Intelligent search method and system for oil and gas field surface engineering cases

Abstract

The invention discloses an intelligent search method and system for oil and gas field surface engineering cases, and the method comprises the steps: building an oil and gas field surface engineering case library covering multiple modes based on a knowledge graph technology, and providing resource support for the intelligent search system. Meanwhile, modeling and analysis are conducted on retrieval information of the users, the most relevant oil and gas field ground engineering case search result is customized for each user through a recommendation algorithm based on a decision-making tree, better user experience is provided, and efficient management of ground engineering projects of oil and gas enterprises is facilitated.

Description

Technical Field

[0001] The present invention belongs to the fields of natural language processing technology, machine learning technology and information retrieval technology, and specifically relates to an intelligent search method and system for oil and gas field surface engineering cases. Background Art

[0002] In recent years, the scale of surface engineering construction projects in oil and gas fields has continued to expand, the number has increased, and the construction period has become more urgent. The resulting surface construction management tasks have increased dramatically. Therefore, there is an urgent need to adopt information technology to support the efficient management of project construction.

[0003] The existing oil and gas field surface engineering case search system is mainly based on traditional search technology, which mainly relies on keyword matching to determine the relevance of search results. However, keyword matching often has limitations and cannot fully understand the user's query intent, resulting in limited accuracy and coverage of search results. Due to the algorithmic limitations of search engines, traditional search technology often has difficulty in providing accurate and useful search results when dealing with complex engineering problems. As a result, users need to spend a lot of time and energy to screen and verify the quality of search results. Traditional search technology faces an increasing amount of information, resulting in an overload of search results.

[0004] At the same time, engineering cases usually contain multiple types of data such as text, images, and videos. However, traditional search technologies are difficult to effectively process so many modal data and cannot fully utilize information such as images and videos, thus limiting the accuracy and diversity of search results. Traditional search technologies often cannot provide personalized search results based on user preferences and historical behaviors. The engineering case search results obtained by users lack customization and cannot meet their different needs. Summary of the invention

[0005] In order to solve the problems that the existing oil and gas field surface engineering case search technology based on traditional search technology limits the accuracy or diversity of search results and cannot meet different needs, the present invention provides an intelligent search method and system for oil and gas field surface engineering cases. The present invention can provide more accurate and diversified search results, meet users' needs for engineering cases, and improve search efficiency and user experience.

[0006] The present invention is achieved through the following technical solutions:

[0007] An intelligent search method for oil and gas field surface engineering cases, the intelligent search method comprising:

[0008] Case classification and database management: Classify the oil and gas field surface engineering cases uploaded by users, pre-process the oil and gas field surface engineering cases, extract feature keywords, and build a case library based on the knowledge graph;

[0009] Receive user search information: Receive the search information entered by the user through the built search interface;

[0010] Case search: Two search modes are provided for users, including case search mode and intelligent search mode. When users select case search mode, matching is performed based on the basic information and complete information of the project, and the matching result is the project entity. When users select intelligent search mode, matching is performed based on the keywords in the search information entered by the user, and the basic information of the project case and the nine disassembly item elements are matched, and the matching result is all relevant project case files containing the keywords.

[0011] Search result display: supports two sorting modes, including sorting by the importance of case information and sorting by the relevance of search results to search information;

[0012] Search log generation: record historical search information, including search information structured fields and search results;

[0013] Intelligent case recommendation: Analyze the historical search information, extract user search features, and use a decision tree-based recommendation algorithm to perform intelligent case recommendation.

[0014] The intelligent search method proposed in the present invention is based on knowledge graph technology to establish a multi-modal oil and gas field surface engineering case library to provide resource support for the intelligent search system. At the same time, the present invention also models and analyzes the user's retrieval information, and adopts a decision tree-based recommendation algorithm to tailor the most relevant oil and gas field surface engineering case search results for each user, thereby providing a better user experience and helping oil and gas companies to achieve efficient management of surface engineering projects.

[0015] As a preferred embodiment, the present invention classifies the oil and gas field surface engineering cases uploaded by users, and specifically classifies the oil and gas field surface engineering cases according to single engineering cases and typical construction cases;

[0016] The extracted feature keywords include basic information and complete information of the engineering case;

[0017] The basic information of the engineering case includes: project name, project number, project location, keywords, start time, acceptance time, transfer time, whether it is digitally handed over, project pictures, oil and gas field type, engineering type and special environment;

[0018] The complete information includes: construction unit, design unit, supervision unit and construction unit.

[0019] As a preferred embodiment, after completing the storage operation of oil and gas field surface engineering cases, the method of the present invention further includes:

[0020] Combined with the oil and gas field surface engineering implementation standards and the reference opinions of field experts on the importance of various types of information, the Delphi method and information weight method are used to calculate the weights of various types of information respectively;

[0021] The weighted average method is then used to assign weights to case information.

[0022] As a preferred embodiment, after receiving the search information input by the user, the method of the present invention further includes:

[0023] An entity extraction model based on LSTM algorithm is used to extract retrieval information entities.

[0024] As a preferred embodiment, the process of establishing the entity extraction model based on the LSTM algorithm of the present invention specifically includes:

[0025] Select some historical retrieval information and divide it into training set and test set according to the preset ratio;

[0026] Using BIO annotation to perform entity annotation on the retrieval information in the training set and the test set;

[0027] Using the training set as input data, training an entity extraction model based on the LSTM algorithm;

[0028] Using the test set to verify the extraction effect of the trained entity extraction model, comparing the annotation information extracted by the entity extraction model from the search information in the test set with the manually labeled annotation information, and calculating the accuracy thereof;

[0029] The model with a correctness greater than the threshold after testing is used to extract the search information input by the user, and the entity extraction based on the LSTM algorithm is completed. The present invention also uses an entity extraction model based on the LSTM algorithm to process the user's search information, ensuring that the intelligent search system can better understand the user's query intention, not only relying on keyword matching, but can provide more accurate and precise search results based on context and semantic relevance.

[0030] As a preferred embodiment, the nine disassembly item elements of the present invention include: case overview, key parameters, operation time, construction drawing documents, construction, construction measures, HSE measures, process record data, pictures and images.

[0031] As a preferred embodiment, the method of sorting according to the importance of case information of the present invention is specifically sorting according to the weight of the importance of case information;

[0032] The method of sorting the search results according to their relevance to the search information is specifically implemented through a vector space model.

[0033] As a preferred embodiment, the method of sorting the search results according to the relevance of the search information of the present invention specifically includes:

[0034] The BOW model is used to segment the word vectors generated by the search information and search results;

[0035] The TF-IDF algorithm is used to calculate the TF-IDF value of each element in the word vector of the search information after word segmentation and the word vector of the search result after word segmentation;

[0036] The word vectors of the search information and the word vectors of the search results are reconstructed according to the TF-IDF value, and the cosine similarity between the two reconstructed word vectors is calculated. The corresponding sorting is performed according to the cosine similarity calculation result.

[0037] As a preferred embodiment, the present invention adopts a decision tree-based recommendation algorithm for intelligent case recommendation, which specifically includes:

[0038] The historical search information is used as the training data set, and the case feature keywords are used as the feature set, which are input into the model for training;

[0039] If all cases in the training data set belong to the same label, the decision tree is a single-node tree, and the label is used as the label of the node. The decision tree is output and the corresponding oil and gas field surface engineering case output is matched in the case library according to the label;

[0040] If the feature set is empty, the decision tree is a single-node tree. The label of the class with the largest number of cases in the training data set is used as the label of the node. The decision tree is output and the corresponding oil and gas field surface engineering case output is matched in the case library according to the label;

[0041] If the feature set is not empty, calculate the information gain of the features in the feature set for the training data set, and select the feature with the largest information gain;

[0042] If the information gain of the feature with the largest information gain is less than the initial threshold, the decision tree is a single-node tree, and the label of the class with the largest number of cases in the training data set is used as the label of the node. The decision tree is output and the corresponding oil and gas field surface engineering case output is matched in the case library according to the label;

[0043] If the new gain of the feature with the largest information gain is greater than the initial threshold, then for each possible value of the feature, the training data set is divided into several non-empty subsets, and the labels of the classes with the largest number of cases in the several non-empty subsets are used as node labels to construct child nodes, and the child nodes and their subtrees constitute a decision tree;

[0044] For any node in the decision tree, the corresponding non-empty subset is used as the training set. The feature with the largest information gain is removed from the original feature set as the feature set. The above steps are recursively called until only the label of the class with the largest number of cases in the subset is on the subtree. According to the label, the corresponding oil and gas field surface engineering case output is matched in the case library.

[0045] On the other hand, the present invention also proposes an intelligent search system for oil and gas field surface engineering cases, the intelligent search system comprising:

[0046] A case library, wherein the case library is used to store oil and gas field surface engineering cases;

[0047] A search information input module, which provides a search interface for users to input search information;

[0048] Case search module, which provides two search modes for users, including case search mode and intelligent search mode; when the user selects the case search mode, matching is performed based on the basic information and complete information of the engineering project, and the matching result is the engineering project entity; when the user selects the intelligent search mode, matching is performed based on the keywords in the search information input by the user, and the matching result is all related engineering case files containing the keywords;

[0049] A search result display module, which is used to display the search results and supports two sorting modes, including sorting by importance of case information and sorting by relevance of the search results to the search information;

[0050] A search log generation module, wherein the search log generation module is used to record historical search information, the search information including search information structured fields and search results;

[0051] and an intelligent recommendation module, wherein the intelligent recommendation module analyzes the historical search information, extracts user search features, and uses a decision tree-based recommendation algorithm to perform intelligent case recommendations.

[0052] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0053] 1. The intelligent search method and system proposed in the present invention can provide more accurate, relevant and diversified search results, meet users' needs for engineering cases, and improve search efficiency and user experience;

[0054] 2. Based on the knowledge graph technology, the present invention establishes a multi-modal oil and gas field surface engineering case library to provide resource support for the intelligent search system. At the same time, by modeling and analyzing the user's historical search information, a decision tree-based recommendation algorithm is used to tailor the most relevant oil and gas field surface engineering case search results for each user, providing technical support for the efficient management of oil and gas field surface engineering projects;

[0055] 3. The present invention also uses an entity extraction model based on the LSTM algorithm to process user retrieval information, ensuring that the intelligent search system can better understand the user's query intention, not only relying on keyword matching, but also providing more accurate and precise search results based on context and semantic relevance. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:

[0057] Figure 1 The figure is a flow chart of the intelligent search method according to an embodiment of the present invention.

[0058] Figure 2 This is a functional block diagram of an intelligent search system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.

[0060] Example:

[0061] The existing oil and gas field surface engineering case search system based on traditional search technology is difficult to provide accurate and useful search results due to the limitations of traditional search technology, and is inefficient and limits the diversity of search results, and cannot meet the different needs of users. In view of this, this embodiment proposes an intelligent search method for oil and gas field surface engineering cases. The intelligent search method proposed in this embodiment uses knowledge graph technology, machine learning algorithms, etc. to provide more accurate and diverse search results, meet users' personalized needs for engineering cases, and improve search efficiency.

[0062] like Figure 1 As shown, the intelligent search method proposed in this embodiment specifically includes the following steps:

[0063] Step 1: Case classification and storage management: Classify the oil and gas field surface engineering cases uploaded by users, pre-process the case texts using OCR and NLP technologies, extract feature keywords, and build a case library based on the knowledge graph;

[0064] Step 2, receiving user search information: the interaction of information search is that the user and the oil and gas field surface engineering case library build a search interface for receiving the search information input by the user;

[0065] Step 3, case search: two search modes are provided for users, including case search mode and intelligent search mode; when the user selects the case search mode, it mainly matches based on the basic information and complete information of the engineering case, and the matching result is the engineering project entity; when the user selects the intelligent search mode, it matches the basic information of the engineering case and the nine disassembly item elements according to the keywords in the user input information, and the matching result is all related engineering case files containing the keywords;

[0066] Step 4, search result display: supports two sorting modes, including sorting by the importance of case information and sorting by the relevance of search results to search information. When the user selects the case search mode, the search results are sorted by the importance of the entity in the knowledge graph structure; when the user selects the intelligent search mode, the search results are sorted by the relevance of the search results to search information;

[0067] Step 5, retrieval log generation: record historical retrieval information as training data for the intelligent recommendation model. The retrieval information includes retrieval information structured fields and retrieval results.

[0068] Step 6, intelligent case recommendation: Analyze historical retrieval information, extract user retrieval features, and use a decision tree-based recommendation algorithm to perform intelligent case recommendation.

[0069] Furthermore, the classification step of step 1 specifically classifies the oil and gas field surface engineering cases according to single engineering cases and typical construction cases, and the extracted characteristic keywords include basic information and complete information of the engineering cases.

[0070] Among them, the basic information includes: project name, project number, project location, keywords, start time, acceptance time, transfer time, whether it is digitally handed over, project pictures, oil and gas field type, project type, special environment; complete information includes: construction unit, design unit, supervision unit, and construction unit.

[0071] Furthermore, after completing the storage operation of the oil and gas field surface engineering case in step 1, it also includes:

[0072] Combined with the implementation standards of oil and gas field surface engineering and the reference opinions of field experts on the importance of various types of information, the Delphi method and information weight method are used to calculate the weights of various types of information respectively, and then the weighted average method is used to assign weights to case information.

[0073] Furthermore, after receiving the user's search information in step 2, an entity extraction model based on the LSTM algorithm is used to extract search information entities, wherein the specific process of establishing and applying the model includes:

[0074] Step 201, select part of the historical search information and divide it into a training set and a test set according to a ratio of 7:3; it should be noted that the division according to the ratio of 7:3 is only an exemplary description and is not limited to this, and the division can be performed according to actual needs;

[0075] Step 202, using BIO annotation to perform entity annotation on the retrieval information in the training set and the test set; wherein the BIO annotation is a three-digit annotation system, B represents the beginning of entity X, I represents the structure of entity X, and O represents an entity that does not belong to any type;

[0076] Step 203, using the training set as input data, training an entity extraction model based on the LSTM algorithm;

[0077] Step 204, using the test set to verify the entity extraction effect of the trained model, compare the annotation information extracted by the entity extraction model from the search information in the test set with the annotation information manually marked by the on-site staff, and calculate their accuracy. The accuracy calculation formula is as follows (1):

[0078]

[0079] Where A is the accuracy of entity extraction; T l The number of characters labeled for the model that is the same as the human annotation; F l The number of characters that differ between the model's labels and the human labels.

[0080] Step 205, use the model with a correct rate greater than 90% after testing to extract the search information input by the user, and complete the entity extraction based on the LSTM algorithm. It should be noted that the selection of a model with a correct rate greater than 90% in this embodiment is only an exemplary description and is not limited to this. It can be selected according to actual needs.

[0081] Furthermore, the nine disassembly elements are case overview, key parameters, operation time, construction drawing documents, construction, construction measures, HSE measures, process record data, pictures and images.

[0082] Furthermore, the method of sorting by importance of case information in step 4 is specifically to sort by the weight of the importance of case information; sorting by the relevance of the search results and the search information is specifically implemented by a vector space model, and the specific process includes:

[0083] Step 401, using the BOW model to segment the word vectors generated from the search information and the search results;

[0084] Step 402, using the TF-ID algorithm, calculate the TF-IDF value of each element in the word vector a after word segmentation of the search information and the word vector β after word segmentation of the search result, and the calculation formula is as follows:

[0085] TF-IDF=TF×IDF (2)

[0086]

[0087]

[0088] In the formula, TF is the term frequency; IDF is the inverse document frequency; X i is the total number of times a search information segment appears in the search results; X is the total number of words in the case text; Y is the total number of cases included in the search results; Y i The number of cases that contain the search information segmentation in the search results;

[0089] Step 403, reconstruct the word vector α' of the search information and the word vector β' of the search result according to the TF-IDF value, calculate the cosine similarity between the two, and perform corresponding sorting according to the calculation results. The calculation formula is as follows (5).

[0090]

[0091] Furthermore, the specific implementation process of using the decision tree-based recommendation algorithm for case intelligent recommendation in step 6 includes:

[0092] Step 601, historical search information is used as a training data set D, and case feature keywords are used as a feature set B, and are input into the model for training;

[0093] Step 602: If all cases in D belong to the same label S k , then the decision tree T is a single-node tree, and the label S k As the label of the node, output the decision tree T, according to the label S k Match the corresponding oil and gas field surface engineering cases in the case library and output them as intelligent recommendation results;

[0094] Step 603, if Then T is a single-node tree, and the label S of the class with the largest number of cases in D isk As the label of the node, output the decision tree T, according to the label S k Match the corresponding oil and gas field surface engineering cases in the case library and output them as intelligent recommendation results;

[0095] Step 604, if Then calculate the information gain of feature pair D in B and select feature B with the largest information gain K , the calculation formula is as follows (6)-(8):

[0096] g(D,B)=H(D)-H(D|B) (6)

[0097]

[0098]

[0099] In the formula, g(D,B) is the information gain of feature set B to training data set D, which means that feature x i The information uncertainty of the training data set D is reduced by the information of the training data set D; H(D) is the entropy of the training data set D; H(D|B) is the feature x i Conditional entropy under given conditions;

[0100] Step 605, if B K If the information gain of is less than the initial threshold ε, then T is a single-node tree, and the label S of the class with the largest number of cases in D is k As the label of the node, output the decision tree T, according to the label S k Match the corresponding oil and gas field surface engineering cases in the case library and output them as intelligent recommendation results;

[0101] Step 606, if B K The information gain of B is greater than the initial threshold ε, then K For every possible value bi, according to B K =b i , split D into several non-empty subsets D i , D i The label S of the class with the largest number of cases k As the child node label, construct the child node, and the child node and its subtree constitute T;

[0102] Step 607: for the i-th child node, use D i As the training set, B-{B K} is the feature set, and steps 601 to 606 are called recursively until the subtree T i Only S k Class, according to label S kThe corresponding oil and gas field surface engineering cases are matched in the case library and output as intelligent recommendation results.

[0103] The intelligent search method proposed in this embodiment can provide more accurate, relevant and diversified search results, meet users' needs for engineering cases, and improve search efficiency and user experience.

[0104] Based on the same technical concept as above, this embodiment also proposes an intelligent retrieval system for oil and gas field surface engineering cases, such as Figure 2 As shown, the intelligent detection system includes:

[0105] Case library, which is used to store oil and gas field surface engineering cases;

[0106] A search information input module, which provides a search interface for users to input search information;

[0107] Case retrieval module, which provides two retrieval modes for users, including case retrieval mode and intelligent retrieval mode;

[0108] The search result display module is used to display the search results and supports two sorting modes, including sorting by the importance of case information and sorting by the relevance of the search results to the search information;

[0109] A search log generation module, which can collect and count the user's historical search information;

[0110] And, the intelligent recommendation module can analyze the user's historical search information, extract user search features, and use a decision tree-based recommendation algorithm to make intelligent case recommendations.

[0111] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0112] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0113] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0114] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0115] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An intelligent search method for oil and gas field surface engineering cases, It is characterized in that The intelligent search method comprises: Case classification and database management: Classify the oil and gas field surface engineering cases uploaded by users, pre-process the oil and gas field surface engineering cases, extract feature keywords, and build a case library based on the knowledge graph; Receive user search information: Receive the search information entered by the user through the built search interface; Case search: Two search modes are provided for users, including case search mode and intelligent search mode. When users select case search mode, matching is performed based on the basic information and complete information of the project, and the matching result is the project entity. When users select intelligent search mode, matching is performed based on the keywords in the search information entered by the user, and the basic information of the project case and the nine disassembly item elements are matched, and the matching result is all relevant project case files containing the keywords. Search result display: supports two sorting modes, including sorting by the importance of case information and sorting by the relevance of search results to search information; Search log generation: record historical search information, including search information structured fields and search results; Intelligent case recommendation: Analyze the historical search information, extract user search features, and use a decision tree-based recommendation algorithm to perform intelligent case recommendation.

2. According to claim 1, an intelligent search method for oil and gas field surface engineering cases, It is characterized in that Classify the oil and gas field surface engineering cases uploaded by users, and specifically classify the oil and gas field surface engineering cases into single engineering cases and typical construction cases; The extracted feature keywords include basic information and complete information of the engineering case; The basic information of the engineering case includes: project name, project number, project location, keywords, start time, acceptance time, transfer time, whether it is digitally handed over, project pictures, oil and gas field type, engineering type and special environment; The complete information includes: construction unit, design unit, supervision unit and construction unit.

3. According to claim 2, an intelligent search method for oil and gas field surface engineering cases, It is characterized in that After completing the storage of oil and gas field surface engineering cases, it also includes: Combined with the oil and gas field surface engineering implementation standards and the reference opinions of field experts on the importance of various types of information, the Delphi method and information weight method are used to calculate the weights of various types of information respectively; The weighted average method is then used to assign weights to case information.

4. According to claim 1, an intelligent search method for oil and gas field surface engineering cases, It is characterized in that After receiving the search information input by the user, it also includes: An entity extraction model based on LSTM algorithm is used to extract retrieval information entities.

5. According to claim 4, an intelligent search method for oil and gas field surface engineering cases, It is characterized in that The process of establishing the entity extraction model based on the LSTM algorithm specifically includes: Select some historical retrieval information and divide it into training set and test set according to the preset ratio; Using BIO annotation to perform entity annotation on the retrieval information in the training set and the test set; Using the training set as input data, training an entity extraction model based on the LSTM algorithm; Using the test set to verify the extraction effect of the trained entity extraction model, comparing the annotation information extracted by the entity extraction model from the search information in the test set with the manually labeled annotation information, and calculating the accuracy thereof; The model with a tested accuracy greater than the threshold is used to extract the search information input by the user and complete the entity extraction based on the LSTM algorithm.

6. The intelligent search method for oil and gas field surface engineering cases according to claim 1, It is characterized in that The nine dismantling elements include: case overview, key parameters, operation time, construction drawings, construction, construction measures, HSE measures, process record data, pictures and images.

7. The intelligent search method for oil and gas field surface engineering cases according to claim 1, It is characterized in that The method of sorting by the importance of case information is specifically to sort according to the weight of the importance of case information; The method of sorting the search results according to their relevance to the search information is specifically implemented through a vector space model.

8. The intelligent search method for oil and gas field surface engineering cases according to claim 7, It is characterized in that The specific methods of sorting the search results according to their relevance to the search information include: The BOW model is used to segment the word vectors generated by the search information and search results; The TF-IDF algorithm is used to calculate the TF-IDF value of each element in the word vector of the search information after word segmentation and the word vector of the search result after word segmentation; The word vectors of the search information and the word vectors of the search results are reconstructed according to the TF-IDF value, and the cosine similarity between the two reconstructed word vectors is calculated. The corresponding sorting is performed according to the cosine similarity calculation result.

9. The intelligent search method for oil and gas field surface engineering cases according to claim 1, It is characterized in that A decision tree-based recommendation algorithm is used for intelligent case recommendation, including: The historical search information is used as the training data set, and the case feature keywords are used as the feature set, which are input into the model for training; If all cases in the training data set belong to the same label, the decision tree is a single-node tree, and the label is used as the label of the node. The decision tree is output and the corresponding oil and gas field surface engineering case output is matched in the case library according to the label; If the feature set is empty, the decision tree is a single-node tree. The label of the class with the largest number of cases in the training data set is used as the label of the node. The decision tree is output and the corresponding oil and gas field surface engineering case output is matched in the case library according to the label; If the feature set is not empty, calculate the information gain of the features in the feature set for the training data set, and select the feature with the largest information gain; If the information gain of the feature with the largest information gain is less than the initial threshold, the decision tree is a single-node tree, and the label of the class with the largest number of cases in the training data set is used as the label of the node. The decision tree is output and the corresponding oil and gas field surface engineering case output is matched in the case library according to the label; If the new gain of the feature with the largest information gain is greater than the initial threshold, then for each possible value of the feature, the training data set is divided into several non-empty subsets, and the labels of the classes with the largest number of cases in the several non-empty subsets are used as node labels to construct child nodes, and the child nodes and their subtrees constitute a decision tree; For any node in the decision tree, the corresponding non-empty subset is used as the training set. The feature with the largest information gain is removed from the original feature set as the feature set. The above steps are recursively called until only the label of the class with the largest number of cases in the subset is on the subtree. According to the label, the corresponding oil and gas field surface engineering case output is matched in the case library.

10. An intelligent search system for oil and gas field surface engineering cases, It is characterized in that The intelligent search system comprises: A case library, wherein the case library is used to store oil and gas field surface engineering cases; A search information input module, which provides a search interface for users to input search information; Case search module, which provides two search modes for users, including case search mode and intelligent search mode; when the user selects the case search mode, matching is performed based on the basic information and complete information of the engineering project, and the matching result is the engineering project entity; when the user selects the intelligent search mode, matching is performed based on the keywords in the search information input by the user, and the matching result is all related engineering case files containing the keywords; A search result display module, which is used to display the search results and supports two sorting modes, including sorting by importance of case information and sorting by relevance of the search results to the search information; A search log generation module, wherein the search log generation module is used to record historical search information, the search information including search information structured fields and search results; and an intelligent recommendation module, wherein the intelligent recommendation module analyzes the historical search information, extracts user search features, and uses a decision tree-based recommendation algorithm to perform intelligent case recommendations.

Images