Multi-level and multi-type planning content difference identification and conflict elimination method

Abstract

The present invention discloses a multi-level and multi-type planning content difference identification and conflict elimination method, comprising: obtaining planning text content, and converting the obtained planning text content into planning text data of a specified format, and performing knowledge annotation processing on the converted planning text data; extracting the annotated planning text data, using a knowledge extraction method, segmenting the planning text semantics, and generating planning structured data; based on the generated planning structured data, organizing the relationship between planning elements and / or indicators according to the entity-relationship-entity triple model for planning knowledge, and constructing a planning knowledge graph; based on the constructed planning knowledge graph, utilizing the linkage relationship between planning semantics to identify planning conflicts, and outputting the identified planning conflict results. The present invention improves the accuracy and reliability of the planning collaboration process, effectively solves the problem of multi-level and multi-type planning content difference identification and conflict elimination, and can improve the collaboration efficiency of different planning types.

Description

Technical Field

[0001] The present invention relates to the technical field of planning data processing, and in particular to a multi-level and multi-type planning content difference identification and conflict elimination method, an intelligent terminal and a storage medium. Background Art

[0002] Spatial planning serves as a guide for spatial development, a blueprint for sustainable development, and the foundation for all types of development, protection, and construction activities. Under the new circumstances, how to coordinate various related plans within the planning process is a pressing issue that demands consideration and resolution. Spatial planning is a complex undertaking, involving a range of stakeholders, including urban planners, city managers, architects, and developers. These stakeholders require collaboration within the spatial planning process. Therefore, from a rational and systematic perspective, it is necessary to establish a common "planning context" or "planning elements" for the coordination of these plans, enabling effective communication within the planning process.

[0003] In existing technologies, planning documents come from different sources, such as government websites, natural resources bureaus, and development planning bureaus. Furthermore, there are different types of plans, including master plans, detailed plans, and zoning plans. Due to the differences in planning document sources and types, conflicts between planning elements and indicators can easily arise due to a lack of communication and coordination. Specifically, existing technologies are not very accurate in identifying differences in multi-level and multi-type planning content during the planning process. Some conflicts require manual resolution, which fails to improve the efficiency of collaboration across different planning types.

[0004] Therefore, existing navigation technology needs to be improved and developed. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to address the above-mentioned shortcomings of the existing technology and provide a multi-level and multi-type planning content difference identification and conflict resolution method, intelligent terminal, and storage medium. The present invention obtains planning elements from different types of planning texts, collaboratively constructs a planning information knowledge base, and uses the knowledge graph as the basis for identifying and resolving planning content differences. The present invention improves the accuracy and reliability of the planning collaboration process, effectively solves the problem of identifying and resolving conflicts in multi-level and multi-type planning content, and can improve the collaborative efficiency of different planning types.

[0006] The technical solutions adopted by the present invention to solve the problem are as follows:

[0007] A multi-level and multi-type planning content difference identification and conflict elimination method, wherein the method comprises:

[0008] Acquire planning text content, convert the acquired planning text content into planning text data in a specified format, and perform knowledge annotation processing on the converted planning text data;

[0009] Extract the annotated planning text data, use knowledge extraction to segment the planning text semantics, and generate planning structured data;

[0010] Based on the generated planning structured data, the relationships between planning elements and / or indicators are organized into planning knowledge according to the entity-relationship-entity triple model to construct a planning knowledge graph;

[0011] Based on the constructed planning knowledge graph, the linkage relationship between planning semantics is utilized to identify planning conflicts and output the identified planning conflict results.

[0012] In the multi-level and multi-type planning content difference identification and conflict elimination method, the step of obtaining planning text content and converting the obtained planning text content into planning text data in a specified format includes:

[0013] Obtain the content of each planning text to form a multi-source planning text original data set;

[0014] The planning text data in the multi-source planning text original data set is preprocessed by converting the original text into TXT format to obtain planning text data in TXT format.

[0015] The multi-level and multi-type planning content difference identification and conflict elimination method, wherein the step of performing knowledge annotation processing on the converted planning text data includes:

[0016] Get the converted planning text data in the specified format;

[0017] Generating original text sequences and annotating text corpus for the planned text data in the specified format;

[0018] The BIO annotation method is used to perform entity category annotation and text relationship annotation respectively, and a planning text entity sequence based on the BIO annotation and a planning text relationship sequence based on the BIO annotation are obtained respectively, forming the annotated planning text data.

[0019] The multi-level and multi-type planning content difference identification and conflict elimination method, wherein the steps of extracting the annotated planning text data, using a knowledge extraction method, segmenting the planning text semantics, and generating planning structured data include:

[0020] The BERT-BiLSTM-CRF model is used to extract knowledge from the annotated planning text data;

[0021] Obtain a planning text entity sequence of the planning text data to obtain an entity training data set; obtain a planning text relationship sequence of the planning text data to obtain a relationship training data set;

[0022] Input the entity training dataset into the BERT-BiLSTM-CRF model for entity recognition, save the optimal model, and predict the entity results;

[0023] Input the relationship training data set into the BERT-BiLSTM-CRF model for relationship extraction, save the optimal model, and predict the relationship results;

[0024] The predicted entity results and predicted entity results are used to generate planning structured data.

[0025] The multi-level and multi-type planning content difference identification and conflict elimination method, wherein the steps of organizing planning knowledge based on the generated planning structured data and the relationships between planning elements and / or indicators according to the entity-relationship-entity triple model to construct a planning knowledge graph include:

[0026] Based on the generated planning structured data, the relationships between planning elements and / or indicators in the planning structured data are organized according to entity-relationship-entity triples;

[0027] Using the Neo4j graph database, seven groups of graph node labels are set; the seven groups of graph node labels include: a PName node label representing the specific name of the plan, a PPurpose node label representing the respective planning objectives contained in the plan, a PBasis node label representing the specific planning basis, a PRange node label representing the administrative scope involved in the plan, a PIndex node label representing the specific indicators involved in the planning content, a PTime node label representing the expected completion target period of the compiled statutory plan, and a PAttributes node label representing the content of multi-attribute elements;

[0028] Based on the specific entities in the set node labels and the relationships between labels, the relationships between the PName node label and the other six groups of node labels are constructed, as well as the relationships between the entities contained in the labels in the PName node label, PIndex node label, and PAttributes node label.

[0029] The constructed node labels and the relationship information between node labels are presented in the form of entity-relationship-entity triples to complete the construction of the planning knowledge graph.

[0030] The multi-level and multi-type planning content difference identification and conflict elimination method, wherein the steps of identifying planning conflicts based on the constructed planning knowledge graph and utilizing the linkage relationship between planning semantics and outputting the identified planning conflict results include:

[0031] Based on the constructed planning knowledge graph, the linkage relationship between planning semantics is utilized to obtain query planning types and indicators, automatically identify differences in planning content, identify planning conflicts, and issue reminders.

[0032] A multi-level and multi-type planning content difference identification and conflict elimination device, comprising:

[0033] An acquisition conversion module is used to acquire planning text content, convert the acquired planning text content into planning text data in a specified format, and perform knowledge annotation processing on the converted planning text data;

[0034] The extraction and segmentation module is used to extract the annotated planning text data, use knowledge extraction to segment the planning text semantics, and generate planning structured data;

[0035] A knowledge graph construction module is used to organize planning knowledge based on the generated planning structured data by organizing the relationships between planning elements and / or indicators according to the entity-relationship-entity triple model to construct a planning knowledge graph;

[0036] The conflict identification module is used to identify planning conflicts based on the constructed planning knowledge graph and utilize the linkage relationship between planning semantics, and output the identified planning conflict results.

[0037] The multi-level and multi-type planning content difference identification and conflict elimination device, wherein the acquisition and conversion module includes:

[0038] An acquisition unit, used to acquire the content of each planning text to form a multi-source planning text original data set;

[0039] The conversion processing unit is used to perform a pre-processing of the original text conversion of the planning text data in the multi-source planning text original data set into the TXT format to obtain the planning text data in the TXT format;

[0040] The tagging unit is used to generate original text sequences and annotate text corpora for the planning text data in the specified format; adopt the BIO tagging method to perform entity category tagging and text relationship tagging respectively, obtain the planning text entity sequence based on BIO tagging and the planning text relationship sequence based on BIO tagging respectively, and form the planning text data processed by the tagging.

[0041] An intelligent terminal includes a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by one or more processors, including the method for executing any one of the methods described above.

[0042] A non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to perform any of the methods described above.

[0043] Beneficial effects of the present invention: The present invention provides a multi-level and multi-type planning content difference identification and conflict elimination method, intelligent terminal and storage medium, which involves the acquisition of planning knowledge, the establishment of knowledge relationships and the elimination of planning conflicts. The method of the present invention comprises: first, obtaining the planning text content and pre-processing the planning text content, including text collection and knowledge annotation; second, extracting the planning knowledge. For the existing national land space planning text, the BERT-BiLSTM-CRF model is used to extract the planning name, planning basis, planning indicators and planning-related attribute information in the planning text, and integrate the planning knowledge. Third, based on the knowledge extraction results, the relationship between planning elements and / or indicators is organized according to the "entity-relationship-entity" triple model to construct a planning knowledge graph. Finally, based on the knowledge graph, the linkage relationship between planning semantics is used to identify planning conflicts. The present invention effectively solves the problem of multi-level and multi-type planning content difference identification and conflict elimination, and can improve the collaborative efficiency of different planning types. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0045] Figure 1 It is a flowchart of a multi-level and multi-type planning content difference identification and conflict elimination method provided by an embodiment of the present invention.

[0046] Figure 2 It is a text preprocessing diagram of the multi-level and multi-type planning content difference identification and conflict elimination method provided by an embodiment of the present invention.

[0047] Figure 3 This is a schematic diagram of the knowledge extraction process of the multi-level and multi-type planning content difference identification and conflict elimination method provided by an embodiment of the present invention.

[0048] Figure 4 It is a schematic diagram of the entity and relationship classification extraction process of the multi-level and multi-type planning content difference identification and conflict elimination method provided by an embodiment of the present invention.

[0049] Figure 5 This is a principle block diagram of a multi-level and multi-type planning content difference identification and conflict elimination device provided by an embodiment of the present invention.

[0050] Figure 6 This is a block diagram of the internal structure of the smart terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of the present invention more clear and distinct, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0052] It should be noted that if the embodiments of the present invention involve directional indications (such as up, down, left, right, front, back, etc.), the directional indications are only used to explain the relative position relationship, movement status, etc. between the various components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly.

[0053] Spatial planning serves as a guide for spatial development, a blueprint for sustainable development, and the foundation for all types of development, protection, and construction activities. Under the new circumstances, how to coordinate various related plans within the planning process is a pressing issue that demands consideration and resolution. Spatial planning is a complex undertaking, involving a range of stakeholders, including urban planners, city managers, architects, and developers. These stakeholders require collaboration within the spatial planning process. Therefore, from a rational and systematic perspective, it is necessary to establish a common "planning context" or "planning elements" for the coordination of these plans, enabling effective communication within the planning process.

[0054] In existing technologies, planning documents come from different sources, such as government websites, natural resources bureaus, and development planning bureaus. Furthermore, there are different types of plans, including master plans, detailed plans, and zoning plans. Due to the differences in planning document sources and types, conflicts between planning elements and indicators can easily arise due to a lack of communication and coordination. Specifically, existing technologies are not very accurate in identifying differences in multi-level and multi-type planning content during the planning process. Some conflicts require manual resolution, which fails to improve the efficiency of collaboration across different planning types.

[0055] As a knowledge graph of a large-scale semantic network, we can leverage the advantages of knowledge graphs in logical expression to explore how to solve the coordination problems of various types of planning.

[0056] Therefore, the present invention introduces knowledge graph technology, and proposes a new method for identifying and eliminating differences in planning content for conflicts between different levels and types of planning content; it also proposes a method for identifying and eliminating differences in multi-level and multi-type planning content. By integrating the planning elements of different types of planning texts, a planning information knowledge base is collaboratively constructed, and planning content differences are identified and conflicts are eliminated based on the knowledge graph. While improving the accuracy and reliability of the planning collaboration process, the present invention also helps the planning subject enhance logical thinking, which is of great significance to my country's land space planning.

[0057] Exemplary Methods

[0058] like Figure 1 As shown in , an embodiment of the present invention provides a multi-level and multi-type planning content difference identification and conflict elimination method, which can be applied to smart terminals. In an embodiment of the present invention, the method includes the following steps:

[0059] Step S100: Acquire planning text content, convert the acquired planning text content into planning text data in a specified format, and perform knowledge annotation processing on the converted planning text data;

[0060] In an embodiment of the present invention, the content of each planning text is first obtained to form a multi-source planning text original data set; then the original text of the planning text data in the multi-source planning text original data set is pre-processed by converting it into txt format to obtain planning text data in txt format.

[0061] During specific implementation, the present invention obtains the converted planning text data in a specified format; performs original text sequence generation and text corpus annotation on the planning text data in the specified format (txt format); for example, adopts the BIO annotation method to perform entity category annotation and text relationship annotation respectively, and obtains the planning text entity sequence based on BIO annotation and the planning text relationship sequence based on BIO annotation respectively, forming the annotated planning text data.

[0062] The BIO annotation is as follows: each element is labeled "BX", "IX", or "O". "BX" indicates that the segment where the element is located is of type X and the element is at the beginning of the segment. "IX" indicates that the segment where the element is located is of type X and the element is in the middle of the segment. "O" indicates that the element is not of any type.

[0063] Step S200: extract the annotated planning text data, use knowledge extraction to segment the planning text semantics, and generate planning structured data;

[0064] Among them, when extracting the annotated planning text data, the present invention needs to obtain the spatial planning text and adopt the BERT-BiLSTM-CRF model to extract the planning name, planning basis, planning indicators and planning-related attribute information in the planning text to integrate planning knowledge;

[0065] Specifically, the BERT-BiLSTM-CRF model is used to extract knowledge from the annotated planning text data;

[0066] Obtain a planning text entity sequence of the planning text data to obtain an entity training data set; obtain a planning text relationship sequence of the planning text data to obtain a relationship training data set;

[0067] Input the entity training dataset into the BERT-BiLSTM-CRF model for entity recognition, save the optimal model, and predict the entity results;

[0068] Input the relationship training data set into the BERT-BiLSTM-CRF model for relationship extraction, save the optimal model, and predict the relationship results;

[0069] The predicted entity results and predicted entity results are used to generate planning structured data.

[0070] Step S300: Based on the generated planning structured data, the relationships between planning elements and / or indicators are organized into planning knowledge according to the entity-relationship-entity triple model to construct a planning knowledge graph;

[0071] In the embodiment of the present invention, based on the generated planning structured data, the relationships between planning elements and / or indicators in the planning structured data are sorted according to entity-relationship-entity triples;

[0072] Using the Neo4j graph database, seven groups of graph node labels are set; the seven groups of graph node labels include: the PName node label representing the specific name of the plan, the PPurpose node label representing the respective planning objectives contained in the plan, the PBasis node label representing the specific planning basis, the PRange node label representing the administrative scope involved in the plan, the PIndex node label representing the specific indicators involved in the planning content, the PTime node label representing the expected completion target period of the compiled statutory plan, and the PAttributes node label representing the content of multiple attribute elements.

[0073] Based on the specific entities in the set node labels and the relationships between labels, the relationships between the PName node label and the other six groups of node labels are constructed, as well as the relationships between the entities contained in the labels in the PName node label, PIndex node label, and PAttributes node label.

[0074] The constructed node labels and the relationship information between node labels are presented in the form of entity-relationship-entity triples to complete the construction of the planning knowledge graph.

[0075] Step S400: Based on the constructed planning knowledge graph, the linkage relationship between planning semantics is utilized to automatically identify planning conflicts and output the identification results.

[0076] Optionally, the present invention is specifically described as follows regarding step S100 of acquiring planning text content, converting the acquired planning text content into planning text data in a specified format, and performing knowledge annotation processing on the converted planning text data:

[0077] Step S101: Acquire the contents of each planning text to form a multi-source planning text original data set;

[0078] Step S102: performing pre-processing of converting the original text into TXT format on the planning text data in the multi-source planning text original data set to obtain planning text data in TXT format.

[0079] In the embodiment of the present invention, the content of each planning text is obtained to form a multi-source planning text original data set, and then the multi-source planning text original data set is converted into txt format using python tools such as pdfminer (PDF information extraction) and python-docx (WORD and other information extraction). Figure 2 The figure shows a schematic diagram of text preprocessing, in which the text is preprocessed into TXT format. Python is a general-purpose computer programming language.

[0080] Step S103: generating an original text sequence and annotating the text corpus for the planning text data in the specified format; using the BIO annotation method to perform entity category annotation and text relationship annotation, respectively, to obtain a planning text entity sequence based on the BIO annotation and a planning text relationship sequence based on the BIO annotation, thereby forming the annotated planning text data;

[0081] The BIO notation is to mark each element as "BX", "IX", or "O". "BX" indicates that the segment containing the element is of type X and the element is at the beginning of the segment. "IX" indicates that the segment containing the element is of type X and the element is in the middle of the segment. "O" indicates that the element does not belong to any type.

[0082] The BIO annotation method is used to address sequence labeling in NLP (natural language processing) processes. It is feasible to use it to annotate planning texts with entity categories. The BIO annotation method divides an entity (Single) into its beginning (Begin), its interior (Inside), and its non-entity portion (Outside). To address the issues that need to be addressed in planning collaboration, entity definitions are developed for planning elements such as the planning type, planning period, implementation scope, planning objectives, planning indicators, planning basis, and other planning attributes. Common relationships within the text primarily include "is," "located in," "amongst others," "occupy," "adjust," "increase," and "decrease." The BIO annotation method generates labels for each entity and outputs output that meets the input corpus standards of the word vector generation layer. The specific entity label annotation definitions are shown in Table 1.

[0083] Table 1 Planning text entity type definition

[0084]

[0085] The entity relationship annotation definition is shown in Table 2.

[0086] Table 2 Planning text relationship definition

[0087]

[0088] In this embodiment of the present invention, using Python libraries such as re and python-docx, we generate a raw text sequence and annotate the planned text data, obtaining a BIO-annotated planned text data sequence. The BIO-annotated planned text is shown in Table 3, and the relationships between the BIO-annotated planned text are shown in Table 4.

[0089] Table 3 Examples of planning text entity sequences based on BIO annotation (partial)

[0090]

[0091]

[0092] Table 4 Examples of planning text relationship sequences based on BIO annotation (partial)

[0093]

[0094] In the embodiment of the present invention, the steps of extracting the annotated planning text data in step S200, using a knowledge extraction method to segment the planning text semantics, and generating planning structured data specifically include:

[0095] Step S201: extract knowledge from the annotated planning text data using the BERT-BiLSTM-CRF model;

[0096] Step S202: Acquire a planning text entity sequence of the planning text data to obtain an entity training data set; acquire a planning text relationship sequence of the planning text data to obtain a relationship training data set;

[0097] Step S203: Input the entity training dataset into the BERT-BiLSTM-CRF model for entity recognition, save the optimal model, and predict the entity results;

[0098] Step S204: Input the relationship training data set into the BERT-BiLSTM-CRF model to extract the relationship, save the optimal model, and predict the relationship results;

[0099] Step S205: Generate planning structured data from the predicted entity results and the predicted entity results.

[0100] Specifically, the present invention uses knowledge extraction based on the BERT-BiLSTM-CRF model. The BERT-BiLSTM-CRF model is an overall model architecture based on a transformer-based bidirectional encoder representation, a bidirectional long short-term memory neural network, and a conditional random field.

[0101] Specifically, the overall model architecture of the knowledge extraction link of the present invention is composed of three modules: BERT, BiLSTM and CRF. The specific process is as follows: Figure 3 As shown in the figure, the spatial planning document for a certain city, XX, is input into the input layer. Word embedding, including token embedding, fragment embedding, and position embedding, is then performed, followed by BERT encoding. In this embodiment of the present invention, the BERT model, used as a pre-trained model, obtains a vector representation for each word by inputting the planning document. The word vectors are then input into the BiLSTM module, encoded, and finally passed through the CRF layer to output the maximum probability result.

[0102] The BERT model is a context-based model that can generate different embedding models according to the context of words, and understand the contextual meaning of each word based on every word in a sentence in relation to all other words. In Chinese, there are also cases of polysemy. For example, the word "burden" can refer to a mental burden or the punchline in a prepared joke, which is the essence of the entire language expression content. Therefore, when generating word embeddings, it is necessary to generate word embeddings according to the context. To increase the semantic representation of word vectors, by broadening the embedding ability of word vectors and adding multiple word embeddings, the relationship features between different characters are fully learned.

[0103] The Bi-directional Long Short-Term Memory (BiLSTM) neural network model is a group of LSTM models with two directions. In natural language processing, it is mainly used to process contextual text information. LSTM uses a one-way order from front to back and cannot encode in the reverse direction. When facing classification tasks with different levels of meaning in text, the words or characters between the front and back contexts have obvious correlations, and BiLSTM can more effectively capture the dependencies between contexts.

[0104] CRF (in natural language processing) includes two elements, namely the position element and the corresponding value element. When each position is assigned the corresponding element value according to a certain distribution, the set composed of the position element and the value element is a random field.

[0105] In the embodiments of the present invention, the process of entity and relationship classification extraction is as Figure 4 shown: Figure 4 This is the process of entity and relationship classification extraction in the method embodiments of the present invention. As Figure 4 shown, the embodiments of the present invention based on the BERT-BiLSTM-CRF model are divided into the following steps:

[0106] Step (1): Input the labeled entity and relationship planning corpus into the BERT-BiLSTM-CRF model.

[0107] As Figure 4 shown, input the labeled entity training dataset into the BERT-BiLSTM-CRF model for entity recognition, and input the labeled relationship training dataset into the BERT-BiLSTM-CRF model for relationship extraction.

[0108] Step (2): First, perform preprocessing through the BERT model to obtain word vectors and send them into the BiLSTM layer.

[0109] The entity training data set and relationship training data set to be input are preprocessed by the BERT model in different channels to obtain word vectors, which are then sent to the BiLSTM layer.

[0110] Step (3) uses the BiLSTM model to obtain the processed word vector results.

[0111] Step (4) classifies entities and relations based on the probability distribution of label results.

[0112] After encoding the word vector results of the entity training dataset to be processed, the CRF (natural language processing) module outputs the result with the maximum probability, saves the optimal model, and predicts the entity result;

[0113] After encoding the word vector results of the processed relationship training data set, the CRF (natural language processing) module outputs the result with the maximum probability, saves the optimal model, and predicts the relationship result.

[0114] Furthermore, in an embodiment of the present invention, regarding step S300, based on the generated planning structured data, the relationships between planning elements and / or indicators are organized into planning knowledge according to the entity-relationship-entity triple model, and the construction of the planning knowledge graph specifically includes:

[0115] Step S301: Based on the generated planning structured data, the relationships between planning elements and / or indicators in the planning structured data are sorted according to entity-relationship-entity triples;

[0116] Step S302: Using the Neo4j graph database, set seven groups of graph node labels; the seven groups of graph node labels include: a PName node label representing the specific name of the plan, a PPurpose node label representing the respective planning objectives contained in the plan, a PBasis node label representing the specific planning basis, a PRange node label representing the administrative scope involved in the plan, a PIndex node label representing the specific indicators involved in the plan content, a PTime node label representing the expected completion target period of the compiled statutory plan, and a PAttributes node label representing the content of multiple attribute elements;

[0117] Step S303: Based on the specific entities in the set node tags and the relationships between tags, the relationships between the PName node tag and the other six groups of node tags are established, as well as the relationships between the entities contained in the tags in the PName node tag, the PIndex node tag, and the PAttributes node tag.

[0118] Step S304: Present the constructed node labels and the relationship information between the node labels in the form of entity-relationship-entity triples to complete the construction of the planning knowledge graph.

[0119] Specifically, in the embodiments of the present invention, graph-based storage offers flexibility. The present invention utilizes the Neo4j graph database, based on triple RDF structures ("entity-relationship-entity"), to store and display knowledge graphs. Neo4j is a high-performance, NoSQL graph database that stores structured data on the network rather than in tables.

[0120] In this embodiment of the present invention, based on the generated planning structured data, the relationships between planning elements and / or indicators in the planning structured data are organized according to "entity-relationship-entity" triples, and seven sets of graph node labels are assigned. In this embodiment of the present invention, the relationships between planning elements and / or indicators include: between planning elements and indicators, between planning elements and planning elements, and between indicators and indicators.

[0121] In specific implementation, the seven groups of graph node labels set in the embodiment of the present invention include: PName node label (representing the specific name of the plan), PPurpose node label (representing the respective planning objectives contained in the plan), PBasis node label (representing the specific planning basis), PRange node label (representing the administrative scope involved in the plan), PIndex node label (representing the specific indicator items involved in the planning content, mainly including total indicators, incremental indicators, efficiency indicators and other indicator items according to different planning texts), PTime node label (representing the expected deadline for the compilation of the statutory plan to complete the planning goals), and PAttributes node label (for example, including multi-attribute elements such as spatial layout, land use zoning, and functional zoning).

[0122] Based on the specific entities in these node labels and the relationships between labels, we can build relationships between the PName node label and the other six groups of node labels, as well as relationships between the entities contained in the labels in the PName node label, PIndex node label, and PAttributes node label.

[0123] In this embodiment of the present invention, the relationship information between nodes in the knowledge graph is presented as "entity-relationship-entity" triples. In the specific implementation of the relationship, some planning attribute information is displayed as a unique relationship between two planning entities. For example, if planning document A is a master plan and it states that the building land area does not exceed 100 square kilometers, then its related subordinate plan B must be implemented within the scope specified by document A. This prevents conflicts between plans.

[0124] In the embodiment of the present invention, in step S400, based on the constructed planning knowledge graph, the linkage relationship between planning semantics is utilized to identify planning conflicts, and the identified planning conflicts are output as follows:

[0125] In an embodiment of the present invention, based on the constructed planning knowledge graph, the linkage relationship between planning semantics is utilized to obtain query planning types and indicators, automatically identify differences in planning content, identify planning conflicts and give reminders. For example, for a new planning text, based on the entity semantic relationship in the knowledge graph, it is possible to automatically identify differences in planning content by querying planning types and indicators to identify planning conflicts, output identified planning conflict results, and reduce related planning conflicts. For example: Planning text A is a master plan, which mentions that the building land does not exceed 100 square kilometers, then its related subordinate plan B needs to be executed within the scope specified in text A. This can prevent conflicts between plans.

[0126] As can be seen from the above, the present invention is aimed at multi-level and multi-type planning, and adopts a knowledge graph construction method. Under the conditions of clarifying existing elements, relationships, thresholds, etc., it is conducive to the identification and elimination of planning conflict differences.

[0127] Exemplary devices

[0128] like Figure 5 As shown in , an embodiment of the present invention provides a multi-level and multi-type planning content difference identification and conflict elimination device, which includes:

[0129] An acquisition and conversion module 510 is used to acquire planning text content, convert the acquired planning text content into planning text data in a specified format, and perform knowledge annotation processing on the converted planning text data;

[0130] The extraction and segmentation module 520 is used to extract the annotated planning text data, use knowledge extraction to segment the planning text semantics, and generate planning structured data;

[0131] The knowledge graph construction module 530 is used to organize the planning knowledge according to the entity-relationship-entity triple model based on the generated planning structured data, and construct a planning knowledge graph;

[0132] The conflict identification module 540 is used to identify planning conflicts based on the constructed planning knowledge graph and utilize the linkage relationship between planning semantics, and output the identified planning conflict results, as described above.

[0133] Wherein, the acquisition conversion module includes:

[0134] An acquisition unit, used to acquire the content of each planning text to form a multi-source planning text original data set;

[0135] The conversion processing unit is used to perform a pre-processing of the original text conversion of the planning text data in the multi-source planning text original data set into the TXT format to obtain the planning text data in the TXT format;

[0136] The tagging unit is used to generate the original text sequence and annotate the text corpus for the planning text data in the specified format; adopt the BIO tagging method to perform entity category tagging and text relationship tagging respectively, obtain the planning text entity sequence based on BIO tagging and the planning text relationship sequence based on BIO tagging respectively, and form the planning text data processed by the tagging, as described above.

[0137] Based on the above embodiment, the present invention also provides an intelligent terminal, whose principle block diagram can be shown as follows: Figure 6 As shown. The intelligent terminal includes a processor, a memory, a network interface, a display screen, and a sensor connected via a system bus. The processor of the intelligent terminal is used to provide computing and control capabilities. The memory of the intelligent terminal includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the intelligent terminal is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a multi-level and multi-type planning content difference identification and conflict elimination method is implemented. The display screen of the intelligent terminal can be a liquid crystal display screen or an electronic ink display screen.

[0138] Those skilled in the art will understand that Figure 6 The principle block diagram shown in the figure is only a block diagram of a partial structure related to the solution of the present invention and does not constitute a limitation on the smart terminal to which the solution of the present invention is applied. The specific smart terminal may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0139] In one embodiment, a smart terminal is provided, comprising a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by one or more processors. The one or more programs include instructions for performing the following operations:

[0140] Acquire planning text content, convert the acquired planning text content into planning text data in a specified format, and perform knowledge annotation processing on the converted planning text data;

[0141] Extract the annotated planning text data, use knowledge extraction to segment the planning text semantics, and generate planning structured data;

[0142] Based on the generated planning structured data, the relationships between planning elements and / or indicators are organized into planning knowledge according to the entity-relationship-entity triple model to construct a planning knowledge graph;

[0143] Based on the constructed planning knowledge graph, the linkage relationship between planning semantics is utilized to identify planning conflicts and output the identified planning conflict results.

[0144] In the multi-level and multi-type planning content difference identification and conflict elimination method, the step of obtaining planning text content and converting the obtained planning text content into planning text data in a specified format includes:

[0145] Obtain the content of each planning text to form a multi-source planning text original data set;

[0146] The planning text data in the multi-source planning text original data set is preprocessed by converting the original text into TXT format to obtain planning text data in TXT format.

[0147] The step of performing knowledge annotation processing on the converted planning text data includes:

[0148] Get the converted planning text data in the specified format;

[0149] Generating original text sequences and annotating text corpus for the planned text data in the specified format;

[0150] The BIO annotation method is used to perform entity category annotation and text relationship annotation respectively, and a planning text entity sequence based on the BIO annotation and a planning text relationship sequence based on the BIO annotation are obtained respectively, forming the annotated planning text data.

[0151] The steps of extracting the annotated planning text data, segmenting the planning text semantics using a knowledge extraction method, and generating planning structured data include:

[0152] The BERT-BiLSTM-CRF model is used to extract knowledge from the annotated planning text data;

[0153] Obtain a planning text entity sequence of the planning text data to obtain an entity training data set; obtain a planning text relationship sequence of the planning text data to obtain a relationship training data set;

[0154] Input the entity training dataset into the BERT-BiLSTM-CRF model for entity recognition, save the optimal model, and predict the entity results;

[0155] Input the relationship training data set into the BERT-BiLSTM-CRF model for relationship extraction, save the optimal model, and predict the relationship results;

[0156] The predicted entity results and predicted entity results are used to generate planning structured data.

[0157] The steps of organizing planning knowledge based on the generated planning structured data and organizing the relationships between planning elements and / or indicators according to the entity-relationship-entity triple model to construct a planning knowledge graph include:

[0158] Based on the generated planning structured data, the relationships between planning elements and / or indicators in the planning structured data are organized according to entity-relationship-entity triples;

[0159] Using the Neo4j graph database, seven groups of graph node labels are set; the seven groups of graph node labels include: a PName node label representing the specific name of the plan, a PPurpose node label representing the respective planning objectives contained in the plan, a PBasis node label representing the specific planning basis, a PRange node label representing the administrative scope involved in the plan, a PIndex node label representing the specific indicators involved in the planning content, a PTime node label representing the expected completion target period of the compiled statutory plan, and a PAttributes node label representing the content of multi-attribute elements;

[0160] Based on the specific entities in the set node labels and the relationships between labels, the relationships between the PName node label and the other six groups of node labels are constructed, as well as the relationships between the entities contained in the labels in the PName node label, PIndex node label, and PAttributes node label.

[0161] The constructed node labels and the relationship information between node labels are presented in the form of entity-relationship-entity triples to complete the construction of the planning knowledge graph.

[0162] The multi-level and multi-type planning content difference identification and conflict elimination method, wherein the steps of identifying planning conflicts based on the constructed planning knowledge graph and utilizing the linkage relationship between planning semantics and outputting the identified planning conflict results include:

[0163] Based on the constructed planning knowledge graph, the linkage relationship between planning semantics is utilized to obtain the query planning type and indicator status, automatically identify the differences in planning content, identify planning conflicts and issue reminders, and avoid planning conflicts, as described above.

[0164] Those skilled in the art will appreciate that all or part of the processes in the above-described embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the above-described embodiments. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAM bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0165] In summary, the present invention discloses a multi-level and multi-type planning content difference identification and conflict elimination method, intelligent terminal and storage medium, which involves the acquisition of planning knowledge, the establishment of knowledge relationships and the elimination of planning conflicts. The method of the present invention comprises: first, obtaining the planning text content and pre-processing the planning text content, including text collection and knowledge annotation; second, extracting the planning knowledge. For the existing national land space planning text, the BERT-BiLSTM-CRF model is used to extract the planning name, planning basis, planning indicators and planning-related attribute information in the planning text, and integrate the planning knowledge. Third, based on the knowledge extraction results, the relationship between planning elements and / or indicators is organized according to the "entity-relationship-entity" triple model to construct a planning knowledge graph. Finally, based on the knowledge graph, the linkage relationship between planning semantics is used to identify planning conflicts. The present invention effectively solves the problem of multi-level and multi-type planning content difference identification and conflict elimination, and can improve the collaborative efficiency of different planning types.

[0166] It should be understood that the application of the present invention is not limited to the above examples. For those skilled in the art, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A multi-level and multi-type planning content difference identification and conflict elimination method, characterized by: The method comprises: Acquire planning text content, convert the acquired planning text content into planning text data in a specified format, and perform knowledge annotation processing on the converted planning text data; Extract the annotated planning text data, use knowledge extraction to segment the planning text semantics, and generate planning structured data; Based on the generated planning structured data, the relationships between planning elements and / or indicators are organized into planning knowledge according to the entity-relationship-entity triple model to construct a planning knowledge graph; Based on the constructed planning knowledge graph, the linkage relationship between planning semantics is utilized to identify planning conflicts and output the identified planning conflict results; The step of performing knowledge annotation processing on the converted planning text data includes: Get the converted planning text data in the specified format; Generating original text sequences and annotating text corpus for the planned text data in the specified format; Using the BIO annotation method to perform entity category annotation and text relationship annotation respectively, respectively obtaining a planning text entity sequence based on the BIO annotation and a planning text relationship sequence based on the BIO annotation, to form the annotated planning text data; Entity definitions are performed on the planning type, planning period, planning implementation scope, planning purpose, planning indicators, planning basis, and other planning attributes contained in the planning text; original text sequence generation and text corpus annotation are performed on the planning text data to obtain a planning text data sequence based on BIO annotation; Label each element as "BX", "IX" or "O". "BX" means that the fragment containing this element is of type X and this element is at the beginning of this fragment. "IX" means that the fragment containing this element is of type X and this element is in the middle of this fragment. "O" means that it does not belong to any type. The steps of organizing planning knowledge based on the generated planning structured data and the relationships between planning elements and / or indicators according to the entity-relationship-entity triple model to construct a planning knowledge graph include: Based on the generated planning structured data, the relationships between planning elements and / or indicators in the planning structured data are organized according to entity-relationship-entity triples; Using the Neo4j graph database, seven groups of graph node labels are set; the seven groups of graph node labels include: a PName node label representing the specific name of the plan, a PPurpose node label representing the respective planning objectives contained in the plan, a PBasis node label representing the specific planning basis, a PRange node label representing the administrative scope involved in the plan, a PIndex node label representing the specific indicators involved in the planning content, a PTime node label representing the expected completion target period of the compiled statutory plan, and a PAttributes node label representing the content of multi-attribute elements; Based on the specific entities in the set node labels and the relationships between labels, the relationships between the PName node label and the other six groups of node labels are constructed, as well as the relationships between the entities contained in the labels in the PName node label, PIndex node label, and PAttributes node label. The constructed node labels and the relationship information between node labels are presented in the form of entity-relationship-entity triples to complete the construction of the planning knowledge graph; The steps of obtaining the planning text content and converting the obtained planning text content into planning text data in a specified format include: Obtain the content of each planning text to form a multi-source planning text original data set; The planning text data in the multi-source planning text original data set is preprocessed by converting the original text into TXT format to obtain planning text data in TXT format.

2. The multi-level and multi-type planning content difference identification and conflict elimination method according to claim 1 is characterized in that: The steps of extracting the annotated planning text data, segmenting the planning text semantics using a knowledge extraction method, and generating planning structured data include: The BERT-BiLSTM-CRF model is used to extract knowledge from the annotated planning text data; Obtain a planning text entity sequence of the planning text data to obtain an entity training data set; obtain a planning text relationship sequence of the planning text data to obtain a relationship training data set; Input the entity training dataset into the BERT-BiLSTM-CRF model for entity recognition, save the optimal model, and predict the entity results; Input the relationship training data set into the BERT-BiLSTM-CRF model for relationship extraction, save the optimal model, and predict the relationship results; The predicted entity results and predicted entity results are used to generate planning structured data.

3. The multi-level and multi-type planning content difference identification and conflict elimination method according to claim 1 is characterized in that: The steps of identifying planning conflicts based on the constructed planning knowledge graph and utilizing the linkage relationship between planning semantics and outputting the identified planning conflict results include: Based on the constructed planning knowledge graph, the linkage relationship between planning semantics is utilized to obtain query planning types and indicators, automatically identify differences in planning content, identify planning conflicts, and issue reminders.

4. A multi-level and multi-type planning content difference identification and conflict elimination device, characterized in that: include: An acquisition conversion module is used to acquire planning text content, convert the acquired planning text content into planning text data in a specified format, and perform knowledge annotation processing on the converted planning text data; The extraction and segmentation module is used to extract the annotated planning text data, use knowledge extraction to segment the planning text semantics, and generate planning structured data; A knowledge graph construction module is used to organize planning knowledge based on the generated planning structured data by organizing the relationships between planning elements and / or indicators according to the entity-relationship-entity triple model to construct a planning knowledge graph; The conflict identification module is used to identify planning conflicts based on the constructed planning knowledge graph and utilize the linkage relationship between planning semantics, and output the identified planning conflict results; The step of performing knowledge annotation processing on the converted planning text data includes: Get the converted planning text data in the specified format; Generating original text sequences and annotating text corpus for the planned text data in the specified format; Using the BIO annotation method to perform entity category annotation and text relationship annotation respectively, respectively obtaining a planning text entity sequence based on the BIO annotation and a planning text relationship sequence based on the BIO annotation, to form the annotated planning text data; Entity definitions are performed on the planning type, planning period, planning implementation scope, planning purpose, planning indicators, planning basis, and other planning attributes contained in the planning text; original text sequence generation and text corpus annotation are performed on the planning text data to obtain a planning text data sequence based on BIO annotation; Label each element as "BX", "IX" or "O". "BX" means that the fragment containing this element is of type X and this element is at the beginning of this fragment. "IX" means that the fragment containing this element is of type X and this element is in the middle of this fragment. "O" means that it does not belong to any type. The steps of organizing planning knowledge based on the generated planning structured data and the relationships between planning elements and / or indicators according to the entity-relationship-entity triple model to construct a planning knowledge graph include: Based on the generated planning structured data, the relationships between planning elements and / or indicators in the planning structured data are organized according to entity-relationship-entity triples; Using the Neo4j graph database, seven groups of graph node labels are set; the seven groups of graph node labels include: a PName node label representing the specific name of the plan, a PPurpose node label representing the respective planning objectives contained in the plan, a PBasis node label representing the specific planning basis, a PRange node label representing the administrative scope involved in the plan, a PIndex node label representing the specific indicators involved in the planning content, a PTime node label representing the expected completion target period of the compiled statutory plan, and a PAttributes node label representing the content of multi-attribute elements; Based on the specific entities in the set node labels and the relationships between labels, the relationships between the PName node label and the other six groups of node labels are constructed, as well as the relationships between the entities contained in the labels in the PName node label, PIndex node label, and PAttributes node label. The constructed node labels and the relationship information between node labels are presented in the form of entity-relationship-entity triples to complete the construction of the planning knowledge graph; The acquisition conversion module includes: An acquisition unit, used to acquire the content of each planning text to form a multi-source planning text original data set; The conversion processing unit is used to perform a pre-processing of the original text conversion of the planning text data in the multi-source planning text original data set into the TXT format to obtain the planning text data in the TXT format; The tagging unit is used to generate original text sequences and annotate text corpora for the planning text data in the specified format; adopt the BIO tagging method to perform entity category tagging and text relationship tagging respectively, obtain the planning text entity sequence based on BIO tagging and the planning text relationship sequence based on BIO tagging respectively, and form the planning text data processed by the tagging.

5. An intelligent terminal, characterized in that: The device comprises a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by one or more processors, wherein the one or more programs include the method according to any one of claims 1 to 3. 6 . A non-transitory computer-readable storage medium, which, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to perform the method according to any one of claims 1 to 3.

Images