A method for constructing a knowledge graph of water and soil conservation measures in black soil regions

Abstract

The application discloses a kind of black soil region water and soil conservation measure configuration knowledge graph construction methods, comprising the following steps: S1, the construction of knowledge storage module;S2, the construction of knowledge expression module;S3, the construction of knowledge reasoning module;S4, the construction of knowledge display module.Through the present application, the water and soil conservation configuration knowledge of black soil region can form a knowledge graph.The knowledge graph is a visual display of related abstract concepts and information.By searching a certain spatio-temporal scene of soil erosion, the correlation between water and soil conservation measures and the scene, the benefit change or environmental change before and after the measure configuration, and the farmers' economic interests are obtained, thereby providing tool support for the popularization and technical promotion of water and soil conservation configuration knowledge.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of popular science and technical popularization of water and soil conservation configuration knowledge in black soil regions, in particular to a construction method of a knowledge graph of water and soil conservation configuration measures in black soil regions. BACKGROUND

[0002] Due to the combined action of natural and human factors, soil erosion is serious in the black soil region of Northeast China, leading to soil nutrient loss and soil fertility decline, thus causing deterioration of the ecological environment. In the black soil region, rainfall is concentrated in summer, mainly in the form of heavy rain, so rainwater runoff erosion is dominant in summer. In addition, there are wind erosion, snowmelt erosion, and freeze-thaw erosion. The topographic factors in the black soil region are divided into black soil flat river and hillside gully regions, and farming-pastoral ecotone regions, which increase the complexity of soil erosion types. Unreasonable human factors, such as over-reclamation, excessive fertilization, and the use of pesticides, also contribute to soil erosion. Finally, farmers in the black soil region lack knowledge of water and soil conservation configuration and scientific management concepts of soil, and there is an urgent need for popular science and technical popularization of water and soil conservation configuration knowledge. However, there is a lack of tools for abstract knowledge popularization and technical popularization, resulting in poor popularization and promotion effect and low efficiency. SUMMARY

[0003] In view of the above technical problems in the related art, the present application provides a construction method of a knowledge graph of water and soil conservation configuration measures in black soil regions, which can solve the above problems.

[0004] To achieve the above technical purposes, the technical solution of the present application is as follows:

[0005] A construction method of a knowledge graph of water and soil conservation configuration measures in black soil regions, comprising the following steps:

[0006] S1, construction of a knowledge storage module, first, determine the knowledge source, collect the knowledge, second, sort the knowledge, construct the black soil erosion scene, according to the suitable erosion scene type of water and soil conservation measures, mark, classify and arrange the water and soil conservation measure name and its related attributes, finally, store the sorted knowledge;

[0007] S2, construction of a knowledge expression module, first, extract the knowledge, provide raw materials and content for the construction of the knowledge graph, second, fuse different water and soil conservation configuration knowledge in black soil regions, make the implicit knowledge of water and soil conservation configuration explicit, finally, express the knowledge, adopt the form of entity-relation-entity, entity-attribute, condition[condition-[condition]-…]-entity for knowledge expression;

[0008] S3, the construction of the knowledge reasoning module, firstly, the knowledge of the event is extracted, and a knowledge multi-dimensional event vector A is formed, secondly, based on the third law of geography, fuzzy similar concepts are used for knowledge matching and reasoning, and the similarity between vector A and vector B is calculated based on the cosine similarity algorithm; finally, the calculated similarity is compared with the given similarity threshold value, and the similarity relationship between vector A and vector B is judged.

[0009] S4, the construction of the knowledge display module, the graph matching method is used for the visual display of the knowledge graph, and the nodes are represented by circles and the relationships are represented by solid lines in the visual interface.

[0010] Further, the knowledge collection in S1 includes collecting and analyzing different black soil area slope scale water and soil conservation measure types, measure names, descriptions, spatial suitability, measure implementation effect, suitable erosion scene, and measure cost related attribute data.

[0011] Further, the black soil area includes black soil alluvial and hilly areas, and the spatial suitability includes suitable slope, suitable slope position, land use type, wherein the suitable slope position includes ridge, uphill, middle slope, downhill, and valley, the suitable slope includes flat slope, gentle slope and steep slope, and the land use type includes farmland, forest land, grassland, water area and urban (engineering) land; the implementation effect of the measure includes water saving effect, sediment reduction effect and soil improvement.

[0012] Further, the black soil erosion scene in S1 includes hydraulic erosion, wind erosion and freeze-thaw erosion, according to the erosion scene type suitable for soil and water conservation measures, the potential measures and their configuration rules which can be preferentially selected are screened, wherein the configuration rules include the following steps: first, dividing according to the suitable slope position to obtain the first level regional division; second, dividing according to the slope, dividing the type area of suitable slope position according to the slope to obtain the second level regional division; then, dividing according to the land use type, dividing the second level regional division according to the land use type to obtain the third level regional division; finally, based on the attributes of erosion scene-suitable slope position-suitable slope-land use type-measure suitability, the rules are configured.

[0013] Further, the entity in S2 includes measure type, erosion scene, spatial location and time scale; the relationship in S2 includes belongs to, suitable for and contains; the attribute in S2 includes water and soil conservation measure attribute, erosion scene attribute, spatial location attribute and time scale attribute; the condition in S2 includes greater than, equal to, less than and contains.

[0014] Further, the formula of the cosine similarity algorithm in S3 is:

[0015]

[0016] wherein A represents a multi-dimensional event vector A (x1, y1, z1), B represents a multi-dimensional event vector B (x 2， y2, z2); x i , y i , z i (i = 1 or 2) is the knowledge extracted from the reasoning event, and the knowledge includes entities, attributes, and relationships.

[0017] Further, the nodes in the S4 include entities, attributes, and conditions.

[0018] The application has the following beneficial effects: the application can form a knowledge graph of water and soil conservation configuration knowledge in the black soil area, the knowledge graph is a visual display of related abstract concepts and information, by searching a certain spatio-temporal scene of soil erosion, the correlation between the water and soil conservation measures and the scene, the benefit change or environmental change before and after the measure configuration, and the economic benefit of the farmers themselves are obtained, thereby providing tool support for popularization and technical promotion of the water and soil conservation configuration knowledge. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings in the following description only constitute some embodiments of the application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0020] The application will be further described in detail below according to the drawings.

[0021] Figure 1 is a graph of the water and soil conservation measures in the hilly and gully area described in the embodiments of the application;

[0022] Figure 2 is a graph of the water and soil conservation measures in the hilly and gully area described in the embodiments of the application;

[0023] Figure 3 is a graph of the water and soil conservation measures in the hilly and gully area described in the embodiments of the application;

[0024] Figure 4 is a flow chart of the construction of the spatial configuration rules of the water and soil conservation measures on the slope scale in the black soil area described in the embodiments of the application;

[0025] Figure 5 is a schematic diagram of the construction and application of the knowledge graph of the spatial configuration of the water and soil conservation measures on the slope scale in the black soil area described in the embodiments of the application;

[0026] Figure 6A schematic diagram of the water and soil conservation measures mainly of hydraulic erosion, wind erosion and freeze-thaw erosion according to the embodiment of the present application is shown in the figure.

[0027] Figure 7 A spatial configuration layout of the slope scale water and soil conservation measures in the black soil region according to the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present application will be clearly and completely described in combination with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments in the present application belong to the scope of protection of the present application.

[0029] The construction method of the knowledge graph of the water and soil conservation measures in the black soil region according to the embodiment of the present application mainly includes a knowledge storage module, a knowledge expression module, a knowledge reasoning module and a knowledge display module, wherein the knowledge storage module includes knowledge collection, knowledge analysis (scene construction) and knowledge storage; the knowledge expression module includes knowledge extraction, knowledge fusion and knowledge expression; the knowledge reasoning module is based on the third law of geography, that is, the more similar the geographical environment is, the more similar the geographical features are, and the fuzzy similar concept is adopted to match and reason the knowledge; and the knowledge display module displays all the knowledge in the knowledge base and the reasoning results in the form of a graph.

[0030] In an embodiment of the present application, the determination of the knowledge source in the construction process of the knowledge storage module is very important, and the identification of the knowledge source of the slope scale water and soil conservation measures in the black soil region is the basis of knowledge extraction. It is a very important stage to determine the sources of structured data, semi-structured and unstructured data for constructing the knowledge graph. Taking the types of the slope scale water and soil conservation measures in the black soil hilly and gully region, the farming-pastoral ecotone region and the black soil hilly and gully region as examples, the knowledge sources are as follows: the existing attribute data of the slope scale water and soil conservation measures in the black soil region, the regularized measure configuration rules, the professional experience of the field experts (this part of knowledge is implicit and difficult to be specified, but can give guidance in the model evaluation stage), the related literature or cases of the slope scale water and soil conservation measures in the black soil region, etc. In this stage, the knowledge sources and the characteristic elements, the applicable scenarios of the water conservation measures are determined in cooperation with the field experts.

[0031] In the construction process of the knowledge storage module, the knowledge collection, that is, the collection and analysis of the types, names, descriptions, spatial suitability (suitable slope, suitable slope position, land use type) of the slope scale water and soil conservation measures in the black soil hilly and gully region, the farming-pastoral ecotone region and the black soil hilly and gully region, and the implementation effect (water saving effect, sediment reduction effect, soil improvement, etc.), suitable erosion scene, measure cost and other attribute data of the measures are as follows:

[0032] Table 1. Attributes of Soil and Water Conservation Measures at Slope Scale in the Black Soil Region

[0033]

[0034] Among them: in terms of slope, 1 represents a flat slope, 2 represents a gentle slope, and 3 represents a steep slope; in terms of suitable slope location, 1 represents a ridge, 2 represents an uphill slope, 3 represents a medium slope, 4 represents a downhill slope, and 5 represents a valley; in terms of water-saving effect and sediment reduction effect, 1 represents none, 2 represents weak, 3 represents moderate, and 4 represents strong.

[0035] After knowledge collection is completed, the knowledge needs to be organized and black soil erosion scenarios constructed, categorized as: water erosion, wind erosion, and freeze-thaw erosion. Based on the appropriate erosion scenario type for soil and water conservation measures, the names of the measures and their related attributes are marked, classified, and organized, such as... Figures 1-3 As shown, 15 soil and water conservation measures have been collected in the black soil plains and hilly areas, 11 in the hilly and gully areas, and 11 in the agro-pastoral transition areas, along with their attribute data. During the construction of black soil erosion scenarios, based on the suitable erosion scenario types for the soil and water conservation measures, potential measures and their configuration rules were selected for priority selection. The configuration rule base is spatially discretized according to the combination of "suitable slope position - suitable slope gradient - land use type". Based on the geomorphological characteristics of the black soil region, slope position types are described, including at least one of the following: ridge, uphill, middle slope, downhill, and valley; slope gradient types include at least one of the following: flat slope, gentle slope, and steep slope; land use types include at least one of the following: cultivated land, forest land, grassland, water area, urban (engineering) land, and unused land. The division steps are as follows: First, slope location division: The area is divided according to slope location to obtain the first-level regional division; second, slope gradient division: At least one of the slope location types is divided according to slope gradient to obtain the second-level regional division; then, land use type division: The second-level regional division is further divided according to land use type to obtain the third-level regional division. Based on the above three-level regional division method, the "slope location-slope gradient-land use" combination type is determined. For example... Figure 4 As shown, the final configuration is based on the attributes of "erosion scenario - suitable slope position - suitable slope gradient - land use type - suitability of measures". Finally, the organized knowledge is stored in the form of relational database tables.

[0036] In one embodiment of the present application, for the construction of the knowledge expression module, firstly, knowledge extraction is carried out, mainly including: named entity recognition, relationship (or soil conservation measure attribute) extraction, event (or erosion scene) extraction. The process of further discovering the implicit knowledge on the basis of knowledge extraction is the key step of knowledge graph modeling, which provides the original materials and content for the construction of knowledge graph; secondly, the knowledge of soil conservation measure configuration in different black soil regions (black soil alluvial and hilly gully region, farming and pasturing interlaced region) is fused (on the basis of knowledge source determination and knowledge extraction, the required domain knowledge for constructing the knowledge graph has been obtained, including measure attribute and configuration term, rule and instance, etc. In this stage, the knowledge is expressed by ontology technology. Among them, the measure attribute and configuration term are described by network ontology language OWL, which clearly defines the class, attribute and instance of the measure knowledge; the rule is described by semantic web rule language SWRL, which supports intelligent reasoning. The knowledge graph expressed by semantic expression is stored and managed by Neo4j knowledge base), the implicit knowledge of soil conservation measure configuration is made explicit, that is, the display labeling of knowledge and cases such as "measure-attribute-suitability (condition)-erosion scene", which greatly simplifies the later knowledge expression process; finally, the knowledge is expressed in the form of (entity-relation-entity), (entity-attribute), (condition [condition-[condition]-…]-entity) and the like. Among them, the entity includes: measure type, erosion scene, spatial location, time scale and the like; the relationship includes: belongs to, is suitable for, contains and the like; the attribute includes: soil conservation measure attribute, erosion scene attribute, spatial location attribute, time scale attribute and the like; the condition includes: greater than, equal to, less than, contains and the like logical information.

[0037] In one embodiment of the present application, in the construction of the knowledge reasoning module, firstly, the entity, attribute and relationship knowledge of the reasoning event (the question raised by the user) is extracted, forming a multi-dimensional event vector such as event vector A (x1, y1, z1, …); secondly, based on the third law of geography, the more similar the geographical environment is, the more similar the geographical features are, a fuzzy similar concept is adopted, vectors A (x1, y1, z1) and B (x2, y2, z2) represent two events, the cosine similarity algorithm is used to calculate the similarity between vector A and vector B, and the calculation formula is as follows:

[0038]

[0039] The cosine of the included angle is in the range of [-1, 1], the greater the cosine of the included angle indicates the smaller the included angle of two vectors, the smaller the cosine of the included angle indicates the greater the included angle of two vectors, and the cosine of the included angle takes the maximum value 1 when the directions of two vectors coincide, that is, the more similar two vectors are, the greater the similarity f between two BMPs is, and vice versa, finally, based on the similarity threshold given by an expert, such as 0.5, if the similarity of two event vectors is greater than the threshold, then it is determined that the two event vectors are similar, in the relationship of the event vector A, it is applicable to the vector B, and vice versa; if no event vector greater than the threshold is searched in the knowledge base, it is determined that there is no similar event vector. Based on the above assumptions, the question event A proposed by the user, such as a query of the measure configuration, the measure configuration scenario and scheme of the event vector B are applicable to the question event A, that is, the answer to the question, otherwise, no answer to the question is found.

[0040] In an embodiment of the present application, as shown in Figures 5-7 The present application can realize the visual display of the knowledge graph by using Neo4j Desktop software, for example, the relationship of the slope scale soil and water conservation measures in the black soil region mainly subjected to water erosion, wind erosion, snowmelt erosion and freeze-thaw erosion, which can be queried and displayed by the user, the visual display of the knowledge graph adopts a graph matching method, which is a logical operation based on search matching, and as long as the stored knowledge and knowledge reasoning results in the present knowledge graph meet the requirements, they are displayed in the form of a graph. In the visual interface, the nodes (entities, attributes, conditions, etc.) are represented by circles, and the relationships (belonging to, applicable to, containing, etc.) are represented by solid lines.

[0041] The above only describes the preferred embodiments of the present application and is not intended to limit the present application, and any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1.A method for constructing a knowledge graph of water and soil conservation measures in black soil regions, characterized in that, The method comprises the following steps: S1, construction of a knowledge storage module, first determining a knowledge source, collecting knowledge, secondly, carding the knowledge, constructing a black soil erosion scene, marking, classifying and arranging the name of the soil and water conservation measure and its related attributes according to the suitable erosion scene type of the soil and water conservation measure, and finally storing the carded knowledge; The knowledge collection in S1 comprises collecting and carding and analyzing different black soil region slope scale soil and water conservation measure types, measure names, descriptions, spatial suitability, measure implementation effects, suitable erosion scenes and measure cost related attribute data; The black soil region comprises a black soil hilly and gully region, a hilly and gully region and a farming and pasturing interlaced region; the spatial suitability comprises suitable slope, suitable slope position and land use type, wherein the suitable slope position comprises a ridge, an upper slope, a middle slope, a lower slope and a valley, the suitable slope comprises a flat slope, a gentle slope and a steep slope, and the land use type comprises farmland, forest land, grassland, water area and urban land; the measure implementation effect comprises water saving effect, sediment reduction effect and soil improvement; The black soil erosion scene in S1 comprises hydraulic erosion, wind erosion and freeze-thaw erosion, and according to the suitable erosion scene type of the soil and water conservation measure, potential measures that can be preferentially selected and their configuration rules are screened, wherein the configuration rules comprise the following steps: first, dividing according to the suitable slope position to obtain a first level regional division; second, dividing according to the slope to divide the type region of the suitable slope position according to the slope to obtain a second level regional division; then, dividing according to the land use type to divide the second level regional division according to the land use type to obtain a third level regional division; and finally, based on the attributes of the erosion scene-suitable slope position-suitable slope-land use type-measure suitability, the rules are configured; S2, construction of a knowledge expression module, first extracting the knowledge to provide original materials and content for the construction of a knowledge graph, secondly, fusing the configuration knowledge of different black soil region soil and water conservation measures to make the implicit knowledge of the soil and water conservation measure configuration explicit, and finally expressing the knowledge in the form of entity-relation-entity, entity-attribute and condition[condition-[condition]-…]-entity; S3, construction of a knowledge reasoning module, first extracting the knowledge of the reasoning event to form a knowledge multi-dimensional event vector A, secondly, based on the third law of geography, using fuzzy similar concepts to match and reason the knowledge, and based on the cosine similarity algorithm to calculate the similarity between the vector A and the matching vector B; and finally comparing the calculated similarity with a given similarity threshold to determine the similarity relationship between the vector A and the vector B; S4, construction of a knowledge display module, using a graph matching method to visually display the knowledge graph, and in the visual interface, using a circle to represent a node and a solid line to represent a relationship. 2.The method of claim 1, wherein, The entity in S2 includes measure type, erosion scene, spatial position, and time scale; the relationship in S2 includes belonging, being applicable to, and containing; the attribute in S2 includes soil and water conservation measure attribute, erosion scene attribute, spatial position attribute, and time scale attribute; and the condition in S2 includes being greater than, being equal to, being less than, and containing. 3.The method of claim 1, wherein, The formula of the cosine similarity algorithm in S3 is: ， where A represents a multidimensional event vector A (x 1 , y 1 , z 1 B(x multidimensional event vector The node in S4 includes entity, attribute, and condition. 2 , y 2 , z 2 )；x i 、 y i 、 z i For the extracted knowledge of the reasoning event, i = 1 or 2, the knowledge includes entities, attributes, and relationships. 4.The method of claim 1, wherein, ​

Images