Power grid project review data generation method and device, equipment and storage medium

Abstract

The invention relates to a power grid project review data generation method and device, equipment and a storage medium. The method comprises the following steps: acquiring project attribute data extracted from power grid project data; screening out a plurality of target review objects matched with the project attribute data from a plurality of preset review objects; obtaining project review data of each target review object for the power grid project data, extracting review feature data from the plurality of project review data, the review feature data representing technical feasibility features, execution risk features and compliance features corresponding to the power grid project data; and according to the review feature data, determining target project review data from the plurality of project review data. By adopting the method, the standardization of review data generation can be improved.

Description

Technical Field

[0001] This application relates to the field of power grid review technology, and in particular to a method, apparatus, equipment and storage medium for generating review data for power grid projects. Background Technology

[0002] Power grid projects are major infrastructure construction projects, characterized by large-scale investment, complex technology, and high safety requirements. Therefore, conducting scientific, rigorous, and efficient reviews of power grid projects and generating high-quality review opinions is a crucial step in ensuring the project's investment benefits and technical feasibility. Review opinions not only provide direct evidence for project decisions but also serve as important guidance for subsequent design optimization and risk management.

[0003] However, the review of power grid projects in related technologies largely relies on the personal experience of review experts, and the lack of standardization in the review process is due to the differences in the experts' field backgrounds and experience. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for generating review data for power grid projects that can improve the standardization of review, in response to the above-mentioned technical problems.

[0005] Firstly, this application provides a method for generating review data for power grid projects, including:

[0006] Obtain project attribute data extracted from power grid project data;

[0007] Select multiple target review objects that match the project attribute data from a number of preset review objects;

[0008] Obtain project review data for each target review object related to power grid project data, extract review feature data from multiple project review data, and the review feature data characterizes the technical feasibility characteristics, execution risk characteristics, and compliance characteristics corresponding to the power grid project data;

[0009] Based on the review characteristic data, the target project review data is determined from multiple project review data.

[0010] Secondly, this application also provides a device for generating review data for power grid projects, comprising:

[0011] The data acquisition module is used to acquire project attribute data extracted from power grid project data;

[0012] The review object determination module is used to filter out multiple target review objects that match the project attribute data from a number of preset review objects;

[0013] The review feature data extraction module is used to obtain project review data for each target review object for power grid project data, and extract review feature data from multiple project review data. The review feature data characterizes the technical feasibility characteristics, execution risk characteristics and compliance characteristics corresponding to the power grid project data.

[0014] The target review data generation module is used to determine the target project review data from multiple project review data based on review feature data.

[0015] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in any of the above-described embodiments of the power grid project review data generation method.

[0016] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps in any of the above embodiments of the power grid project review data generation method.

[0017] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps in any of the above-described embodiments of the power grid project review data generation method.

[0018] The aforementioned method, apparatus, computer equipment, computer-readable storage medium, and computer program product for generating review data for power grid projects first acquire project attribute data extracted from power grid project data. Second, they select multiple target review objects from a pool of pre-defined review objects that match the project attribute data. Third, they acquire project review data for these target review objects on the power grid project data. Review feature data is then extracted from this data, characterizing the technical feasibility, execution risk, and compliance features corresponding to the power grid project data. Finally, based on the review feature data, target project review data is determined from the multiple project review data sets. This method, by extracting review feature data to generate target project review data, improves the accuracy of review data generation and facilitates adaptation to diverse project review data scenarios, thus enhancing the standardization of review data generation. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is an application environment diagram of a method for generating review data for power grid projects in one embodiment.

[0021] Figure 2 This is a flowchart illustrating a method for generating review data for a power grid project in one embodiment;

[0022] Figure 3 This is a flowchart illustrating a method for generating review data for a power grid project, as described in another embodiment.

[0023] Figure 4 This is a flowchart illustrating a method for generating review data for a power grid project in yet another embodiment;

[0024] Figure 5 A structural block diagram of a power grid project review data generation device in one embodiment;

[0025] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0027] The method for generating review data for power grid projects provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104, or it can be located in the cloud or on another network server.

[0028] Specifically, the operator can upload project attribute data extracted from the power grid project data to the server 104 via terminal 102, and then send a review data generation message to the server 104 via terminal 102. The server 104 obtains the project attribute data extracted from the power grid project data and selects multiple target review objects that match the project attribute data from multiple preset review objects. Next, it obtains the project review data for each target review object for the power grid project data, and extracts review feature data from the multiple project review data. The review feature data characterizes the technical feasibility characteristics, execution risk characteristics, and compliance characteristics corresponding to the power grid project data. Finally, based on the review feature data, it determines the target project review data from the multiple project review data.

[0029] The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle systems, and projection devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted displays. Head-mounted displays can be virtual reality (VR) devices, augmented reality (AR) devices, and smart glasses. The server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0030] In one exemplary embodiment, such as Figure 2 As shown, a method for generating review data for power grid projects is provided, which can be applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps (hereinafter referred to as S): S100 to S400. Wherein:

[0031] S100: Obtain project attribute data extracted from power grid project data.

[0032] The data for power grid projects can include, but is not limited to, structured numerical data, text report data, and spatial geographic data. Understandably, the content of these data will vary depending on the application scenario. For example, structured numerical data can be quantitative indicators such as project investment amount, equipment capacity, and load forecast; text report data can be unstructured documents such as feasibility study reports, technical solutions, and approval contracts; and spatial geographic data can be geospatial information such as project site coordinates, route maps, and topographic information. Project attribute data can include, but is not limited to, project type, technical field, and voltage level.

[0033] In practice, power grid project data, including structured numerical data, text report data, and spatial geographic data, can be obtained through the power grid's project management system. For example, the power grid project to be reviewed could be a newly constructed substation project. Its structured data could include budget sheets, substation electrical equipment capacity, load forecasts, etc.; text report data could include construction plans, approved contracts, etc.; and spatial geographic data could include project site coordinates, transmission line routes, topography, etc. After obtaining the power grid project data, the project attribute data to be extracted can be identified through keyword recognition, field recognition, etc., and then project attribute data such as project type, technical field, and voltage level can be extracted.

[0034] S200 selects multiple target review objects from a set of preset review objects that match the project attribute data.

[0035] The review object can be any of the review objects. In specific implementation, historical project review data for multiple review objects is acquired. Historical project attribute data is extracted from this data. A similarity analysis is performed between the project attribute data and the historical project attribute data of each review object. The similarity scores are then ranked, and a predetermined number of target review objects are selected from highest to lowest. To improve the accuracy of the review data, the predetermined number can be a value of at least 3. In other implementations, a scoring method can be used to evaluate the matching degree between historical project attribute data and project attribute data under different project attribute data dimensions. Then, a weighted summation method is used to determine a comprehensive score. Based on the comprehensive score, a predetermined number of target review objects are selected from highest to lowest.

[0036] S300: Obtain project review data for each target review object for power grid project data, extract review feature data from multiple project review data, and the review feature data characterizes the technical feasibility characteristics, execution risk characteristics, and compliance characteristics corresponding to the power grid project data.

[0037] The review feature data can include technical feasibility assessment data, execution risk assessment data, and compliance assessment data. Technical feasibility assessment data can include, but is not limited to, multi-dimensional data such as text and charts assessing the technical feasibility of power grid project data. Execution risk assessment data can include assessment data on the execution risks of power grid project data across multiple dimensions, including safety, economy, and environment. Compliance assessment data can include assessment data on whether power grid project data complies with national / industry regulations and standards.

[0038] In practice, after selecting multiple target review objects, the server can push power grid project data to these objects, allowing each object to submit its own project review data to the server via its terminal. The server can then perform semantic analysis on each project review data point to identify the text locations of technical feasibility assessment data, execution risk assessment data, and compliance assessment data. Based on these text locations, the server can extract the corresponding review feature data. Alternatively, keywords can be pre-defined for each of the technical feasibility assessment, execution risk assessment, and compliance assessment data. During the extraction process, these keywords are identified within the project review data, and their location information is used to extract the corresponding review feature data.

[0039] S400 determines the target project review data from multiple project review data based on review feature data.

[0040] In practice, for each dimension of the review feature data, the sentiment category of the review feature data under that dimension can be identified through preset keywords. Then, by statistically analyzing the proportion of different sentiment categories, the review feature data of the sentiment category with the largest proportion is integrated to obtain the review data of the target project. For example, the sentiment category can include categories that represent affirmation or negation, and the preset keywords can be words that represent affirmation or negation.

[0041] In the aforementioned method for generating review data for power grid projects, firstly, project attribute data is extracted from the power grid project data. Secondly, multiple target review objects matching the project attribute data are selected from a plurality of preset review objects. Thirdly, project review data for the power grid project data is obtained from the multiple target review objects. Review feature data is extracted from the multiple project review data. The review feature data characterizes the technical feasibility characteristics, execution risk characteristics, and compliance characteristics corresponding to the power grid project data. Finally, based on the review feature data, the target project review data is determined from the multiple project review data. Thus, by extracting review feature data to generate target project review data, the accuracy of review data generation is improved. Simultaneously, it facilitates adaptation to differentiated project review data scenarios and improves the standardization of review data generation.

[0042] In one exemplary embodiment, such as Figure 3 As shown, based on the review characteristic data, the target project review data is determined from multiple project review data, including S410 to S430, where:

[0043] S410: For each review feature data, identify the sentiment category of the review feature data.

[0044] The emotion categories can include positive, negative, and neutral.

[0045] In practical applications, a review data recognition model can be pre-built based on a pre-trained model (such as BERT) to extract review feature data and the corresponding sentiment categories from project review data. Specifically, project review data corresponding to historical power grid projects is obtained, labeled, and supervised labels containing review feature data and sentiment categories are generated. Then, the model parameters of the pre-built review data recognition model are adjusted using the labeled historical power grid project review data. During the adjustment process, the review data recognition model is trained to a convergent state, resulting in the trained review data recognition model.

[0046] In practice, project review data can be used as input to call a trained review data recognition model, which outputs sentiment categories for technical feasibility assessment data, execution risk assessment data, and compliance assessment data across three dimensions: technical feasibility features, risk assessment features, and compliance features. Alternatively, the review feature data can be extracted from the project review data, or it can be predicted by inputting the project review data into a trained review data recognition model.

[0047] In other implementations, an emotion category recognition model can be pre-built based on a pre-trained model. Then, project review data corresponding to historical power grid project data is obtained, and emotion categories are labeled on this historical data. Next, the labeled historical project review data is used to adjust the model parameters of the emotion category recognition model. During the adjustment process, the emotion category recognition model is trained to convergence, resulting in a trained emotion category recognition model. Specifically, review feature data is input into the trained emotion category recognition model to obtain the emotion category corresponding to the review feature data.

[0048] S420, determine the number of target review objects corresponding to each sentiment category in each review feature data, and determine the sentiment category corresponding to the maximum number as the target sentiment category of the review feature data.

[0049] In practice, for each dimension of review feature data, it can be identified whether different sentiment categories exist. If different sentiment categories exist, the review feature data for that dimension is marked as a review conflict. For each dimension of review feature data, the number of target review objects with positive and negative sentiment categories is counted separately. The sentiment category with the largest number is determined as the target sentiment category for the current dimension of review feature data. For example, for the technical feasibility assessment data dimension, suppose there are 8 target review objects. Under the current review feature data dimension, there are 6 target review objects with positive sentiment categories and 2 target review objects with negative sentiment categories. The target review objects with the positive sentiment category have the largest number of objects, so the target sentiment category for the technical feasibility assessment data dimension is determined to be positive.

[0050] S430, determine the matching degree of review data for multiple projects based on the number of target review objects corresponding to each target emotion category.

[0051] In practice, for each dimension (category) of review feature data, the ratio of the number of target review objects corresponding to the target sentiment category to the total number of target review objects is determined, thus obtaining the review data matching degree of the review feature data for that dimension. Continuing the example above, under the technical feasibility assessment data dimension, the proportion of target review objects for the target sentiment category is 0.75, meaning the review data matching degree between the technical feasibility assessment data of multiple target review objects is 0.75. When the number of target review objects corresponding to different sentiment categories is equal, this review feature data is marked as target review feature data. Subsequently, the data matching degree of this target review feature data can be directly determined as the target value (e.g., 0), triggering the subsequent re-review process. Specifically, this can involve increasing the number of target review objects, obtaining the project review data of the newly added target review objects, extracting the target review feature data from the project review data, identifying the sentiment category of the target review feature data, determining the target sentiment category corresponding to the updated target review feature data for multiple target review feature data, and then determining the review data matching degree of the target review feature data based on the number of target review objects corresponding to the updated target sentiment category.

[0052] S440, based on the matching degree of the review data, determines the target project review data from multiple project review data.

[0053] In practice, it is determined whether the matching degree of the review feature data of each dimension is higher than the preset review data matching degree threshold. If they are all higher than the preset review data matching degree threshold, the review feature data corresponding to the target sentiment category is integrated for the review feature data of each dimension to obtain the review data of the target project.

[0054] In this embodiment, by introducing emotion categories to determine the matching degree of review data, the consistency of project review data is quantified, which helps to improve the accuracy and standardization of review data.

[0055] In an exemplary embodiment, target project review data is determined from multiple project review data based on the review data matching degree, including S441 to S442, wherein:

[0056] S441, If ​​the matching degree of the review data is higher than the preset matching degree threshold of the review data, generate the target project review data based on the project review data of the target review object corresponding to the target sentiment category.

[0057] In practice, the matching degree of the review data corresponding to each dimension's review feature data is compared with a preset review data matching degree threshold. If the review feature data of each dimension is higher than the preset review data matching degree threshold, the matching degree of the review data corresponding to each dimension's review feature data can be further compared with a preset target review data matching degree threshold (target review data matching degree threshold > review data matching degree threshold). For example, the preset review data matching degree threshold is 0.6, and the preset target data matching degree threshold is 0.8. If the matching degree of each review data is higher than 0.6, it is further compared with the target data matching degree threshold of 0.8. If there is review feature data with a matching degree higher than the preset target data matching degree threshold, a pre-configured review result template will be called to generate the first review result text. Otherwise, the distribution index of the review feature data is statistically analyzed, and a second review result text is generated based on the distribution index and the review feature data. The distribution index can represent the quantity of different sentiment categories. For example, the first review result text includes "Project XX has high technical feasibility and the solution is stable and reliable". The second review result text includes "Regarding the compliance of Project XX: Two review subjects determined that it complies with the specifications, while one review subject pointed out that there are different risk assessment results on the 'XX Standard' clause. It is recommended to pay close attention to this issue and provide supplementary explanations." By traversing the review feature data across various dimensions, the target project review data containing the first or second review result texts corresponding to the review feature data for each dimension is obtained.

[0058] S442, if the matching degree of the review data is not higher than the preset matching degree threshold of the review data, obtain the target project review data of multiple target review objects. The target project review data is generated by the target review object for the power grid project data.

[0059] In practice, the matching degree of the review data corresponding to each review feature data is compared with the preset review data matching degree threshold (such as 60%). If the matching degree of each review data is not higher than the preset review data matching degree threshold, the re-review process is triggered. Here, the implementation steps of the re-review process in the above embodiment can be referred to. Specifically, it includes the complete process of adding review objects, obtaining the project review data of the review objects, extracting review feature data, identifying sentiment categories, determining the review data matching degree, and generating target project review data.

[0060] In this embodiment, the review data of different projects is standardized by the matching degree of the review data to obtain the review data of the target project, which helps to improve the accuracy of the review data.

[0061] In one exemplary embodiment, the project attribute data includes project type and technical field. Multiple target review objects matching the project attribute data are selected from a plurality of preset review objects, including:

[0062] Obtain the file information of multiple preset review subjects.

[0063] The archival information may include, but is not limited to, professional fields, experience levels, and historical power grid projects.

[0064] In practice, the file information of each review object can be integrated in advance to build a review object file database. By calling the database, the file information of multiple preset review objects can be obtained.

[0065] Based on multiple archival information, project types, and technical fields, multiple target review objects are selected from multiple preset review objects.

[0066] In practical applications, a pre-built model for matching review objects can be constructed: Historical matching training data is acquired, including project attribute data of historical power grid projects, corresponding review object profile information, and feedback scoring data from object reviews. Then, the project attributes and object profile information are represented as feature vectors to construct high-dimensional feature vectors. A supervised learning algorithm is used, with the feedback scoring data as the supervisory signal, to train the model. Through training, the model learns the complex mapping relationship between project features and object attributes, ultimately outputting a model capable of scoring the matching degree between a given project and any object.

[0067] In practice, a review object matching model is used to match project type, technical field, and voltage level with archival information in the review object case library to obtain a matching score. Based on the matching score, at least three target review objects are selected from the preset review objects to form a review object group. The project data is then pushed to a dedicated review platform on the review object's end via a data transmission channel.

[0068] In this embodiment, data matching is used to automatically match review objects, which helps to improve review efficiency and reliability.

[0069] In one exemplary embodiment, such as Figure 4 As shown, after determining the target project review data based on the review data matching degree, the method also includes S500:

[0070] S500 integrates project data, project attribute data, review feature data, sentiment category and target project review data to obtain review case data.

[0071] In practice, after obtaining the target project review data, the project data, project attribute data, review characteristic data, and the sentiment category of the object in this review process are linked and stored with the target project review data to obtain review case data and build a review case library, so as to provide historical similar cases for retrieval and recommendation for new power grid projects to be reviewed.

[0072] After obtaining the project attribute data extracted from the power grid project data, the method also includes:

[0073] S320 performs text similarity analysis and semantic similarity analysis on the power grid project data in the power grid project data and the review case data, and obtains the text similarity analysis results and semantic similarity analysis results.

[0074] S340. Based on the text similarity analysis results and semantic similarity analysis results, determine the target project review data corresponding to the project data.

[0075] In practice, the first step is to calculate the similarity (e.g., cosine similarity) between the project attribute data of the power grid project to be reviewed and the project attribute data of historical power grid project data in the review case library. This yields a text similarity analysis result. Based on the text similarity analysis result, a candidate case set is selected from the review case library. For example, candidate cases with text similarity scores higher than a preset text similarity threshold are selected.

[0076] Subsequently, using a pre-trained natural language processing model, the text report data of the new power grid projects to be reviewed and the text report data of candidate cases are converted into semantic vectors, respectively. The similarity between the semantic vectors is determined by a similarity algorithm to obtain the semantic similarity analysis results.

[0077] Finally, the text similarity analysis results and semantic similarity analysis results can be weighted and fused to obtain a comprehensive similarity score. The comprehensive similarity scores can be sorted from high to low to select a preset number of target review case data. Subsequently, the target project review data of the target review case data can be determined as the target project review data of the power grid project data to be reviewed, or the target project review data of the selected target review case data can be pushed to the target review object as an auxiliary decision-making reference.

[0078] In this embodiment, by constructing a case library, a closed-loop project review data generation process is formed, which improves the efficiency of review data generation.

[0079] In an exemplary embodiment, before obtaining the project attribute data extracted from the power grid project data, the method further includes: performing data preprocessing on the power grid project data, the preprocessing including at least one of missing value imputation, word segmentation, stop word removal, data format conversion, and standardization.

[0080] In practice, for structured numerical data, interpolation can be used to fill in missing values. Specifically, check if there are any missing items in the data (such as missing investment amounts, capacity, or other key values). For fields with missing values, use interpolation to fill them in. Combine this with the 3σ principle or box plot method to identify and process outlier data points, ensuring the accuracy and consistency of the numerical data.

[0081] Subsequently, the text report data was processed through word segmentation and stop word removal, and named entity recognition technology was used to extract key entity information related to the power grid project. Specifically, the feasibility study report, design plan, and other documents were lexicalized using a Chinese word segmentation tool, and meaningless stop words were removed. Then, a named entity recognition model based on the BERT-BiLSTM-CRF architecture was used to accurately extract key entity information such as key technical parameters, equipment models, implementation standards, and responsible units.

[0082] Next, coordinate system and format conversion are performed on the spatial geographic data; specifically, GIS data from different sources are uniformly converted to the national geodetic coordinate system, and various spatial data formats are standardized to preset formats (such as GeoJSON) to ensure the consistency of spatial benchmarks.

[0083] Finally, the processed data is mapped to unified standardized fields to generate standardized power grid project data.

[0084] In this embodiment, preprocessing the project data helps improve the reliability of data review.

[0085] To provide a clearer explanation of the method for generating review data for power grid projects provided in this application, a specific embodiment is described below, which includes the following steps:

[0086] S1, retrieve project attribute data extracted from power grid project data.

[0087] S2, perform data preprocessing on power grid project data, including at least one of missing value imputation, word segmentation, stop word removal, data format conversion, and standardization.

[0088] S3: Obtain the file information of multiple preset review objects, and select multiple target review objects from the multiple preset review objects based on the multiple file information, project type and technical field.

[0089] S4 receives project review data for each target review object for power grid project data, extracts review feature data from multiple project review data, and the review feature data characterizes the technical feasibility characteristics, execution risk characteristics and compliance characteristics corresponding to the power grid project data.

[0090] S5. For each review feature data, identify the sentiment category of the review feature data, determine the number of target review objects corresponding to each sentiment category in each review feature data, and take the sentiment category corresponding to the maximum number as the target sentiment category of the review feature data. Based on the number of target review objects corresponding to each target sentiment category, determine the review data matching degree of each review feature data.

[0091] S6. When the matching degree of the review data is higher than the preset first review data matching degree threshold, the target project review data is generated based on the project review data of the target review object corresponding to the target sentiment category. When the matching degree of the review data is not higher than the preset review data matching degree threshold, the target project review data of multiple target review objects is obtained. The target project review data is generated by the target review object for the power grid project data.

[0092] S7 integrates power grid project data, project attribute data, review characteristic data, sentiment category and target project review data to obtain review case data.

[0093] In other embodiments, after executing S4, the process can jump to execute S7, and then execute S8: perform text similarity analysis and semantic similarity analysis on the power grid project data in the power grid project data and review case data, obtain the text similarity analysis results and semantic similarity analysis results, and determine the target project review data corresponding to the power grid project data based on the text similarity analysis results and semantic similarity analysis results.

[0094] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0095] In one exemplary embodiment, such as Figure 5 As shown, a power grid project review data generation device 600 is provided, including: a data acquisition module 610, a review object determination module 620, a review feature data extraction module 630, and a target review data generation module 640, wherein:

[0096] The data acquisition module 610 is used to acquire project data of the power grid project to be reviewed;

[0097] The review object determination module 620 is used to extract project attribute data from project data and select multiple target review objects that match the project attribute data from multiple preset review objects.

[0098] The review feature data extraction module 630 is used to receive project review data from multiple target review objects, extract review feature data from multiple project review data, and characterize the technical feasibility, execution risks and compliance of power grid projects.

[0099] The target review data generation module 640 is used to determine the review data matching degree of multiple project review data based on review feature data, and to determine the target project review data based on the review data matching degree.

[0100] In an exemplary embodiment, the target review data generation module 640 is further configured to: identify the sentiment category of each review feature data; determine the number of target review objects corresponding to each sentiment category in each review feature data, and take the sentiment category corresponding to the maximum number as the target sentiment category of the review feature data; determine the review data matching degree of each review feature data according to the number of target review objects corresponding to each target sentiment category; and determine the target project review data from multiple project review data according to the review data matching degree.

[0101] In an exemplary embodiment, the target review data generation module 640 is further configured to generate target project review data based on the project review data of the target review object corresponding to the target sentiment category when the review data matching degree is higher than the preset first review data matching degree threshold; and to obtain target project review data of multiple target review objects when the review data matching degree is not higher than the preset review data matching degree threshold, wherein the target project review data is generated by the target review object for the power grid project data.

[0102] In an exemplary embodiment, the review object determination module 620 is further configured to obtain file information of multiple preset review objects; and select multiple target review objects from the multiple preset review objects based on the multiple file information, project type and technical field.

[0103] In an exemplary embodiment, the power grid project review data generation device 600 further includes a case retrieval module 650, which integrates project data, project attribute data, review feature data, sentiment category, and target project review data to obtain review case data; performs text similarity analysis and semantic similarity analysis on the power grid project data and the power grid project data in the review case data to obtain text similarity analysis results and semantic similarity analysis results; and determines the target project review data corresponding to the power grid project data based on the text similarity analysis results and semantic similarity analysis results.

[0104] In an exemplary embodiment, the power grid project review data generation device 600 further includes a data preprocessing module 660 for preprocessing the power grid project data. The preprocessing includes at least one of missing value filling, word segmentation, stop word removal, data format conversion, and standardization.

[0105] Each module in the aforementioned power grid project review data generation device 600 can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0106] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 6 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores data. The I / O interfaces allow the processor to exchange information with external devices. The communication interface allows communication with external terminals via a network connection. When executed by the processor, the computer program implements a method for generating review data for a power grid project.

[0107] Those skilled in the art will understand that Figure 6The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0108] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in any of the above embodiments of the power grid project review data generation method.

[0109] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in any of the above embodiments of the power grid project review data generation method.

[0110] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in any of the above embodiments of the power grid project review data generation method.

[0111] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0112] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0113] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0114] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for generating review data for power grid projects, characterized in that, The method includes: Obtain project attribute data extracted from power grid project data; Select multiple target review objects that match the project attribute data from a variety of preset review objects; Obtain project review data for each target review object in relation to the power grid project data, and extract review feature data from multiple project review data. The review feature data characterizes the technical feasibility characteristics, execution risk characteristics, and compliance characteristics corresponding to the power grid project data. Based on the review feature data, the target project review data is determined from multiple project review data.

2. The method according to claim 1, characterized in that, The step of determining the target project review data from multiple project review data based on the review feature data includes: For each of the aforementioned review feature data, the sentiment category of the review feature data is identified respectively; Determine the number of target review objects corresponding to each sentiment category in each review feature data, and take the sentiment category corresponding to the maximum number as the target sentiment category of the review feature data; The review data matching degree of each review feature data is determined based on the number of target review objects corresponding to each of the target emotion categories. Based on the matching degree of the review data, the target project review data is determined from multiple project review data.

3. The method according to claim 2, characterized in that, The step of determining the target project review data from multiple project review data based on the review data matching degree includes: If the matching degree of the review data is higher than the preset first review data matching degree threshold, target project review data is generated based on the project review data of the target review object corresponding to the target emotion category. If the matching degree of the review data is not higher than the preset matching degree threshold of the review data, the target project review data of multiple target review objects is obtained, and the target project review data is generated by the target review object for the power grid project data.

4. The method according to claim 1, characterized in that, The project attribute data includes project type and technical field; the step of selecting multiple target review objects that match the project attribute data from multiple preset review objects includes: Obtain the file information of multiple preset review subjects; Based on the multiple types of file information, project types, and technical fields, multiple target review objects are selected from the multiple preset review objects.

5. The method according to claim 3, characterized in that, After determining the target project review data based on the review data matching degree, the method further includes: By integrating the power grid project data, project attribute data, review characteristic data, sentiment category, and target project review data, review case data is obtained. After obtaining the project attribute data extracted from the power grid project data, the method further includes: Text similarity analysis and semantic similarity analysis are performed on the power grid project data and the power grid project data in the review case data to obtain the text similarity analysis results and semantic similarity analysis results; Based on the text similarity analysis results and the semantic similarity analysis results, the target project review data corresponding to the power grid project data is determined.

6. The method according to any one of claims 1 to 5, characterized in that, Before obtaining the project attribute data extracted from the power grid project data, the method further includes: The power grid project data is preprocessed, including at least one of missing value filling, word segmentation, stop word removal, data format conversion, and standardization.

7. A device for generating review data for power grid projects, characterized in that, The device includes: The data acquisition module is used to acquire project attribute data extracted from power grid project data; The review object determination module is used to select multiple target review objects that match the project attribute data from a plurality of preset review objects; The review feature data extraction module is used to obtain project review data for each target review object for the power grid project data, and extract review feature data from multiple project review data. The review feature data characterizes the technical feasibility features, execution risk features and compliance features corresponding to the power grid project data. The target review data generation module is used to determine the target project review data from multiple project review data based on the review feature data.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images