the real-time final object corresponding to the geological exploration description features of the real-time initial object based on the position relationship between the real-time start queue position and each final queue position.

[0185] Step 1010, the geological exploration data theme extraction thread determines the real-time object from the secondary geological exploration data to be analyzed according to the geological exploration description features of the real-time initial object and the geological exploration description features of the real-time final object.

[0186] Step 1012, the geological exploration data theme extraction thread performs weighted sum calculation on the geological exploration description features of the initial object and the geological exploration description features of the final object to obtain the target integration features corresponding to the object, and clusters each geological exploration description feature in the geological exploration description feature queue with the target integration features respectively to obtain each target clustering result to be processed corresponding to the object;

[0187] Step 1014: The geological exploration data theme extraction thread inputs each clustering result of the target to be processed into the geological exploration data semantic extraction thread for geological exploration data semantic recognition, obtains the geological exploration data semantic types corresponding to each clustering result of the target to be processed, and obtains the geological exploration data semantic analysis result corresponding to the object based on the geological exploration data semantic types corresponding to each clustering result of the target to be processed.

[0188] On the basis of the above, a secondary analysis device for geological exploration data based on artificial intelligence is provided. The device includes:

[0189] A feature queue obtaining module, configured to obtain the secondary geological exploration data to be analyzed, perform feature extraction processing on the secondary geological exploration data to be analyzed, and obtain a geological exploration description feature queue;

[0190] A description feature obtaining module, configured to perform start topic description content recognition in combination with the geological exploration description feature queue to obtain the geological exploration description features of the start object, and perform final topic description content recognition in combination with the geological exploration description feature queue to obtain the geological exploration description features of the final object;

[0191] A feature integration module, configured to integrate based on the geological exploration description features of each object to obtain the integrated features corresponding to the object;

[0192] A feature clustering module, configured to cluster each geological exploration description feature in the geological exploration description feature queue with the integrated features respectively to obtain each clustering result of the target to be processed corresponding to the object;

[0193] A result analysis module, configured to perform topic geological exploration data semantic recognition on each clustering result of the target to be processed, obtain the geological exploration data semantic types corresponding to each clustering result of the target to be processed, and obtain the geological exploration data semantic analysis result corresponding to the object in combination with the geological exploration data semantic types corresponding to each clustering result of the target to be processed.

[0194] On the basis of the above, a secondary analysis system for geological exploration data based on artificial intelligence is shown, including a processor and a memory that communicate with each other. The processor is configured to read and execute a computer program from the memory to implement the above method.

[0195] On the basis of the above, a computer-readable storage medium is further provided, and the computer program stored thereon implements the above method when running.

[0196] In summary, based on the above solution, by obtaining the secondary geological exploration data to be analyzed, performing feature extraction processing on the secondary geological exploration data to be analyzed, a geological exploration description feature queue is obtained. Based on the geological exploration description feature queue, the identification of the theme description content is performed to obtain the geological exploration description features of each object, and the object is determined based on the geological exploration description features of each object. Based on the geological exploration description features of each object, integration is performed to obtain the integrated features corresponding to the object. Each geological exploration description feature in the geological exploration description feature queue is respectively clustered with the integrated features to obtain each to-be-processed target clustering result corresponding to the object. The semantic recognition of the theme geological exploration data is performed on each to-be-processed target clustering result to obtain the geological exploration data semantic types corresponding to each to-be-processed target clustering result, and the geological exploration data semantic analysis result corresponding to the object is obtained based on the geological exploration data semantic types corresponding to each to-be-processed target clustering result. Since the integrated features corresponding to the object are clustered with each geological exploration description feature in the geological exploration description feature queue to obtain each to-be-processed target clustering result, and then each to-be-processed target clustering result is used for the semantic recognition of the theme geological exploration data, the semantic correlation between the geological exploration data semantics and the geological exploration data semantics of the object can be extracted in the semantic recognition of the theme geological exploration data, the accuracy of the obtained geological exploration data semantic types is improved, and then the geological exploration data semantic analysis result corresponding to the object is obtained based on the geological exploration data semantic types, that is, the accuracy of the obtained geological exploration data semantics is improved, thereby improving the analysis accuracy of the geological exploration data.

[0197] It should be understood that the systems and their modules shown above can be implemented in various ways. For example, in some embodiments, the systems and their modules can be implemented through hardware, software, or a combination of software and hardware. Among them, the hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art can understand that the above methods and systems can be implemented using computer-executable instructions and / or included in the processor control code. For example, such code is provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The systems and their modules of the present application can be implemented not only by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips and transistors, or programmable hardware devices such as field programmable gate arrays and programmable logic devices, but also by software executed by various types of processors, or by a combination of the above hardware circuits and software (for example, firmware).

[0198] It should be noted that the beneficial effects that may be produced by different embodiments are different. In different embodiments, the beneficial effects that may be produced can be any one or several combinations of the above, or any other beneficial effects that may be obtained.

Claims

1. A secondary analysis method for geological exploration data based on artificial intelligence, characterized in that, The method includes: Obtain the secondary geological exploration data to be analyzed, perform feature extraction processing on the secondary geological exploration data to be analyzed, and obtain a geological exploration description feature queue; wherein, the secondary geological exploration data is obtained by integrating the data after the first exploration and then conducting the second exploration to determine whether there is any ore information that has been missed underground. Combine the geological exploration description feature queue to identify the start topic description content, obtain the geological exploration description features of the start object, and combine the geological exploration description feature queue to identify the final topic description content, obtain the geological exploration description features of the final object; determine the object from the secondary geological exploration data to be analyzed by combining the geological exploration description features of the start object and the geological exploration description features of the final object. Integrate based on the geological exploration description features of each object to obtain the integrated features corresponding to the object. Cluster each geological exploration description feature in the geological exploration description feature queue with the integrated features respectively to obtain each target clustering result to be processed corresponding to the object. Perform semantic recognition of the topic geological exploration data on each target clustering result to be processed, obtain the geological exploration data semantic types corresponding to each target clustering result to be processed, and combine the geological exploration data semantic types corresponding to each target clustering result to be processed to obtain the geological exploration data semantic analysis result corresponding to the object.

2. The method according to claim 1, characterized in that, The integrating based on the geological exploration description features of each object to obtain the integrated features corresponding to the object includes: Perform weighted sum calculation on the geological exploration description features of the start object and the geological exploration description features of the final object to obtain the target integrated features corresponding to the object.

3. The method according to claim 2, characterized in that, The performing weighted sum calculation on the geological exploration description features of the start object and the geological exploration description features of the final object to obtain the target integrated features corresponding to the object includes: Calculate the start weighted vector corresponding to the geological exploration description features of the start object based on a preset start confidence level, and calculate the final weighted vector corresponding to the geological exploration description features of the final object based on a preset final confidence level. Calculate the vector sum of the start weighted vector and the final weighted vector to obtain the target integrated features corresponding to the object.

4. The method according to claim 1, wherein The geological exploration description features of the start object include no less than two, and the geological exploration description features of the final object include no less than two; the determining the object from the secondary geological exploration data to be analyzed by combining the geological exploration description features of the start object and the geological exploration description features of the final object includes: Determine the geological exploration description features of the real-time start object from each geological exploration description feature of the start object. Obtain the real-time start queue position of the geological exploration description features of the real-time start object in the geological exploration description feature queue, and obtain each final queue position of each geological exploration description feature of the final object in the geological exploration description feature queue. Determine the geological exploration description features of the real-time final object corresponding to the geological exploration description features of the real-time start object by combining the positional relationship between the real-time start queue position and each final queue position. Determine a real-time object from the secondary geological exploration data to be analyzed according to the geological exploration description features of the real-time start object and the geological exploration description features of the real-time final object.

5. The method according to claim 4, wherein The determining of the geological exploration description features of the real-time final object corresponding to the geological exploration description features of the real-time start object by combining the positional relationship between the real-time start queue position and each final queue position includes: Select each final queue position by combining the real-time start queue position to obtain each target final queue position; Calculate the positional differences between the real-time start queue position and each of the target final queue positions respectively, and determine the minimum positional difference from each positional difference; Use the geological exploration description features of the final object corresponding to the minimum positional difference as the geological exploration description features of the real-time final object.

6. The method according to claim 1, wherein The integrating based on the geological exploration description features of each object to obtain the integrated features corresponding to the object includes: Obtain the preset confidence levels and the total number of vectors corresponding to the geological exploration description features of each object; Perform weighted average calculation on the geological exploration description features of each object by combining the preset confidence levels and the total number of vectors corresponding to the geological exploration description features of each object to obtain the integrated features corresponding to the object.

7. The method according to claim 1, characterized in that, The method further includes: Input the secondary geological exploration data to be analyzed into a geological exploration data topic extraction thread for topic extraction to obtain the output object and the geological exploration data semantic analysis result corresponding to the object; The geological exploration data topic extraction thread is pre-trained using an artificial intelligence thread based on training examples, and the training examples include training geological exploration data, the object directory corresponding to the training geological exploration data, and the geological exploration data semantic directory corresponding to the training geological exploration data.

8. The method according to claim 7, wherein The geological exploration data topic extraction thread includes an object extraction thread and a geological exploration data semantic extraction thread; the inputting of the secondary geological exploration data to be analyzed into the geological exploration data topic extraction thread for topic extraction to obtain the output object and the geological exploration data semantic analysis result corresponding to the object includes: Perform feature extraction processing on the secondary geological exploration data to be analyzed to obtain the geological exploration description feature queue; Input the geological exploration description feature queue into the object extraction thread for topic description content recognition to obtain the geological exploration description features of each object, and determine the object by combining the geological exploration description features of each object; Integrate the geological exploration description features of each object to obtain the integrated features corresponding to the object, and cluster each geological exploration description feature in the geological exploration description feature queue with the integrated features respectively to obtain each to-be-processed target clustering result corresponding to the object; Input each to-be-processed target clustering result into the geological exploration data semantic extraction thread for geological exploration data semantic recognition to obtain the geological exploration data semantic types corresponding to each to-be-processed target clustering result, and obtain the geological exploration data semantic analysis result corresponding to the object by combining the geological exploration data semantic types corresponding to each to-be-processed target clustering result; Among them, the object extraction thread includes a start object extraction thread and a final object extraction thread; the step of inputting the geological exploration description feature queue into the object extraction thread to perform topic description content recognition, obtaining the geological exploration description features of each object, and determining the object in combination with the geological exploration description features of each object includes: Inputting the geological exploration description feature queue into the start object extraction thread to perform start description content recognition, obtaining the geological exploration description features of the start object, and inputting the geological exploration description feature queue into the final object extraction thread to perform final description content recognition, obtaining the geological exploration description features of the final object; Determining the object from the secondary geological exploration data to be analyzed in combination with the geological exploration description features of the start object and the geological exploration description features of the final object; Among them, the training of the geological exploration data topic extraction thread includes the following steps: Performing feature extraction processing on the training geological exploration data in the training example to obtain a training geological exploration description feature queue; Inputting the training geological exploration description feature queue into the original object extraction thread to perform topic description content recognition, obtaining the geological exploration description features of each training object, and determining the original object in combination with the geological exploration description features of each training object; Integrating in combination with the geological exploration description features of each training object to obtain the original integrated features corresponding to the original object, and clustering each training geological exploration description feature in the training geological exploration description feature queue with the original integrated features respectively to obtain each original to-be-processed target clustering result corresponding to the original object; Inputting each original to-be-processed target clustering result into the original geological exploration data semantic extraction thread to perform geological exploration data semantic recognition, obtaining the original geological exploration data semantic types corresponding to each original to-be-processed target clustering result, and obtaining the original geological exploration data semantics corresponding to the original object in combination with the original geological exploration data semantic types corresponding to each original to-be-processed target clustering result; Calculating object evaluation index information by using the object evaluation index algorithm in combination with the original object and the object directory corresponding to the training geological exploration data, and calculating geological exploration data semantic evaluation index information by using the geological exploration data semantic evaluation index algorithm in combination with the original geological exploration data semantics and the geological exploration data semantic directory corresponding to the training geological exploration data; Determining thread evaluation index information in combination with the object evaluation index information and the geological exploration data semantic evaluation index information, and optimizing the original object extraction thread and the original geological exploration data semantic extraction thread in combination with the thread evaluation index information. When the training is completed, the geological exploration data topic extraction thread is obtained; Among them, the step of performing feature extraction processing on the training geological exploration data in the training example to obtain a training geological exploration description feature queue includes: Inputting the training geological exploration data in the training example into the geological exploration data feature distribution thread to perform feature distribution, obtaining the training geological exploration description feature queue, and the geological exploration data feature distribution thread is obtained by training in advance using an artificial intelligence thread based on a feature distribution training example. Among them, the original object extraction thread includes an original start object extraction thread and an original end object extraction thread; the step of inputting the training geological exploration description feature queue into the original object extraction thread to perform topic description content recognition, obtaining the geological exploration description features of each training object, and determining the original object in combination with the geological exploration description features of each training object includes: Inputting the training geological exploration description feature queue into the original start object extraction thread to perform start description content recognition, and obtaining the geological exploration description features of the original start object; Inputting the training geological exploration description feature queue into the original end object extraction thread to perform end description content recognition, and obtaining the geological exploration description features of the original end object; Determining the original object from the training geological exploration data in combination with the geological exploration description features of the original start object and the geological exploration description features of the original end object; The step of calculating the object evaluation index information by using the object evaluation index algorithm in combination with the original object and the object directory corresponding to the training geological exploration data includes: Calculating the start object evaluation index information by using the start object evaluation index algorithm in combination with the geological exploration description features of the original start object and the start object directory in the object directory; Calculating the end object evaluation index information by using the end object evaluation index algorithm in combination with the geological exploration description features of the original end object and the end object directory in the object directory, and obtaining the object evaluation index information by combining the start object evaluation index information and the end object evaluation index information.

9. The method according to claim 1, wherein The step of determining the object from the secondary geological exploration data to be analyzed in combination with the geological exploration description features of the start object and the geological exploration description features of the end object includes: obtaining the object start description attribute corresponding to the geological exploration description features of the start object, the object end description attribute corresponding to the geological exploration description features of the end object, and the description attribute between the object start description attribute and the object end description attribute from the secondary geological exploration data to be analyzed, and obtaining the object.

10. An artificial intelligence-based secondary analysis system for geological exploration data, characterized in that, It includes a processor and a memory that communicate with each other, and the processor is configured to read and execute a computer program from the memory to implement the method according to any one of claims 1-9.

Images