CN116187812A - Evaluation method and system for sustainable utilization of cultivated land resources based on geographic knowledge - Google Patents

Evaluation method and system for sustainable utilization of cultivated land resources based on geographic knowledge Download PDF

Info

Publication number
CN116187812A
CN116187812A CN202211643785.5A CN202211643785A CN116187812A CN 116187812 A CN116187812 A CN 116187812A CN 202211643785 A CN202211643785 A CN 202211643785A CN 116187812 A CN116187812 A CN 116187812A
Authority
CN
China
Prior art keywords
index
evaluation
module
knowledge
indexes
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202211643785.5A
Other languages
Chinese (zh)
Inventor
康昕怡
马丹虹
范辉辉
黄雍怀
李伊黎
支盼丁
罗小梅
陈轶文
黄润智
胡淼
范娟
许耿然
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangzhou Alpha Software Information Technology Co ltd
SURVEYING AND MAPPING INSTITUTE LANDS AND RESOURCE DEPARTMENT OF GUANGDONG PROVINCE
Original Assignee
Guangzhou Alpha Software Information Technology Co ltd
SURVEYING AND MAPPING INSTITUTE LANDS AND RESOURCE DEPARTMENT OF GUANGDONG PROVINCE
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangzhou Alpha Software Information Technology Co ltd, SURVEYING AND MAPPING INSTITUTE LANDS AND RESOURCE DEPARTMENT OF GUANGDONG PROVINCE filed Critical Guangzhou Alpha Software Information Technology Co ltd
Priority to CN202211643785.5A priority Critical patent/CN116187812A/en
Publication of CN116187812A publication Critical patent/CN116187812A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/205—Parsing
    • G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/02—Knowledge representation; Symbolic representation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/067—Enterprise or organisation modelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/26—Government or public services

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Theoretical Computer Science (AREA)
  • Strategic Management (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Educational Administration (AREA)
  • Tourism & Hospitality (AREA)
  • Development Economics (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Game Theory and Decision Science (AREA)
  • Computational Linguistics (AREA)
  • Operations Research (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Quality & Reliability (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • Primary Health Care (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses a method and a system for evaluating sustainable utilization of cultivated land resources based on geographic knowledge, which are oriented to multi-scene evaluation of sustainable utilization of cultivated land resources, and are combined with the geographic knowledge to build a body layer in a top-down modularization manner, wherein the body layer comprises a basic module, an index module and a scene module; under the constraint of the ontology layer, the instance corresponding to each concept of the ontology layer and the attribute and relation thereof are complemented from bottom to top, including entity identification, attribute extraction and relation extraction, so as to obtain a knowledge graph; based on the knowledge graph, extracting subgraphs containing the attention points, the criterion layers and the indexes according to a given rule; the combined weighting model is fused, weights among the attention points, the criterion layers and the indexes in the subgraph are quantized to obtain an evaluation grade, the main obstacle indexes are finally obtained and mapped back to the knowledge graph, the basic indexes are traced back, and the cause diagnosis is assisted, so that the problems that the existing cultivated land resource sustainable utilization evaluation efficiency is low and automatic cause diagnosis cannot be achieved based on the evaluation result are solved.

Description

Evaluation method and system for sustainable utilization of cultivated land resources based on geographic knowledge
Technical Field
The invention relates to the technical field of knowledge graph and arable land resource sustainable utilization evaluation, in particular to a arable land resource sustainable utilization evaluation method based on geographic knowledge.
Background
The sustainable utilization evaluation of cultivated land resources is an important means for realizing efficient utilization and sustainable development of cultivated land resources, relates to multi-scene evaluation of cultivated land quantity, quality and potential, and the index sources relate to a series of rich basic geographic data, normalized citation files, homeland space planning results, homeland investigation data results and the like, the indexes relate to data dispersion and complexity, the same index can be used for different evaluation scenes, and the indexes have association relations. For example, the cultivated land resource can be continuously utilized to evaluate multiple scenes including cultivated land quality evaluation, cultivated land treatment potential evaluation, cultivated land backup resource suitable-for-cultivation evaluation and the like, and although the evaluation purposes are different, the evaluation indexes of the evaluation indexes have cross overlapping, such as the index of the terrain gradient can be used for multiple evaluation scenes including cultivated land quality, cultivated land treatment potential, cultivated land backup resource suitable-for-cultivation and the like;
in addition, the indexes originate from a geographic knowledge system, have association relations, for example, the evaluation of the tillability of the backup resources of the cultivated land of the evaluation scene relates to three criterion layers of natural adaptability, ecological safety and social feasibility, and comprise evaluation indexes such as terrain gradient, irrigation condition, cultivation convenience and the like, the cultivation convenience can establish association relations with the irrigation condition, the irrigation condition can establish association relations with the terrain gradient, and the indexes with the association relations form the evaluation index system together. Meanwhile, the method has an association relationship with indexes outside the evaluation index system, for example, the cultivation convenience can establish an association relationship with the road accessibility of the field and the connection degree.
The conventional continuous utilization of cultivated land resources for multi-scene evaluation is mainly characterized in that an evaluation system is manually established, indexes are selected and combined, index weights are calculated, the cause tracing cannot be automatically performed based on the association relation among the indexes, knowledge organization is lacked, and the problems of long time consumption, low performance and incapability of quickly identifying the causes exist. In addition, for the same evaluation scene, because of the differences of natural endowment and economic development, evaluation emphasis points of different regions are different, so that a criterion layer is required to be introduced in the evaluation to meet the evaluation requirements of different regions and users.
Disclosure of Invention
In order to solve the problems, the invention provides a method and a system for evaluating sustainable utilization of cultivated land resources based on geographic knowledge, which are used for solving the technical problems that the efficiency of evaluating sustainable utilization of cultivated land resources is low and automatic cause diagnosis cannot be performed based on an evaluation result.
The embodiment of the invention provides a geographic knowledge-based farmland resource sustainable utilization evaluation method, which comprises the following steps:
s1: modularizing the cultivated land scene from top to bottom through geographic knowledge reference to construct a body layer;
s2: under the constraint of the ontology layer, complementing the instance corresponding to each concept of the ontology layer and the attribute and relation thereof from bottom to top to obtain a knowledge graph;
s3: based on the knowledge graph, extracting a subgraph containing the attention points, the criterion layers and the indexes;
s4, fusing a combined weighting model, quantifying weights among the attention points, the criterion layers and the indexes in the subgraph, and carrying out continuous utilization evaluation of cultivated land resources to obtain an evaluation grade;
s5, acquiring a main obstacle index according to the evaluation grade, mapping the main obstacle index back to the knowledge graph, tracing the root index and assisting in the cause diagnosis.
Further, in step S1, the ontology layer is divided in a modularized manner by combining with the geographic knowledge, the ontology layer includes a base module, an index module and a scene module after being divided in a modularized manner, corpus such as a table, a textbook, a literature, a normative reference file and the like related to the geographic knowledge system are collected according to the divided modules, a corpus is constructed, existing corpora at home and abroad are collected, reusable ontology and corpus are extracted from the corpus and the ontology, structured data are obtained by packaging the ontology and corpus, the structured data are used for carrying out structural modeling on the base module, the index module and the scene module, the base module, the entity module and the scene module are expressed in a formalized manner by using an ontology construction tool, and top-down module integration is completed by mutual reference among the modules.
Further, in step S2, a web crawler tool is used to obtain data such as natural resource related normative files, documents and reports, basic geographic data such as administrative regions, geological features and the like and data such as homeland space planning results such as three-region three-line and the like are collected to form a database, an instance included in concepts in an ontology layer is mined and extracted from the database by using a deep learning model, a relation among the instances is extracted by a relation extraction method based on templates, a concept attribute list defined in the ontology layer is extracted for the extracted instance, attributes of the instance are filled by a rule-based method and according to elements and association relations of each module in the ontology layer, knowledge alignment is achieved by adopting a similarity function based on the extracted instance and attributes, relations and the like of the instance and a pattern defined in the ontology layer, the instance knowledge is converted into RDF based on the mapping relation, and a missing file in the RDF triplet is completed to obtain a knowledge map.
Further, in step S3, a natural language of the user is subjected to syntactic analysis according to the scene requirement, a semantic dependency relationship between vocabularies is obtained by adopting rule-based dependency syntactic analysis, a tree-like query graph Q is formed, sub-graph matching is performed by adopting a Multi-way Join method based on breadth first, sub-graph expansion is performed based on the knowledge graph structure, and a sub-graph including a focus point, a criterion layer and an index is obtained.
Further, in step S4, a combined weighting model is fused, weights among the attention points, the criterion layers and the indexes in the subgraph are quantized, and natural resource evaluation is performed, so that the method steps of obtaining the evaluation level are as follows:
s41: evaluation index set { M } in subgraph 1 ,M 2 ,…M n And n is a set of n indexes of the same level in the evaluation index system, and n is more than or equal to 2. Grading and grading all index items in the evaluation index set according to the types, the properties and the threshold values of the index items to obtain an index value grading matrix M= (M) 1 ，m 2 ，…，m n ) T Ensuring that all indexes are in the same dimension;
s42: calculating subjective weight of the evaluation index through quantitative analysis of determining importance degree between the sequence relation and the adjacent indexes;
s421: using Delphi method, in index set { M 1 ,M 2 ,…，M n The most important index value M is obtained by screening 1 * And sequentially screening to obtain M 2 * After n-1 selections, the evaluation index set { M } can be determined 1 ,M 2 ,…，M n-1 ，M n Sequence relation index set { M } obtained by sequencing according to importance degree 1 * ,M 2 * ,…，M n-1 * ，M n * }。
S422: quantitative analysis of importance degree between adjacent indexes, and concentration of adjacent evaluation indexes M in sequence relation indexes n-1 * And M n * The degree of importance of (c) is quantified and expressed as:
Figure BDA0004008928480000041
wherein the method comprises the steps of
Figure BDA0004008928480000042
And->
Figure BDA0004008928480000043
Representing adjacent evaluation index->
Figure BDA0004008928480000044
And->
Figure BDA0004008928480000045
Weights of (2);
s423: index weight calculation according to a given r k Assigning a value to obtain an evaluation index
Figure BDA0004008928480000046
Weight of +.>
Figure BDA0004008928480000047
The calculation formula is as follows: />
Figure BDA0004008928480000048
The r is i For a given r k The i-th value is assigned by assignment, so that the subjective weight of the sequence relation index set can be obtained, and further the subjective weight matrix X= (X) of the evaluation index set is obtained 1 ，x 2 ，…,x n ) T The method comprises the steps of carrying out a first treatment on the surface of the ( The subjective weight matrix acquisition method has the beneficial effects that: the distribution and the utilization condition of cultivated land resources can be intuitively judged through subjective weights; )
S43: the objective weight of the evaluation index is calculated, the information entropy of the evaluation index is calculated, and the formula is as follows:
Figure BDA0004008928480000049
wherein e i Represents the information entropy of the ith evaluation index, p represents the number of evaluation objects, m ij An ith evaluation index value indicating a jth evaluation target;
( The beneficial effects of obtaining the index information entropy are as follows: measuring the variance degree of the index value to prepare for the subsequent calculation of the entropy weight; )
Calculating an evaluation index entropy weight, wherein the formula is as follows:
Figure BDA00040089284800000410
and then objective weight vector y= (Y) of the evaluation index set can be obtained 1 ，y 2 ，…,y n ) T And performing optimization fitting on the subjective weight and the objective weight vector to obtain a comprehensive weight matrix, wherein the formula is as follows:
Figure BDA0004008928480000051
wherein Z when the function H (Z) takes a minimum value i The obtained weight is the comprehensive weight matrix Z= (Z) 1 ，z 2 ，…,z n ) T . Weights Z of criterion layers i The sum of the index weights is the sum of the index weights;
s45: calculating an evaluation result score based on index scores and weights by combining a comprehensive factor score method and a restrictive factor method, grading the evaluation result score according to a natural discontinuous method, and multiplying an index score matrix M by a comprehensive weight matrix by adopting the comprehensive factor score method to obtain an evaluation result S, wherein the specific formula is as follows:
Figure BDA0004008928480000052
meanwhile, the combined restriction factor method is evaluated, and if the restriction factor index score is 0, the evaluation result is directly 0.
Further, in step S5, the obstacle degree model is used to identify the main obstacle index according to the evaluation level, map the main obstacle index back to the knowledge graph, take the main obstacle index as a node and perform depth traversal in the knowledge graph to obtain a directed acyclic graph with the main obstacle index as a vertex, optimize the directed acyclic graph according to rule design, traverse the optimized directed acyclic graph, find the index where the node with the degree of 0 is located, and the index is the root index, thereby realizing source tracing.
The arable land resource sustainable utilization evaluation system based on geographic knowledge is characterized by comprising a map construction module, a subgraph extraction module, a calculation evaluation module and an inference diagnosis module, wherein the map construction module comprises an association relation construction module, and the module can realize the steps in any arable land resource sustainable utilization evaluation method based on geographic knowledge.
Preferably, the profile construction module: combining with geographic knowledge, the method is used for constructing a knowledge graph by adopting a mode of constructing a body layer from top to bottom and supplementing examples, attributes and relations of the examples from bottom to top, and comprises a base module, an index module and a scene module;
and a sub-graph extraction module: the method is used for extracting the subgraph meeting scene requirements (such as tillable evaluation of tillable resources) by adopting a method for constructing a query graph Q and matching the subgraph, and comprises a concern point, a criterion layer and an index;
and (3) calculating and evaluating a module: the method is used for determining the index weight of the evaluation index in the subgraph by adopting a subjective weighting model and an entropy weight method, and evaluating by adopting a comprehensive factor score method and a restrictive factor method according to the index grade and the corresponding index weight to obtain an evaluation result.
An inference diagnosis module: and the method is used for tracking the main obstacle index according to the evaluation result, mapping the main obstacle index back to the knowledge graph, and reasoning to obtain the root index influencing the evaluation result so as to complete the source tracing and diagnosis.
The map construction module further comprises:
the association relation construction module: the method is used for establishing an association relationship among the basic module, the index module and the scene module; establishing an association relation among indexes in an index module;
the beneficial effects of the invention are as follows:
1. combining geographical knowledge, following the thought of combining top-down construction and bottom-up supplement, modularizing and structuring to construct a knowledge map, and uniformly expressing the cultivated land resource information, thereby comprehensively expressing multidimensional knowledge of sustainable utilization of cultivated land resources;
2. the embodiment of the invention modularly constructs the body layer based on the content and the characteristics of cultivated land resources, and comprises a basic module, an index module and a scene module, wherein the modules are mutually independent and associated with each other, so that the multiplexing and updating of the body are facilitated, and the convenience of carrying out association analysis on the sustainable utilization scene of the cultivated land resources can be effectively improved;
3. the embodiment of the invention is oriented to scene requirements, and based on the knowledge graph, the subgraphs comprising the attention points, the criterion layers and the indexes are extracted according to the given rule, so that the requirements of different users can be met;
4. according to the embodiment of the invention, a factor weighting model is fused, weights among the attention points, the criterion layers and the indexes in the subgraph are quantized, and the sustainable utilization evaluation of cultivated land resources is carried out, so that the knowledge graph network structure is utilized, and the sustainable utilization evaluation efficiency of the cultivated land resources is improved;
5. according to the evaluation grade, the main obstacle index is obtained, the main obstacle index is mapped back to the knowledge graph, the root index is traced back, the source tracing is completed, the advantage of the dominant relationship of the knowledge graph is utilized, the invisible association between the indexes is rapidly discovered and locked, and the source diagnosis can be effectively assisted.
Drawings
FIG. 1 is a schematic flow chart of an evaluation method for sustainable utilization of cultivated land resources based on geographic knowledge;
FIG. 2 is a schematic diagram of a top-down and bottom-up combined modular knowledge graph construction process provided by the invention;
FIG. 3 is a schematic diagram of a modular knowledge building model provided by the present invention;
FIG. 4 is a simplified structural diagram of an association relationship between indexes according to an embodiment of the present invention;
FIG. 5 is a schematic structural diagram of an evaluation system for sustainable utilization of cultivated land resources based on geographic knowledge according to an embodiment of the present invention;
Detailed Description
The conception, specific structure, and technical effects produced by the present invention will be clearly and completely described below with reference to the embodiments and the drawings to fully understand the objects, aspects, and effects of the present invention. It should be noted that, in the case of no conflict, the embodiments and features in the embodiments may be combined with each other.
In the description of the present invention, a number means one or more, a number means two or more, and greater than, less than, exceeding, etc. are understood to not include the present number, and above, below, within, etc. are understood to include the present number. The description of the first and second is for the purpose of distinguishing between technical features only and should not be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated or implicitly indicating the precedence of the technical features indicated.
As shown in fig. 1, an embodiment of the present invention provides a method for evaluating sustainable utilization of cultivated land resources based on geographical knowledge, including:
s1: modularizing the cultivated land scene from top to bottom through geographic knowledge reference to construct a body layer;
s2: under the constraint of the ontology layer, complementing the instance corresponding to each concept of the ontology layer and the attribute and relation thereof from bottom to top to obtain a knowledge graph;
s3: based on the knowledge graph, extracting a subgraph containing the attention points, the criterion layers and the indexes;
s4, fusing a combined weighting model, quantifying weights among the attention points, the criterion layers and the indexes in the subgraph, and carrying out continuous utilization evaluation of cultivated land resources to obtain an evaluation grade;
s5, acquiring a main obstacle index according to the evaluation grade, mapping the main obstacle index back to the knowledge graph, tracing the root index and assisting in the cause diagnosis.
As shown in fig. 2, further, in step S1, the ontology layer is divided in a modularized manner by combining with the geographic knowledge, the ontology layer includes a basic module, an index module and a scene module after being divided in a modularized manner, corpora such as a table, a textbook, a literature, a normalized reference file and the like related to the geographic knowledge system are collected according to the divided modules, a corpus is constructed, existing ontology libraries at home and abroad are collected, reusable ontologies and corpora are extracted from the corpus and the ontology library, and structured data are obtained by packaging the ontologies and the corpora
Preferably, (1) the base module is constructed. The remarkable characteristics of the knowledge related to cultivated land resources compared with the knowledge in other fields are complex space-time characteristics. Thus, there is a need for structured modeling of space-time elements in a base module, including time and space dimensions:
time dimension: at utc+08:00 is a time standard, three time scales of year, month and day, and a time system is established.
Space dimension: the space dimension refers to national standard 'encoding rules of terrestrial space network', and establishes encoding rules, encoding sequences and encoding calculation methods of natural resource entities; on the basis of referring to the city and county region codes of China province, establishing three-level administrative division scale description of province, city and county/region.
(2) And (5) constructing an index module. The index module is mainly constructed from three aspects of index classification dimension, index type and index relation.
According to the embodiment, the cultivated land resource sustainable utilization evaluation index set is obtained based on the normative file in the constructed corpus, indexes are classified from n dimensions by combining expert domain knowledge, and index dimensions are formed, and can be respectively used for topography, vegetation soil, hydrologic climate, development, utilization, planning management and control and the like. And each dimension corresponds to a plurality of indexes, such as the indexes of landform, associated terrain gradient, altitude and the like, the indexes of vegetation soil, associated soil texture, soil salinization degree and the like, the indexes of groundwater burial depth, annual precipitation and the like, the indexes including cultivation convenience, irrigation conditions and the like are developed and utilized, and the planning management and control comprises the indexes of ecological protection restoration areas, related limitation conditions of national soil space planning and the like.
Among the index types as the index attributes, there are three main types: numerical, boolean, and string-type indices.
Referring to fig. 4, a simplified schematic diagram of an association relationship between indexes according to an embodiment of the present invention is shown:
the embodiment of the invention establishes a relation template based on application requirements, associates the relation between nodes, and establishes a one-way association relation of many-to-many. The form is as follows:
M1——>M2
"M1" is the "result" and "M2" is the "cause", one result may be directed to multiple causes, and multiple results may also be directed to one cause. For example, a place with a steep terrain gradient generally needs to be irrigated by diversion from the upstream of a river, and compared with a place with a slow gradient, the area is dry in irrigation, high in branch channel elevation, long in channel line, and many in curves, and the irrigation condition is poor, so that the terrain gradient is one of the formation factors of the irrigation condition, and meanwhile, the terrain gradient and the irrigation condition also influence cultivation convenience; for another example, soil salinization refers to a process in which salt in the bottom layer of soil or groundwater rises to the surface of the earth along with capillary water, and after evaporation of water, salt is accumulated in the surface layer soil, so that soil salinization is affected by soil pH, soil dryness, and groundwater burial depth. In sum, the irrigation condition and the cultivation convenience can respectively establish an association relationship with the underground gradient, the soil salinization can respectively establish an association relationship with the soil pH value, the soil dryness and the underground water burial depth, and the irrigation condition can also establish an association relationship with the underground water burial depth.
In summary, based on the multi-type information in the index module, an index item-index dimension (MN) relationship, an index item-index item (MM) relationship, and an index-attribute (MA) relationship are constructed.
(3) And constructing a scene module. And referring to the basic module and the index module, and constructing a scene module containing a theme, a focus point and a criterion layer. For example, the subject of sustainable utilization of cultivated land resources can relate to the point of interest of the cultivated land reserve resource, and the evaluation of the cultivated land reserve resource can be performed from three criterion layers of natural suitability, ecological safety and social feasibility, and corresponding indexes can be selected from index modules for the three criterion layers to construct an evaluation system comprising a 'point of interest-criterion layer' level.
Based on the multi-class information in the scene module, a theme-focus (HF) relation and a focus-criterion layer (FC) are constructed.
And carrying out structural modeling on the basic module, the index module and the scene module through the structural data, carrying out formal expression on the multi-type information of the basic module, the entity module and the scene module by utilizing an ontology construction tool, and completing top-down module integration through mutual reference among the modules.
In the embodiment of the invention, geographic knowledge can be collected, the field range of the knowledge graph body is determined according to the target or application requirement, and a body layer is constructed in a modularized manner according to the principles of high aggregation degree in modules and low coupling degree between modules, so that the body is divided into a basic module, an index module and a scene module. The basic module mainly comprises time and space and can be used for supporting the construction of the entity module and the scene module; the index module comprises index items and index dimensions and can support the construction of the scene module; the scene module is specific to application, including theme, focus, etc., and can refer to the basic module and the index module in construction.
The multiple types of information include topics, points of interest, layers of criteria, metrics, attributes, time, and space.
The reference relation among the modules mainly comprises:
criterion layer-index dimension-index entry (CNM) relationship. Multiple index dimensions can be associated according to the characteristics of the criterion layer, wherein the index dimensions are formed by aggregating index items with index connotations. One criterion layer may be associated with multiple index dimensions, depending on the point of interest, or may be a many-to-many relationship.
Index item-attribute-time-space (MATS) relationship. The attribute, time and space of the geographic elements of the index item evaluation mainly reflect the co-occurrence relation of the index item and the attribute, time and space in a specific scene.
Further, in step S2, a web crawler tool is used to obtain data such as natural resource related normative files, documents and reports, basic geographic data such as administrative regions, geological features and the like and data such as homeland space planning results such as three-region three-line and the like are collected to form a database, an instance included in concepts in an ontology layer is mined and extracted from the database by using a deep learning model, a relation among the instances is extracted by a relation extraction method based on templates, a concept attribute list defined in the ontology layer is extracted for the extracted instance, attributes of the instance are filled by a rule-based method and according to elements and association relations of each module in the ontology layer, knowledge alignment is achieved by adopting a similarity function based on the extracted instance and attributes, relations and the like of the instance and a pattern defined in the ontology layer, the instance knowledge is converted into RDF based on the mapping relation, and a missing file in the RDF triplet is completed to obtain a knowledge map.
As shown in fig. 3, the set of index dimensions is defined as E, e= [ w ] 1 ,w 2 ,w 3 ……w u ]The sum of elements of the set E is u, and two index dimensions w i And w i+1 Is of the correlation Sim sum :Sim sum ＝a×Sims(w i ,w i+1 )+b×Simt(w i ,w i+1 )+c×Simm(w i ,w i+1 )
Wherein Sims (w i ,w i+1 ) Semantic facies corresponding to index dimension entitiesSimilarity, simt (w i ,w i+1 ) Corresponding to the classification similarity of the index dimension entities, simm (w i ,w i+1 ) And corresponding to the similarity of the index items, wherein a, b and c are constant values, a, b and c epsilon (0, 1), i is the i element in the set E, and i epsilon (1, n-1).
By calculating the relevance score and then setting two similarity thresholds, the total relevance score Sim is judged sum In which association interval, 3 cases are formalized:
if Sim is sum >t 2 W is then i And w i+1 Matching;
if t 1 <Sim sum <t 2 W is then i And w i+1 A possible match;
if Sim is sum <t 1 W is then i And w i+1 Mismatch.
Wherein t is 1 And t 2 For the lower and upper bounds of the similarity threshold, sim will be sum >t 2 And defining that the index dimensions are all aggregated, and attributing the aggregated index dimensions to an upper criterion layer to realize knowledge alignment.
Preferably, according to the characteristics of cultivated land resource elements, a BERT-CNN-BiLSTM-CRF model is designed for entity identification, specifically:
determining a criterion layer according to entity characteristics in a body layer facing to cultivated land resource elements, and designing a manual labeling mode: and determining a criterion layer according to the information of the evaluation focus point of the entity set corresponding to the index item of the index module in the ontology layer. Taking the attention point of the evaluation of the tillability of the cultivated land reserve resources as an example, the attention point has three types of standard layers of natural suitability, ecological safety and social feasibility, wherein the natural suitability comprises four types of elements of index dimension, index item, attribute and space-time element, and the index item comprises terrain gradient, soil texture and the like; wherein the related entities comprise evaluation object entities such as woodland, grassland, mining site and the like; also included are entities related to the index items, such as terrain and soil, etc.
And (3) adopting a BIO three-segment marking method for various text information in the database. Determining a criterion layer according to natural language, wherein the two tasks are determined as follows: criterion layer classification and criterion layer element extraction. In the rule layer classification task, the whole text of government technical guideline documents, documents and the like is marked according to the rule layer types, and the training unit is divided into the whole text. And extracting the elements at the criterion layer, and dividing and identifying the training unit by using a single sentence.
The BERT-CNN-Bi LSTM-CRF model is designed to perform entity recognition, and based on the traditional depth model BiLSTM-CRF model, the BERT and the model CNN model are introduced, so that semantic information can be enriched, and sentence semantic sparseness interference is reduced. The model is divided into four layers, wherein the first layer is a BERT pre-training word vector model, and training corpus is changed into word vectors; the second layer CNN model receives word vectors trained by the BERT model as input vectors, and the feature vectors are obtained through convolution calculation by means of one-dimensional convolution layers and filling settings of fixed windows by combining word vector features with context information in depth on the premise of not changing word vector dimensions, so that sentence contexts are better combined; the third layer BiLSTM model inputs the feature vector calculated by the rolling sub-layer into BiLSTM to finish the extraction of the context information; the fourth layer is a CRF model, the model scores the result processed by the BiLSTM layer through a transfer matrix, the normalized probability distribution is obtained by using a Softmax function, and the maximum probability set is selected to obtain the final labeling sequence of the CRF layer. F1 value verification is carried out on the training results of each round of epoch, the result retention is carried out with the accuracy threshold value of 70%, and finally the optimal accuracy result is selected through accuracy comparison.
And marking the statement set by using a relation extraction template constructed by the ontology layer based on the intra-module association relation and the inter-module reference relation defined in the ontology layer, and forming a training model. And selecting a PCNN+attribute algorithm to train the annotation set. The CNN/PCNN is used as the presence encoder and a sentence-level presence mechanism is used.
Combining the body structure and the data attribute, forming an attribute template, traversing the entity and sentences with related attributes in a corpus set consisting of documents, government standardability reference files and the like, marking the parts of speech of the sentences by adopting a CRF algorithm, substituting the marked result into a syntactic analyzer for syntactic analysis, adopting a dependency algorithm by the syntactic analyzer, analyzing the syntactic result by matching the grammar template, extracting the attribute, such as a main-predicate structure and the like.
Further, in step S3, a natural language of the user is subjected to syntactic analysis according to the scene requirement, a semantic dependency relationship between vocabularies is obtained by adopting rule-based dependency syntactic analysis, a tree-like query graph Q is formed, sub-graph matching is performed by adopting a Multi-way Join method based on breadth first, sub-graph expansion is performed based on the knowledge graph structure, and a sub-graph including a focus point, a criterion layer and an index is obtained.
Preferably, for example, for sustainable utilization of cultivated land resources, the user proposes "comprehensive consideration of natural, ecological and economic cultivated land backup resource suitable cultivation", and the query graph Q established according to the description is put into a knowledge graph for matching: in the knowledge graph, the connection relation between the 'plowing land reserve resource plowability' and the 'natural suitability' and the 'ecological safety' can be directly matched; however, the method has no connection relation with economic benefit, and considers the economic benefit, the natural suitability and the ecological safety as the criterion layers in the knowledge structure, so that the same relation between the cultivated land reserve resource suitable cultivated property and the economic benefit can be established, and a subgraph meeting the scene requirement is obtained, namely, under the evaluation of the cultivated land reserve resource suitable cultivated property, the three criterion layers of the natural suitability, the ecological safety and the economic benefit and the corresponding indexes thereof are included.
Further, in step S4, a combined weighting model is fused, weights among the attention points, the criterion layers and the indexes in the subgraph are quantized, and the sustainable utilization evaluation of cultivated land resources is carried out, so that the method steps of obtaining the evaluation grade are as follows:
s41: and defining an evaluation index set { M1, M2, … Mn } in the subgraph as a set of n indexes of the same level in the evaluation index system, wherein n is more than or equal to 2. Grading and grading all index items in the evaluation index system according to the types, the properties and the threshold values of the index items to obtain an index value grading matrix M= (M) 1 ，m 2 ，…，m n ) T Ensuring that all indexes are in the same dimension;
preferably, all the evaluation index values within the sub-graph are scored in a grading manner;
and defining an evaluation index set { M1, M2, … Mn } in the subgraph as a set of n indexes of the same level in the evaluation index system, wherein n is more than or equal to 2. Grading and grading all index items in the evaluation index set according to the types, the properties and the threshold values of the index items to obtain an index value grading matrix M= (M) 1 ，m 2 ，…，m n ) T Ensuring that all indexes are in the same dimension. For example, in the evaluation of the suitability for tillable resources of a tillable land, the slope of the terrain is a negative indicator, that is, the greater the slope, the less suitable, and therefore, the threshold range ". Ltoreq.3 °, 3 to 8 °, 8 to 15 °, 15 to 25 °,>25 ° ", index the index value to five scale scores of" 4 (high), 3 (high), 2 (medium), 1 (low), 0 (low) "; for another example, the ecological condition is used as a boolean index, and the index value can be scored into two grades of 4 (feasible) and 0 (infeasible) in the case of only two conditions, namely, the "ecological protection infrared ray" and the "ecological protection red line".
S42: calculating subjective weight of the evaluation index through quantitative analysis of determining importance degree between the sequence relation and the adjacent indexes;
s421: using Delphi method, in index set { M 1 ,M 2 ,…，M n Selecting the most important evaluation index M 1 * Selecting a next most important evaluation index M from the n-1 indexes remaining in the index set 2 * After n-1 selections, the evaluation index set { M } can be determined 1 ,M 2 ,…，M n-1 ，M n Sequence relation index set { M } obtained by sequencing according to importance degree 1 * ,M 2 * ,…，M n-1 * ，M n * }。
S422: quantitative analysis of importance degree between adjacent indexes, and concentration of adjacent evaluation indexes M in sequence relation indexes n-1 * And M n * The degree of importance of (c) is quantified and expressed as:
Figure BDA0004008928480000141
wherein the method comprises the steps of
Figure BDA0004008928480000142
And->
Figure BDA0004008928480000143
Representing adjacent evaluation index->
Figure BDA0004008928480000144
And->
Figure BDA0004008928480000145
Weights of (2);
s423: index weight calculation according to a given r k Assigning a value to obtain an evaluation index
Figure BDA0004008928480000146
Weight of +.>
Figure BDA0004008928480000147
The calculation formula is as follows:
Figure BDA0004008928480000148
the r is i For a given r k The i-th value is assigned by assignment, so that the subjective weight of the sequence relation index set can be obtained, and further the subjective weight matrix X= (X) of the evaluation index set can be obtained 1 ，x 1 ，…,x n ) T ；
S43: the objective weight of the evaluation index is calculated, the information entropy of the evaluation index is calculated, the variance degree of the index is measured, and preparation is made for the calculation of the entropy weight later, wherein the formula is as follows:
Figure BDA0004008928480000151
wherein e i Represents the information entropy of the ith evaluation index, and p represents the evaluation pairNumber of elephants, m ij The ith evaluation index value of the jth evaluation object is represented, and the entropy weight of the evaluation index is calculated, wherein the formula is as follows:
Figure BDA0004008928480000152
and then objective weight vector y= (Y) of the evaluation index set can be obtained 1 ，y 2 ，…,y n ) T And performing optimization fitting on the subjective weight and the objective weight vector to obtain a comprehensive weight matrix, wherein the formula is as follows:
Figure BDA0004008928480000153
wherein Z when the function H (Z) takes a minimum value i The obtained weight is the comprehensive weight matrix Z= (Z) 1 ，z 2 ，…,z n ) T . Weights Z of criterion layers i The sum of the index weights is the sum of the index weights;
s45: calculating an evaluation result score based on index scores and weights by combining a comprehensive factor score method and a restrictive factor method, grading the evaluation result score according to a natural discontinuous method, and multiplying an index score matrix M by a comprehensive weight matrix by adopting the comprehensive factor score method to obtain an evaluation result S, wherein the specific formula is as follows:
Figure BDA0004008928480000154
meanwhile, the combined restriction factor method is evaluated, and if the restriction factor index score is 0, the evaluation result is directly 0.
Further, in step S5, a main obstacle index is obtained according to the evaluation level, and mapped back to the knowledge graph, and the root index is traced back to assist in the cause diagnosis.
S51: identifying a main obstacle index by using an obstacle degree model according to the evaluation grade;
in order to analyze the degree of negative influence of a certain index on an evaluation result due to self investment, a barrier degree model is quoted, and a specific calculation formula is as follows:
Figure BDA0004008928480000161
degree of obstacle o i And when the index is maximum, identifying the index as a main obstacle index.
And S52, mapping the main obstacle index back to the knowledge graph, performing depth traversal in the knowledge graph by taking the main obstacle index as a node to obtain a directed acyclic graph by taking the main obstacle index as a vertex, and optimizing the directed acyclic graph according to rule design.
Based on the association relation network between indexes in the knowledge graph, performing depth-first traversal in the knowledge graph by taking the main obstacle index as a node: and selecting adjacent nodes with the main obstacle index nodes, returning to the previous layer until the last node has no adjacent nodes, and traversing the other adjacent node of the previous node until the traversing is finished, thereby obtaining the directed acyclic graph taking the main obstacle index as the vertex.
Grading and grading all indexes in the directed acyclic graph, designing a reject rule in the same step as S41, and rejecting index nodes with better grading to obtain the optimized directed acyclic graph.
And S53, traversing the optimized directed acyclic graph, finding out an index of a node with the degree of 0, wherein the index is a root index, and realizing source tracing.
As shown in fig. 5, the system for evaluating the sustainable utilization of cultivated land resources based on geographic knowledge is characterized by comprising a map construction module, a sub-graph extraction module, a calculation evaluation module and an inference diagnosis module, wherein the map construction module further comprises an association relation construction module, and the module can realize the steps in any of the evaluation methods for evaluating the sustainable utilization of cultivated land resources based on geographic knowledge.
Preferably, the profile construction module: the method is used for constructing a natural resource evaluation knowledge graph by adopting a mode of constructing a body layer from top to bottom and supplementing examples, attributes and relations of the examples from bottom to top, and comprises a base module, an index module and a scene module;
and a sub-graph extraction module: the method is used for extracting the subgraph meeting scene requirements (such as tillable evaluation of tillable resources) by adopting a method for constructing a query graph Q and matching the subgraph, and comprises a concern point, a criterion layer and an index;
and (3) calculating and evaluating a module: the method is used for determining the index weight of the evaluation index in the subgraph by adopting a G1 subjective weighting model and an entropy weight method, and evaluating by adopting a comprehensive factor score method and a restrictive factor method according to the index grade and the corresponding index weight to obtain an evaluation result.
An inference diagnosis module: and the method is used for tracking the main obstacle index according to the evaluation result, mapping the main obstacle index back to the knowledge graph, and reasoning to obtain the root index influencing the evaluation result so as to complete the source tracing and diagnosis.
The map construction module further comprises:
the association relation construction module: the method is used for establishing an association relationship among the basic module, the index module and the scene module; and establishing an association relation among the indexes in the index module.
Although the present invention has been described in considerable detail and with particularity with respect to several described embodiments, it is not intended to be limited to any such detail or embodiment or any particular embodiment so as to effectively cover the intended scope of the invention. Furthermore, the foregoing description of the invention has been presented in its embodiments contemplated by the inventors for the purpose of providing a useful description, and for the purposes of providing a non-essential modification of the invention that may not be presently contemplated, may represent an equivalent modification of the invention.

Claims (7)

1. The method and the system for evaluating the sustainable utilization of the cultivated land resources based on the geographic knowledge are characterized by comprising the following steps:
s1: modularizing the cultivated land scene from top to bottom through geographic knowledge reference to construct a body layer;
s2: under the constraint of the ontology layer, complementing the instance corresponding to each concept of the ontology layer and the attribute and relation thereof from bottom to top to obtain a knowledge graph;
s3: based on the knowledge graph, extracting a subgraph containing the attention points, the criterion layers and the indexes;
s4, fusing a combined weighting model, quantifying weights among the attention points, the criterion layers and the indexes in the subgraph, and carrying out multi-scene evaluation on sustainable utilization of cultivated land resources to obtain an evaluation grade;
s5, acquiring a main obstacle index according to the evaluation grade, mapping the main obstacle index back to the knowledge graph, tracing the root index and assisting in the cause diagnosis.
2. The method for evaluating the sustainable utilization of cultivated land resources based on geographic knowledge according to claim 1 is characterized in that in step S1, ontology layers are divided in a modularized mode by combining geographic knowledge, the ontology layers comprise a basic module, an index module and a scene module after being divided in a modularized mode, corpora such as a table, a textbook, a literature and a normative reference file related to a geographic knowledge system are collected according to the divided modules, a corpus is constructed, existing ontology libraries at home and abroad are collected, reusable ontologies and corpora are extracted from the corpus and the ontology libraries, structured data are obtained by packaging the ontologies and the corpora, the basic module, the index module and the scene module are modeled in a structuring mode by the structured data, and the multi-type information of the basic module, the entity module and the scene module is expressed in a formalized mode by using ontology construction tools, and top-down module integration is completed by mutual reference among the modules.
3. The method for evaluating the sustainable utilization of cultivated land resources based on geographic knowledge according to claim 1, wherein in step S2, materials such as natural resource related normative files, documents and reports are acquired by using a web crawler tool, basic geographic data such as administrative regions, geological features and the like and homeland space planning result data such as three regions and three lines are collected to form a database, examples contained in concepts in an ontology layer are mined and extracted from the database by using a deep learning model, relationships among the examples are extracted by using a template-based relationship extraction method, a concept attribute list defined in the ontology layer for the extracted examples is filled by a rule-based method according to the attributes of the examples and the module elements and the association relationships of the ontology layer, knowledge alignment is realized by using a similarity function based on the extracted examples, the attributes, the relationships and the like, mapping of the examples, the attributes, the relationships and the like and modes defined in the ontology layer is established, the examples are converted into RDF based on the mapping relationship, and the missing files in the RDF triples are complemented, and knowledge is obtained.
4. The geographical knowledge-based farmland resource sustainable utilization evaluation method according to claim 3, characterized in that natural language of users is subjected to syntactic analysis according to scene requirements, semantic dependency relations among vocabularies are obtained by adopting rule-based dependency syntactic analysis, a tree-like query graph Q is formed, sub-graph matching is performed by adopting a Multi-way Join method based on breadth first, sub-graph expansion is performed based on the knowledge graph structure, and sub-graphs containing attention points, criterion layers and indexes are obtained.
5. The method for evaluating the sustainable utilization of cultivated land resources based on geographical knowledge according to claim 1, wherein in step S4, a combined weighting model is fused, weights among points of interest, criterion layers and indexes in the subgraph are quantized, and the evaluation of the sustainable utilization of cultivated land resources is performed, and the method for obtaining the evaluation grade comprises the following steps:
s41: evaluation index set { M } in subgraph 1 ,M 2 ,…M n And n is a set of n indexes of the same level in the evaluation index system, and n is more than or equal to 2. Grading and grading all index items in the evaluation index set according to the types, the properties and the threshold values of the index items to obtain an index value grading matrix M= (M) 1 ，m 2 ，…，m n ) T Ensuring that all indexes are in the same dimension;
s42: calculating subjective weight of the evaluation index through quantitative analysis of determining importance degree between the sequence relation and the adjacent indexes;
s421: by DelPhenanthrene method, in index set { M 1 ,M 2 ,…，M n The most important index value M is obtained by screening 1 * And sequentially screening to obtain M 2 * After n-1 selections, the evaluation index set { M } can be determined 1 ,M 2 ,…，M n-1 ，M n Sequence relation index set { M } obtained by sequencing according to importance degree 1 * ,M 2 * ,…，M n-1 * ，M n * }。
S422: quantitative analysis of importance degree between adjacent indexes, and concentration of adjacent evaluation indexes M in sequence relation indexes n-1 * And M n * The degree of importance of (c) is quantified and expressed as:
Figure FDA0004008928470000031
wherein the method comprises the steps of
Figure FDA0004008928470000032
And->
Figure FDA0004008928470000033
Representing adjacent evaluation index->
Figure FDA0004008928470000034
And->
Figure FDA0004008928470000035
Weights of (2);
s423: index weight calculation according to a given r k Assigning a value to obtain an evaluation index
Figure FDA0004008928470000036
Weight of +.>
Figure FDA0004008928470000037
The calculation formula is as follows:
Figure FDA0004008928470000038
the r is i For a given r k The i-th value is assigned by assignment, so that the subjective weight of the sequence relation index set can be obtained, and further the subjective weight matrix X= (X) of the evaluation index set is obtained 1 ，x 1 ，…,x n ) T ；
S43: the objective weight of the evaluation index is calculated, the information entropy of the evaluation index is calculated, and the formula is as follows:
Figure FDA0004008928470000039
wherein e i Represents the information entropy of the ith evaluation index, p represents the number of evaluation objects, m ij An ith evaluation index value indicating a jth evaluation target;
calculating an evaluation index entropy weight, wherein the formula is as follows:
Figure FDA00040089284700000310
and then objective weight vector y= (Y) of the evaluation index set can be obtained 1 ，y 2 ，…,y n ) T And performing optimization fitting on the subjective weight and the objective weight vector to obtain a comprehensive weight matrix, wherein the formula is as follows:
Figure FDA00040089284700000311
wherein Z when the function H (Z) takes a minimum value i The obtained weight is the comprehensive weight matrix Z= (Z) 1 ，z 2 ，…,z n ) T . Weights Z of criterion layers i The sum of the index weights is the sum of the index weights;
s45: calculating an evaluation result score based on index scores and weights by combining a comprehensive factor score method and a restrictive factor method, grading the evaluation result score according to a natural discontinuous method, and multiplying an index score matrix M by a comprehensive weight matrix by adopting the comprehensive factor score method to obtain an evaluation result S, wherein the specific formula is as follows:
Figure FDA0004008928470000041
meanwhile, the combined restriction factor method is evaluated, and if the restriction factor index score is 0, the evaluation result is directly 0.
6. The method for evaluating sustainable utilization of cultivated land resources based on geographic knowledge according to claim 1, wherein in step S5, a main obstacle index is identified by using an obstacle degree model according to an evaluation grade, the main obstacle index is mapped back to a knowledge graph, the main obstacle index is used as a node to carry out deep traversal in the knowledge graph to obtain a directed acyclic graph with the main obstacle index as a vertex, the directed acyclic graph is optimized according to a rule design, the optimized directed acyclic graph is traversed to find out an index where the node with the degree of 0 is located, and the index is a root index, thereby realizing source tracing.
7. The arable land resource sustainable utilization evaluation system based on geographic knowledge is characterized by comprising a map construction module, wherein the map construction module comprises: the method comprises an incidence relation construction module, a subgraph extraction module, a calculation evaluation module and an inference diagnosis module, wherein the modules can realize the steps in any of the arable land resource sustainable utilization evaluation methods based on geographic knowledge in any of claims 1-6.
CN202211643785.5A 2022-12-20 2022-12-20 Evaluation method and system for sustainable utilization of cultivated land resources based on geographic knowledge Pending CN116187812A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211643785.5A CN116187812A (en) 2022-12-20 2022-12-20 Evaluation method and system for sustainable utilization of cultivated land resources based on geographic knowledge

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211643785.5A CN116187812A (en) 2022-12-20 2022-12-20 Evaluation method and system for sustainable utilization of cultivated land resources based on geographic knowledge

Publications (1)

Publication Number Publication Date
CN116187812A true CN116187812A (en) 2023-05-30

Family

ID=86443328

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211643785.5A Pending CN116187812A (en) 2022-12-20 2022-12-20 Evaluation method and system for sustainable utilization of cultivated land resources based on geographic knowledge

Country Status (1)

Country Link
CN (1) CN116187812A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117332856A (en) * 2023-11-03 2024-01-02 安徽国麒科技有限公司 Battery knowledge map abstract generation method based on sampling sub-graph strategy

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117332856A (en) * 2023-11-03 2024-01-02 安徽国麒科技有限公司 Battery knowledge map abstract generation method based on sampling sub-graph strategy
CN117332856B (en) * 2023-11-03 2024-02-23 安徽国麒科技有限公司 Battery knowledge map abstract generation method based on sampling sub-graph strategy

Similar Documents

Publication Publication Date Title
Janowicz et al. Know, Know Where, KnowWhereGraph: A densely connected, cross-domain knowledge graph and geo-enrichment service stack for applications in environmental intelligence
Wu et al. Mapping the knowledge domain of smart city development to urban sustainability: a scientometric study
CN112148890B (en) Teaching knowledge graph pedigree system based on network group intelligence
CN103207855B (en) For the fine granularity sentiment analysis system and method for product review information
CN107679221B (en) Time-space data acquisition and service combination scheme generation method for disaster reduction task
Yu Academic Performance Prediction Method of Online Education using Random Forest Algorithm and Artificial Intelligence Methods.
CN112163367B (en) Firefly algorithm and cellular automaton fused city expansion simulation prediction method
Kang et al. Towards landscape visual quality evaluation: methodologies, technologies, and recommendations
CN104598766B (en) Towards the space reasoning by cases method of Comprehensive Evaluation of Reservoirs
CN107784394A (en) Consider that the highway route plan of prospect theory does not know more attribute method for optimizing
CN115098696A (en) Method and device for constructing urban physical examination knowledge graph and storage medium
CN105335510A (en) Text data efficient searching method
CN113779264A (en) Trade recommendation method based on patent supply and demand knowledge graph
Xu et al. CET-4 score analysis based on data mining technology
CN116187812A (en) Evaluation method and system for sustainable utilization of cultivated land resources based on geographic knowledge
CN114819589A (en) Urban space high-quality utilization determination method, system, computer equipment and terminal
CN105160046A (en) Text-based data retrieval method
Liu et al. Evaluation of cultivated land quality using attention mechanism-back propagation neural network
Ross et al. A case-based reasoning system for conflict resolution: design and implementation
Jansen et al. Land-cover harmonisation and semantic similarity: some methodological issues
Li et al. Exploring firms’ innovation capabilities through learning systems
CN117010373A (en) Recommendation method for category and group to which asset management data of power equipment belong
Kim et al. A Study of the Definition and Components of Data Literacy for K-12 AI Education
Chou et al. Spatial knowledge databases as applied to the detection of changes in urban land use
Bai et al. Screening the stones of Venice: Mapping social perceptions of cultural significance through graph-based semi-supervised classification

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination