A landslide questionnaire logic consistency automatic checking method and system based on a knowledge graph

Abstract

This invention provides a method and system for automatically checking the logical consistency of landslide survey forms based on knowledge graphs. The method includes: constructing a landslide knowledge graph; the landslide knowledge graph contains various types of target entities, target attributes of each type of target entity, and constraint relationships between different target entities; mapping the landslide survey form to be inspected based on the landslide knowledge graph to obtain an initial instantiated subgraph of the landslide survey form; performing logical checks on the initial instantiated subgraph based on a preset rule engine, a preset graph matching algorithm, and the constraint relationships in the landslide knowledge graph to obtain the logical conflict check results of the landslide survey form; and generating a structured report based on the logical conflict check results and a preset prompt template library. The solution provided by this invention can achieve automated and intelligent conflict detection of landslide survey forms, while improving detection efficiency and accuracy.

Description

Technical Field

[0001] This invention relates to the field of geological disaster investigation and data processing technology, and in particular to an automatic method and system for checking the logical consistency of landslide survey forms based on knowledge graphs. Background Technology

[0002] Currently, landslide field investigations mainly rely on technicians filling out structured survey forms (Excel or Word) on-site, containing dozens of fields such as basic landslide information, topography, geological structure, landslide characteristics, and deformation activity. After the investigation is completed, the forms need to be manually checked for logical contradictions, factual errors, inconsistencies, and other problems. However, manual self-checking has the following shortcomings:

[0003] 1. The survey form has many fields and complex relationships. Manually checking each item is time-consuming and labor-intensive, with an average of more than 15 minutes required for a single form, resulting in low efficiency.

[0004] 2. Some contradictions can only be discovered by combining multiple parameters (such as the mismatch between the length-to-thickness ratio of the landslide body and the landslide type, or the inconsistency between the stability coefficient and the deformation signs), which are easily overlooked by humans.

[0005] 3. Different technicians have different levels of experience, and the quality of self-inspection varies from person to person, lacking a unified standard.

[0006] Existing auxiliary inspection methods mostly employ fixed rules (such as if-then rules), which can only check pre-defined simple logical relationships. They cannot handle multi-parameter cross-constraints and deep contradictions based on geological mechanisms, nor can they be extended to new inspection rules. Therefore, there is an urgent need for a method that can automatically, comprehensively, and intelligently check the logical consistency of landslide survey forms. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide a method and system for automatic logical consistency checking of landslide survey forms based on knowledge graphs, so as to realize the automated and intelligent contradiction detection of logical consistency of landslide survey forms, and at the same time improve the detection efficiency and accuracy of complex constraints in landslide survey forms.

[0008] To address the aforementioned technical problems, embodiments of the present invention provide an automatic logical consistency check method for landslide survey forms based on knowledge graphs, comprising:

[0009] Construct a landslide knowledge graph; the landslide knowledge graph contains multiple different types of target entities, target attributes of each type of target entity, and constraint relationships between different target entities; each target entity corresponds to a target node in the landslide knowledge graph.

[0010] The landslide knowledge graph is used to map the landslide survey form to be inspected, so as to obtain an initial instantiated subgraph of the landslide survey form to be inspected; the initial instantiated subgraph contains multiple different types of instance entities in the landslide survey form to be inspected, instance attributes corresponding to each instance entity, and association edges between different instance entities; each instance entity corresponds to an instance node in the initial instantiated subgraph.

[0011] Based on the preset rule engine, preset graph matching algorithm and the constraint relationship in the landslide knowledge graph, the initial instantiated subgraph is subjected to logical check processing to obtain the logical conflict check result of the landslide investigation form to be checked.

[0012] Based on the logical conflict check results and the preset prompt template library, a structured report is generated, which includes the location of the logical conflict, the type of logical conflict, and modification suggestions.

[0013] In one embodiment, the landslide knowledge graph is used to map the landslide survey form to be inspected, in order to obtain an initial instantiated subgraph of the landslide survey form to be inspected, including:

[0014] The landslide investigation form is read, and each field in the landslide investigation form is matched with the target entity in the landslide knowledge graph to obtain the instance entity corresponding to each field and the key-value pair set corresponding to each instance entity; in the key-value pair set, the key in each key-value pair represents an attribute, and the value represents the content of the attribute.

[0015] Based on the constraint relationships between different target entities in the landslide knowledge graph, the instance entities corresponding to each field, and the set of key-value pairs corresponding to each instance entity, the initial instantiation subgraph is generated.

[0016] In one embodiment, based on a preset rule engine, a preset graph matching algorithm, and the constraints in the landslide knowledge graph, a logical check is performed on the initially instantiated subgraph to obtain the logical conflict check results of the landslide survey form to be checked, including:

[0017] The landslide knowledge graph is invoked through the preset rule engine, and a logical check is performed on the initial instantiated subgraph based on the constraint relationships in the landslide knowledge graph to obtain a check result and a simplified subgraph; the number of instance nodes in the simplified subgraph is less than the number of instance nodes in the initial instantiated subgraph.

[0018] Based on the preset graph matching algorithm and the landslide knowledge graph, a second logical check is performed on the simplified subgraph to obtain the logical conflict check result.

[0019] In one embodiment, the landslide knowledge graph is invoked through the preset rule engine, and a logical check is performed on the initially instantiated subgraph based on the constraints in the landslide knowledge graph to obtain a check result and a simplified subgraph, including:

[0020] Traverse each instance entity in the initial instantiation subgraph, and based on the constraint relationship, check the current instance entity for one or more of the following: the instance entity's required fields, the numerical range instance attributes, and the instance entity's two-parameter instance attributes, in order to determine the status label of each instance entity and the check result for one instance entity that does not satisfy the constraint relationship.

[0021] Based on the state label of each instance entity, the instance entities in the initial instantiation subgraph are filtered to determine the first instance entity, and the simplified subgraph is constructed based on the first instance entity.

[0022] In one embodiment, based on the preset graph matching algorithm and the landslide knowledge graph, a secondary logical check is performed on the simplified subgraph to obtain the logical conflict check result, including:

[0023] Based on the standard pattern subgraph in the landslide knowledge graph, the simplified subgraph isomorphic search is performed to obtain the search results; the search results include a set of local contradictions; the set of local contradictions represents the set of abnormal instance attribute values ​​existing in the simplified subgraph.

[0024] Based on the meta-path template in the landslide knowledge graph, semantic contradiction detection is performed on the simplified subgraph to determine the semantic contradiction detection result of the simplified subgraph. The semantic contradiction detection result indicates that the instance entity sequence formed by the first instance entity in the simplified subgraph contains logical semantic contradictions.

[0025] Based on the search processing results and the semantic contradiction detection results, the logical conflict check results are determined.

[0026] In one embodiment, based on the standard pattern subgraph in the landslide knowledge graph, a subgraph isomorphic search process is performed on the simplified subgraph to determine the search result, including:

[0027] Based on the target attributes of the target entity in the standard pattern subgraph and the instance attributes of the first instance entity in the simplified subgraph, the node matching likelihood between the target entity in the standard pattern subgraph and the first instance entity in the simplified subgraph is determined.

[0028] Based on the node matching likelihood of the target entity in the standard pattern subgraph and the first instance entity in the simplified subgraph, the overall matching likelihood of the simplified subgraph and the standard pattern subgraph is determined; the overall matching likelihood is used to measure whether there is a subgraph structure in the simplified subgraph that is isomorphic to the standard pattern subgraph, and the instance attribute values ​​of each first instance entity in the subgraph structure satisfy the constraints of the standard pattern subgraph.

[0029] The search processing result is determined based on the overall matching likelihood and the likelihood threshold.

[0030] In one embodiment, determining the node matching likelihood between the target entity in the standard pattern subgraph and the first instance entity in the simplified subgraph, based on the entity attributes of the target entity in the standard pattern subgraph and the instance attributes of the first instance entity in the simplified subgraph, includes:

[0031] The node matching likelihood is determined using the following formula:

[0032]

[0033] in, This represents the node matching likelihood between the target entity u and the first instance entity v in the standard pattern subgraph t. This represents the attribute similarity function between the target entity u and the first instance entity v; This represents the weight of the k-th target attribute of the target entity u; This represents the preset value of the k-th target attribute of target entity u in the standard pattern subgraph t; This represents the actual value of the instance attribute corresponding to the first instance entity v in the simplified subgraph.

[0034] In one embodiment, semantic contradiction detection is performed on the simplified subgraph based on the meta-path template in the landslide knowledge graph to determine the semantic contradiction detection result of the simplified subgraph, including:

[0035] Based on the meta-path template in the landslide knowledge graph, the simplified subgraph is searched to determine at least one instance path in the simplified subgraph; the instance path represents the sequence of instance entities formed by the first instance entity in the simplified subgraph.

[0036] Based on the consistency score and score threshold of the instance path, the instance path containing semantic logical contradictions is identified, and the instance path is identified as the semantic contradiction detection result.

[0037] In one embodiment, the consistency score of the instance path is obtained by the following formula:

[0038]

[0039] in, This represents the consistency score of the instance path P obtained based on the current metapath template r. This represents two adjacent first instance entities in instance path P. The degree of matching; Represents the global consistency factor; , They represent the weights, and ; n represents the number of entities in the first instance.

[0040] Embodiments of the present invention also provide an automatic logical consistency check system for landslide survey forms based on knowledge graphs, comprising:

[0041] The first construction module is used to construct a landslide knowledge graph; the landslide knowledge graph contains multiple different types of target entities, target attributes of each type of target entity, and constraint relationships between different target entities; each target entity corresponds to a target node in the landslide knowledge graph.

[0042] The second construction module is used to perform mapping processing on the landslide investigation form to be inspected based on the landslide knowledge graph, so as to obtain an initial instantiation subgraph of the landslide investigation form to be inspected; the initial instantiation subgraph contains multiple different types of instance entities in the landslide investigation form to be inspected, instance attributes corresponding to each instance entity, and association edges between different instance entities; each instance entity corresponds to an instance node in the initial instantiation subgraph.

[0043] The processing module is used to perform logical checks on the initial instantiated subgraph based on a preset rule engine, a preset graph matching algorithm, and the constraint relationships in the landslide knowledge graph to obtain the logical conflict check results of the landslide investigation form to be inspected; based on the logical conflict check results and a preset prompt template library, a structured report is generated, which includes the logical conflict location, logical conflict type, and modification suggestions.

[0044] The above-described solution of the present invention has at least the following beneficial effects:

[0045] 1. This invention constructs a landslide knowledge graph specific to the landslide field, that is, transforms the landslide formation mechanism and engineering geological experience into a structured target entity-target attribute-constraint network, so as to explicitly express the multi-parameter cross constraints (such as length-thickness ratio, stability coefficient-deformation matching, etc.) in the landslide field. At the same time, this landslide knowledge graph serves as a standardized knowledge base for subsequent inspection of the landslide investigation form, which can improve the inspection efficiency and accuracy of complex constraint relationships in the landslide investigation form.

[0046] 2. When checking the landslide investigation form, the constraints in the landslide knowledge graph are used to perform a logical check on the initial instantiated subgraph based on the preset rule engine to achieve preliminary screening and avoid invalid searches in the future; further, a second logical check is performed on the results after the first logical check based on the preset graph matching algorithm to improve the efficiency and accuracy of the check.

[0047] It should be understood that the implementation of any embodiment of the present invention does not mean that it will simultaneously possess or achieve multiple or all of the above-mentioned beneficial effects. Attached Figure Description

[0048] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0049] The structures, proportions, sizes, etc. illustrated in this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed herein, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0050] Figure 1 This is a flowchart illustrating the automatic logical consistency check method for landslide survey forms based on knowledge graphs provided in an embodiment of the present invention.

[0051] Figure 2 This is a schematic diagram of a landslide knowledge graph provided in an optional embodiment of the present invention;

[0052] Figure 3 This is a schematic diagram of an initial instantiation subgraph provided in an optional embodiment of the present invention;

[0053] Figure 4 This is a schematic block diagram of an automatic logical consistency check system for landslide survey forms based on knowledge graphs, provided in an embodiment of the present invention.

[0054] Figure 5 This is a schematic block diagram of an electronic device provided in an embodiment of the present invention; and

[0055] Figure 6 This is a schematic block diagram of a computing device provided in an embodiment of the present invention. Detailed Implementation

[0056] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0057] It should be understood that the terms "comprising / including," "consisting of," or any other variations are intended to cover non-exclusive inclusion, such that a product, apparatus, process, or method that comprises a list of elements includes not only those elements but may also include, where necessary, other elements not expressly listed, or elements inherent to such a product, apparatus, process, or method. Without further limitation, an element defined by the phrases "comprising / including," "consisting of," does not exclude the presence of additional identical elements in the product, apparatus, process, or method that includes said element.

[0058] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0059] like Figure 1 As shown, embodiments of the present invention propose an automatic logical consistency check method for landslide survey forms based on knowledge graphs, including:

[0060] Step 11, construct a landslide knowledge graph; the landslide knowledge graph contains various types of target entities, target attributes of each type of target entity, and constraint relationships between different target entities; each target entity corresponds to a target node in the landslide knowledge graph;

[0061] Step 12: Based on the landslide knowledge graph, the landslide survey form to be inspected is mapped to obtain the initial instantiated subgraph of the landslide survey form to be inspected; the initial instantiated subgraph contains various types of instance entities in the landslide survey form to be inspected, the instance attributes corresponding to each instance entity, and the association edges between different instance entities; each instance entity corresponds to an instance node in the initial instantiated subgraph.

[0062] Step 13: Based on the preset rule engine, preset graph matching algorithm and constraint relationships in the landslide knowledge graph, perform logical check processing on the initial instantiated subgraph to obtain the logical conflict check results of the landslide survey form to be checked.

[0063] Step 14: Based on the logical conflict check results and the preset prompt template library, generate a structured report. The structured report includes the location of the logical conflict, the type of logical conflict, and modification suggestions.

[0064] In this embodiment, the target entity may include basic landslide information, geological features, hydrological conditions, morphological features, characteristics of the landslide body or slip zone, deformation stability, hazard assessment, prevention and control measures, and threat level. The basic landslide information may include target attributes such as landslide type (push-moving, traction-moving, composite, etc.), main sliding direction (0°~360°), volume, and scale level (small, medium, large, extra-large, etc.). Geological characteristics can include target attributes such as stratigraphic lithology (sandstone, mudstone, silty clay, etc.), dip direction, dip angle, number of joint and fracture groups, and type of control structure surface; hydrological conditions can include target attributes such as groundwater type (pore water, fracture water, karst water, etc.), groundwater depth, and recharge type (rainfall, surface water, artificial, etc.); morphological characteristics can include target attributes such as length, width, thickness, area, volume, slope, aspect, and slope height; sliding body or sliding zone characteristics can include target attributes such as sliding body lithology (clay, gravelly soil, rock blocks, etc.), sliding zone soil name (silty clay, gravelly clay, etc.), sliding surface dip angle, and sliding surface thickness. Deformation stability can include target attributes such as deformation stage (creep, acceleration, failure, rest, etc.), stability coefficient, displacement rate, and crack width; hazard assessment can include target attributes such as current stability status (stable, basically stable, understability, unstable), development trend (tending towards stability, tending towards aggravation, basically stable, etc.), and hazard level (small, medium, large, extra-large, etc.); prevention and control measures can include target attributes such as measure classification (engineering treatment, monitoring and early warning, relocation and avoidance, etc.) and specific measures (anti-slide piles, interception and drainage, mass monitoring and prevention, etc.); and threat disaster can include target attributes such as the number of threatened households, the number of threatened people, and the number of threatened properties.

[0065] Constraints between different target entities can include numerical range constraints, two-parameter matching constraints, multi-parameter cross constraints, and cross-entity consistency constraints. Each type of constraint contains multiple different constraint rules. In a specific implementation example, numerical range constraints can be defined as follows: a reasonable range for the "main sliding direction" is 0°~360°; for the "slip surface angle," it is 10°~45° for soil landslides and 15°~50° for rock landslides. Two-parameter matching constraints can be defined as follows: "landslide type" corresponds to "slip surface morphology"—traction landslides should be "steep at the top and gentle at the bottom" or "linear"; shove landslides should be "gentle at the top and steep at the bottom" or "arc-shaped." Multi-parameter cross constraints can be defined as follows: the length-to-thickness ratio (length / thickness) of the sliding body is usually greater than 30 for traction landslides, less than 30 for shove landslides, and between 10 and 50 for composite landslides. Cross-entity consistency constraints can be implemented as follows: "Current stable status" and "Development trend analysis" should be semantically consistent (stable → tending to be stable); "volume" and "scale level" should be matched according to industry standards (volume less than 100,000 m³ is small, volume between 100,000 m³ and 1,000,000 m³ is medium, volume between 1,000,000 m³ and 10,000,000 m³ is large, and volume greater than 10,000,000 m³ is extra-large).

[0066] Here, the constructed landslide knowledge graph is as follows: Figure 2 As shown, the type of the target entity can be used as a node label, and the target entity is a target node in the graph. Constraints are connected between different target entities in the form of directed constraint edges. Edge labels can be added to the directed constraint edges to indicate the type of constraint relationship, and conditional expressions (involving target attributes within each target entity) can also be attached to the directed constraint edges. This landslide knowledge graph serves as a standardized knowledge base for subsequent inspection of landslide investigation forms, which can improve the efficiency and accuracy of inspecting complex constraint relationships in the landslide investigation forms.

[0067] The initial instantiated subgraph is a data graph dynamically generated based on the landslide knowledge graph and the landslide survey form to be inspected. Here, the type of the landslide survey form to be inspected is not specifically limited; it can be an Excel or Word file. The instance entities (instance nodes) in the initial instantiated subgraph correspond to the specific fields in the survey form. The edges between instance entities are only connection edges set based on experience and do not have any constraint relationships.

[0068] Here, the preset rule engine is a pre-stored component that calls the landslide knowledge graph. It stores various basic inspection rules for landslide survey forms, such as "reasonableness of slip surface dip angle." When checking landslide survey forms, the preset rule engine calls the constraint relationships in the landslide knowledge graph and its stored basic inspection rules to perform a preliminary check on the initially instantiated subgraph. This quickly filters out obvious anomalies, reduces the search space of the subsequent preset graph matching algorithm, avoids invalid searches, and thus improves inspection efficiency. Furthermore, the results of the preliminary check by the preset rule engine are further checked using the preset graph matching algorithm to check for deep logical conflicts that violate geological laws, thereby improving inspection accuracy and avoiding duplicate checks.

[0069] In an optional embodiment of the present invention, step 12 above may include:

[0070] Step 121: Read the landslide survey form to be inspected, and match each field in the landslide survey form with the target entity in the landslide knowledge graph to obtain the instance entity corresponding to each field and the set of key-value pairs corresponding to each instance entity; in the set of key-value pairs, the key in each key-value pair represents the attribute, and the value represents the content of the attribute.

[0071] Step 122: Based on the constraint relationships between different target entities in the landslide knowledge graph, the instance entities corresponding to each field, and the set of key-value pairs corresponding to each instance entity, generate an initial instantiation subgraph.

[0072] In this embodiment, for an Excel-formatted landslide survey form, the fields are located directly by worksheet, row, and column; for a Word-formatted landslide survey form, the location of key fields (such as the text after "landslide type") can be identified first based on a preset form template.

[0073] Furthermore, each field in the landslide investigation form is matched with the target entity and target attribute under the target entity in the landslide knowledge graph (here, the type and name of the instance entity are basically the same as those of the target entity, and the matching between the instance attribute and the target attribute can be based on cross-entity consistency constraints, while using a thesaurus to handle the naming differences between the two, such as "landslide length" matching "length", etc.). Fields that successfully match the target entity are identified as instance entities, and fields that successfully match the target attribute are identified as instance attributes. Since the target attribute belongs to the corresponding target entity, the instance attribute can be assigned to the corresponding instance entity after successful matching, and the specific content of the instance attribute is added to the instance node formed by the instance entity.

[0074] Furthermore, based on the constraints between different target entities in the landslide knowledge graph, association edges are added between different instance entities with related relationships (specifically, an empirical association edge is added from the "landslide basic information instance" pointing to the "morphological feature instance," a derived association edge is added from the "morphological feature instance" pointing to the "hazard assessment instance," and a causal edge is added from the "deformation activity instance" pointing to the "hazard assessment instance") to obtain the initial instantiated subgraph of the landslide investigation form to be inspected (e.g., Figure 3 (As shown).

[0075] In an optional embodiment of the present invention, step 13 above may include:

[0076] Step 131: The landslide knowledge graph is invoked through the preset rule engine, and a logical check is performed on the initial instantiated subgraph based on the constraint relationships in the landslide knowledge graph to obtain the first check result and a simplified subgraph; the number of instance nodes in the simplified subgraph is less than the number of instance nodes in the initial instantiated subgraph. Here, the first logical check on the initial instantiated subgraph based on the preset rule engine includes the check of all basic constraints to avoid repeated searches during the second logical check by the subsequent preset graph matching algorithm, thereby improving the check efficiency.

[0077] Specifically, step 131 above may include:

[0078] Step 1311: Traverse each instance entity in the initial instantiation subgraph and, based on the constraint relationship, check one or more of the following for the current instance entity: the required fields of the instance entity, the instance attributes of the numerical range, and the instance attributes of the instance entity with two parameters, in order to determine the status label of each instance entity and the result of a check that does not meet the constraint relationship.

[0079] Step 1312: Based on the state label of each instance entity, filter the instance entities in the initial instantiation subgraph to determine the first instance entity, and construct a simplified subgraph based on the first instance entity.

[0080] When performing a check on required fields for instance entities, it is necessary to traverse the instance attributes under each instance entity, identify missing key fields (such as "smooth surface pattern" not filled in), and generate a check failure result of "key field missing".

[0081] When performing numerical range instance attribute checks, the numerical range of instance attributes (main sliding direction, sliding surface inclination angle, etc.) under each instance entity can be compared with the numerical constraint range defined in the landslide knowledge graph based on the pre-stored constraint functions in the preset rule engine. A warning is issued if the range is exceeded. Here, the pre-stored constraint function is a general function, and its specific form can be... This pre-stored constraint function only performs numerical comparisons, where, Indicates the range of values ​​for instance properties. These represent the numerical constraint ranges of the corresponding target attributes defined in the landslide knowledge graph. After the preset rule engine receives the initial instantiated subgraph, it simultaneously reads the numerical constraint ranges of the corresponding target attributes from the landslide knowledge graph, performs general logical judgments (comparison, search, existence judgment), and generates corresponding successful check results (numerical range matching) or failed check results (numerical range mismatch).

[0082] When performing a two-parameter instance attribute matching check within an instance entity, for two directly related instance attributes within the same instance entity (such as landslide type → slip surface morphology), the predefined matching rules in the landslide knowledge graph are queried. If the two instance attributes match, a check success result is generated; otherwise, a check failure result is generated.

[0083] Since there are multiple different types of instance attributes under an instance entity, when any of the above checks for the corresponding instance entity generates a check failure result, a check failure status label is added to the instance entity. When all of the above checks for the corresponding instance entity generate check success results, a check success status label is added to the instance entity. Furthermore, based on the status label of each instance entity, the instance entities in the initial instantiation subgraph are filtered. That is, the instance entities marked with the check failure status label are removed from the initial instantiation subgraph, and the instance entities marked with the check success status label are retained. The instance entity marked with the check success status label is determined as the first instance entity. At the same time, based on the association edges between instance entities in the initial instance subgraph, association edges are added between the first instance entities to obtain a simplified subgraph.

[0084] A logical check is performed on the initial instantiated subgraph by a preset rule engine to quickly filter out obvious anomalies in the initial instantiated subgraph, reduce the search space of subsequent preset graph matching algorithms, avoid invalid searches, and thus improve the checking efficiency.

[0085] In an optional embodiment of the present invention, step 13 above may include:

[0086] Step 132: Based on the preset graph matching algorithm and the landslide knowledge graph, a secondary logical check is performed on the simplified subgraph to obtain the logical conflict check results. Here, the preset graph matching algorithm can be the VF2 subgraph isomorphic search algorithm.

[0087] Specifically, step 132 above may include:

[0088] Step 1321: Based on the standard pattern subgraph in the landslide knowledge graph, perform subgraph isomorphic search processing on the simplified subgraph to obtain the search processing result; the search processing result may include a set of local contradictions; the set of local contradictions represents the set of abnormal instance attribute values ​​existing in the simplified subgraph.

[0089] In this embodiment, the standard pattern subgraph is a predefined graph structure in the landslide knowledge graph used to express local, short-distance constraints. This standard pattern subgraph typically involves 2-3 types of target entities and direct constraint edges between these entities. Simultaneously, the standard pattern subgraph corresponds to one or a set of closely related domain experience rules. For example, in a standard pattern subgraph composed of landslide basic information and morphological features, the constraint edges between them can constrain the length-to-thickness ratio and the empirical relationship between landslide type, while the constraint edges also include the length-to-thickness ratio calculation formula and various comparison conditions. It should be noted that the landslide knowledge graph can contain multiple different standard pattern subgraphs. When performing subgraph isomorphic search processing, different standard pattern subgraphs need to be searched sequentially within the simplified subgraph.

[0090] When performing subgraph isomorphic search on a simplified subgraph, the simplified subgraph can be used as the target graph, and a substructure graph isomorphic to the standard pattern subgraph can be searched on it. Specifically, the subgraph isomorphic search process can be to find whether there is an associated edge between two first instance entities on the target graph, and whether the type of the associated edge is the same as the edge label of the constraint edge between two target entities of the same type in the standard pattern subgraph.

[0091] Since each target entity in the standard pattern subgraph has at most one type, while the first instance entity in the simplified subgraph has multiple types, the isomorphic search process of this subgraph can actually be reduced to directly locating the node corresponding to each first instance entity in the simplified subgraph. Here, it is first necessary to pair the target entities in the standard pattern subgraph with the first instance entities in the simplified subgraph, that is, to filter out the target entities and first instance entities of the same type in the standard pattern subgraph and the simplified subgraph to form a set of isomorphic node mappings in the subgraph.

[0092] Furthermore, step 1321 above may include:

[0093] Step 13211: Based on the target attribute of the target entity in the standard pattern subgraph and the instance attribute of the first instance entity in the simplified subgraph, determine the node matching likelihood between the target entity in the standard pattern subgraph and the first instance entity in the simplified subgraph. Here, the node matching likelihood between the target entity and the first instance entity can be calculated directly based on the preset value of the target attribute in the corresponding standard pattern subgraph and the actual value of the instance attribute of the first instance entity. It should be understood that the type of the target entity for calculating the node matching likelihood should be completely consistent with the type of the first instance entity, that is, the two are successfully paired.

[0094] Preferably, the node matching likelihood can be determined using the following formula:

[0095]

[0096] in, This represents the node matching likelihood between the target entity u and the first instance entity v in the standard pattern subgraph t. This represents the attribute similarity function between the target entity u and the first instance entity v; This represents the weight of the k-th target attribute of the target entity u; This represents the preset value (or ideal value) of the k-th target attribute of the target entity u in the standard pattern subgraph t. This represents the actual value of the instance attribute corresponding to the first instance entity v in the simplified subgraph. Here, the calculation is based on the k-th target attribute, which should mean that the k-th attribute exists in both the target entity u and the first instance entity v, and that the k-th attribute has the same type.

[0097] Step 13212: Based on the node matching likelihood of the target entity in the standard pattern subgraph and the first instance entity in the simplified subgraph, determine the overall matching likelihood of the simplified subgraph and the standard pattern subgraph; the overall matching likelihood is used to measure whether there is a subgraph structure in the simplified subgraph that is isomorphic to the standard pattern subgraph, and the instance attribute values ​​of each first instance entity in the subgraph structure satisfy the constraints of the standard pattern subgraph.

[0098] Step 13213: Determine the search processing result based on the overall matching likelihood and the likelihood threshold.

[0099] Preferably, the overall matching likelihood between the simplified subgraph and the standard pattern subgraph can be expressed as:

[0100]

[0101] in, represents the overall matching likelihood between the simplified subgraph and the standard pattern subgraph t; S represents the set of subgraph isomorphic node mappings under the standard pattern subgraph t.

[0102] When the overall matching likelihood between the standard pattern subgraph t and the simplified subgraph is less than the corresponding likelihood threshold, it is determined that there is a contradiction between the simplified subgraph and the current standard pattern subgraph t. The first instance entity with the same target entity type as the current standard pattern subgraph t is marked as a local contradiction. After sequentially searching and processing all standard pattern subgraphs, all local contradictions are integrated to obtain the set of local contradictions.

[0103] In an optional embodiment of the present invention, step 132 above may include:

[0104] Step 1322: Based on the meta-path template in the landslide knowledge graph, perform semantic contradiction detection on the simplified subgraph to determine the semantic contradiction detection result of the simplified subgraph; the semantic contradiction detection result indicates that the instance entity sequence formed by the first instance entity in the simplified subgraph contains logical semantic contradictions.

[0105] Here, the meta-path template is used to describe the multi-step relationships between target entity types in the landslide knowledge graph. The meta-path template consists of a series of target entity types and the relationship edge types between adjacent target entity types (for example, a meta-path template can be represented as: Landslide Basic Information - [Control] → Geological Features - [Impact] → Morphological Features - [Determination] → Hazard Assessment). It can express multi-hop semantic chains with geological significance in domain knowledge. As a query template, this meta-path template can check whether the sequence of instance entities formed by the first instance entity in the simplified subgraph satisfies the domain logical semantic consistency. The sequence of instance entities in the simplified subgraph is a set of actually existing entity instances. The sequence of instance entities in the simplified subgraph can be arranged sequentially according to the entity type order specified in the meta-path template, and there are corresponding relationship edges between adjacent instances.

[0106] It should be noted that the landslide knowledge graph contains multiple different meta-path templates. When performing semantic contradiction checks, the simplified subgraph needs to be checked sequentially through different meta-path templates.

[0107] Specifically, step 1322 above may include:

[0108] Step 13221: Based on the meta-path template in the landslide knowledge graph, perform search processing on the simplified subgraph to determine at least one instance path in the simplified subgraph; the instance path represents the sequence of instance entities formed by the first instance entity in the simplified subgraph.

[0109] Step 13222: Based on the consistency score of the instance path and the score threshold, identify the instance path containing semantic logical contradictions and determine the instance path as the semantic contradiction detection result.

[0110] In this embodiment, for each metapath template, all instance entity sequences that match the metapath template are searched in the simplified subgraph. Specifically, for the current metapath template, a first instance entity (usually only one) of the same type as the starting target entity of the current metapath template is found in the simplified subgraph and used as the starting node. Starting from the starting node, the search proceeds along the specified edge of the same type as the current metapath template, and the first instance entities of the same order and type as the second, third, ..., and last target entities on the metapath template are determined in sequence until the last target entity on the metapath template is successfully found. At this point, all the first instance entities reached by the search are arranged in order to obtain the instance path in the simplified subgraph corresponding to the current metapath template. It should be understood that when performing search processing based on the metapath template, all target entities on the metapath template need to be searched to form a complete instance path. When any target entity on the metapath template is not searched, the search processing of the metapath template is no longer performed, and a field missing search result is immediately generated.

[0111] For any instance path found, preferably, its corresponding consistency score can be calculated using the following formula:

[0112]

[0113] in, This represents the consistency score of the instance path P obtained based on the current metapath template r. This represents two adjacent first instance entities in instance path P. The degree of matching; Represents the global consistency factor; , They represent the weights, and ; n represents the number of entities in the first instance.

[0114] Here, the global consistency factor This global consistency factor can be calculated based on a joint empirical formula of the internal attributes of multiple first instance entities along the instance path P (e.g., the joint distribution of length-to-thickness ratio, slope height, and stability coefficient); It is also predefined by the landslide knowledge graph, ranging from 0 to 1.

[0115] Preferably, two adjacent first instance entities in instance path P Matching degree It can be done through the formula: Calculated; where, Indicates an indicator function; Represents the first instance entity The attribute values ​​that participate in the constraints; Represents the first instance entity The attribute values ​​that participate in the constraints; Represents the first instance entity Pre-set reliability; Represents the first instance entity The pre-set reliability.

[0116] If the consistency score of instance path P in the simplified subgraph under the current metapath template r is less than the corresponding consistency score threshold, then instance path P is determined to have a semantic logical contradiction. After sequentially searching and processing all metapath templates, the semantic contradiction detection result of the simplified subgraph is obtained.

[0117] In an optional embodiment of the present invention, step 132 above may include:

[0118] Step 1323: Based on the search processing results and semantic contradiction detection results, determine the logical conflict check results.

[0119] Specifically, the search processing results can be directly merged with the semantic contradiction detection results to obtain the logical conflict check results.

[0120] The method provided by the above embodiments of the present invention transforms landslide formation mechanisms and engineering geological experience into a structured entity-attribute-constraint network and implements it in the form of a landslide knowledge graph. This explicitly expresses multi-parameter cross-constraints (such as length-thickness ratio, stability coefficient-deformation matching, etc.) in the landslide field, thereby improving the reasoning efficiency of complex constraints. Based on the constraint relationships in the landslide knowledge graph, a basic check is performed through a preset rule engine, and a secondary check is further performed based on a preset graph matching algorithm. This automatically discovers deep logical conflicts that violate geological laws, such as abnormal length-thickness ratios, missing dihedral slope identification, mismatch between stability coefficients and deformation signs, and discrepancies between volume and scale levels, which are complex contradictions dependent on geological mechanisms. On the one hand, this improves the inspection efficiency of landslide survey forms, and on the other hand, it also improves the inspection accuracy and comprehensiveness.

[0121] like Figure 4 As shown, embodiments of the present invention also provide an automatic logical consistency check system 30 for landslide survey forms based on knowledge graphs, which may include:

[0122] The first construction module 31 is used to construct a landslide knowledge graph; the landslide knowledge graph contains various types of target entities, target attributes of each type of target entity, and constraint relationships between different target entities; each target entity corresponds to a target node in the landslide knowledge graph.

[0123] The second construction module 32 is used to map the landslide survey form to be inspected based on the landslide knowledge graph to obtain an initial instantiation subgraph of the landslide survey form to be inspected; the initial instantiation subgraph contains various types of instance entities in the landslide survey form to be inspected, instance attributes corresponding to each instance entity, and association edges between different instance entities; each instance entity corresponds to an instance node in the initial instantiation subgraph.

[0124] Processing module 33 is used to perform logical checks on the initial instantiated subgraph based on the preset rule engine, preset graph matching algorithm and constraint relationships in the landslide knowledge graph to obtain the logical conflict check results of the landslide investigation form to be checked; based on the logical conflict check results and the preset prompt template library, a structured report is generated, which includes the logical conflict location, logical conflict type and modification suggestions.

[0125] It should be noted that this system corresponds to the above-mentioned automatic logical consistency check method for landslide survey forms based on knowledge graphs. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0126] like Figure 5 As shown, embodiments of the present invention also provide an electronic device 50, comprising: a memory 51 for storing one or more computer programs; and one or more processors 52 for executing the one or more computer programs. When the computer programs are run by the processors, they execute the aforementioned automatic logical consistency check method for a knowledge graph-based landslide survey form. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects. The electronic device 50 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components, their connections and relationships, and their functions shown in this invention are merely examples and are not intended to limit the implementation of the invention described and / or claimed herein.

[0127] like Figure 6As shown, electronic device 50 is a computing device or computer system, which may include CPU 501 (computing unit), which can perform various appropriate actions and processes according to a computer program stored in ROM 502 (read-only memory) or a computer program loaded from storage unit 508 into random access RAM 503 (memory). RAM 503 may also store various programs and data required for the operation of device 500. CPU 501, ROM 502, and RAM 503 are interconnected via bus 504. I / O interface 505 (input / output interface) is also connected to bus 504.

[0128] Multiple components in electronic device 500 are connected to I / O interface 505, including: input unit 506, such as keyboard, mouse, etc.; output unit 507, such as various types of monitors, speakers, etc.; storage unit 508, such as disk, optical disk, etc.; and communication unit 509, such as network card, modem, wireless transceiver, etc. Communication unit 509 allows device 500 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0129] CPU 501 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of CPU 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. CPU 501 performs the various methods and processes described above. For example, in some embodiments, the knowledge graph-based automatic logical consistency check method for landslide survey forms can be implemented as a computer software program tangibly contained in a computer-readable storage medium, such as storage unit 508. In some embodiments, part or all of the computer program can be loaded and / or installed on device 500 via ROM 502 and / or communication unit 509. When the computer program is loaded into RAM 503 and executed by CPU 501, one or more steps of the knowledge graph-based automatic logical consistency check method for landslide survey forms described above can be performed. Alternatively, in other embodiments, CPU 501 may be configured by any other suitable means (e.g., by means of firmware) to perform an automatic logical consistency check method for a knowledge graph-based landslide survey form.

[0130] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the methods described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0131] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0132] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0133] In the embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0134] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0135] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0136] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0137] Furthermore, it should be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent solutions of the present invention. Moreover, the steps performing the above series of processes can naturally be executed in the order described, but are not necessarily required to be executed in chronological order; some steps can be executed in parallel or independently of each other. Those skilled in the art will understand that all or any step or component of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or network of computing devices, in hardware, firmware, software, or a combination thereof. This is something that those skilled in the art can achieve by using their basic programming skills after reading the description of the present invention.

[0138] Therefore, the object of the present invention can also be achieved by running a program or a set of programs on any computing device. The computing device can be a known general-purpose device. Therefore, the object of the present invention can also be achieved simply by providing a program product containing program code for implementing the method or apparatus. That is, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any known storage medium or any storage medium developed in the future. It should also be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent to the present invention. Furthermore, the steps for performing the above series of processes can naturally be performed in the order described, but are not necessarily required to be performed in chronological order. Some steps can be performed in parallel or independently of each other.

[0139] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for automatically checking the logical consistency of landslide survey forms based on knowledge graphs, characterized in that, include: Construct a landslide knowledge graph; the landslide knowledge graph contains multiple different types of target entities, target attributes of each type of target entity, and constraint relationships between different target entities; Each target entity corresponds to a target node in the landslide knowledge graph; The landslide knowledge graph is used to map the landslide survey form to be inspected in order to obtain the initial instantiation subgraph of the landslide survey form to be inspected. The initial instantiated subgraph contains various types of instance entities in the landslide investigation table, instance attributes corresponding to each instance entity, and association edges between different instance entities; each instance entity corresponds to an instance node in the initial instantiated subgraph. Based on the preset rule engine, preset graph matching algorithm and the constraint relationship in the landslide knowledge graph, the initial instantiated subgraph is subjected to logical check processing to obtain the logical conflict check result of the landslide investigation form to be checked. Based on the logical conflict check results and the preset prompt template library, a structured report is generated, which includes the location of the logical conflict, the type of logical conflict, and modification suggestions.

2. The method for automatic logical consistency checking of landslide survey forms based on knowledge graphs according to claim 1, characterized in that, Based on the landslide knowledge graph, the landslide survey form to be inspected is mapped to obtain an initial instantiated subgraph of the landslide survey form to be inspected, including: The landslide investigation form is read, and each field in the landslide investigation form is matched with the target entity in the landslide knowledge graph to obtain the instance entity corresponding to each field and the key-value pair set corresponding to each instance entity; in the key-value pair set, the key in each key-value pair represents an attribute, and the value represents the content of the attribute. Based on the constraint relationships between different target entities in the landslide knowledge graph, the instance entities corresponding to each field, and the set of key-value pairs corresponding to each instance entity, the initial instantiation subgraph is generated.

3. The method for automatic logical consistency checking of landslide survey forms based on knowledge graphs according to claim 1, characterized in that, Based on a preset rule engine, a preset graph matching algorithm, and the constraints in the landslide knowledge graph, the initial instantiated subgraph undergoes logical checks to obtain the logical conflict check results of the landslide survey form to be checked, including: The landslide knowledge graph is invoked through the preset rule engine, and a logical check is performed on the initial instantiated subgraph based on the constraint relationships in the landslide knowledge graph to obtain a check result and a simplified subgraph; the number of instance nodes in the simplified subgraph is less than the number of instance nodes in the initial instantiated subgraph. Based on the preset graph matching algorithm and the landslide knowledge graph, a second logical check is performed on the simplified subgraph to obtain the logical conflict check result.

4. The method for automatic logical consistency checking of landslide survey forms based on knowledge graphs according to claim 3, characterized in that, The landslide knowledge graph is invoked through the preset rule engine, and a logical check is performed on the initially instantiated subgraph based on the constraints in the landslide knowledge graph to obtain a check result and a simplified subgraph, including: Traverse each instance entity in the initial instantiation subgraph, and based on the constraint relationship, check the current instance entity for one or more of the following: the instance entity's required fields, the numerical range instance attributes, and the instance entity's two-parameter instance attributes, in order to determine the status label of each instance entity and the check result for one instance entity that does not satisfy the constraint relationship. Based on the state label of each instance entity, the instance entities in the initial instantiation subgraph are filtered to determine the first instance entity, and the simplified subgraph is constructed based on the first instance entity.

5. The method for automatic logical consistency checking of landslide survey forms based on knowledge graphs according to claim 3, characterized in that, Based on the preset graph matching algorithm and the landslide knowledge graph, a secondary logical check is performed on the simplified subgraph to obtain the logical conflict check result, including: Based on the standard pattern subgraph in the landslide knowledge graph, the simplified subgraph isomorphic search is performed to obtain the search results; the search results include a set of local contradictions; the set of local contradictions represents the set of abnormal instance attribute values ​​existing in the simplified subgraph. Based on the meta-path template in the landslide knowledge graph, semantic contradiction detection is performed on the simplified subgraph to determine the semantic contradiction detection result of the simplified subgraph. The semantic contradiction detection result indicates that the instance entity sequence formed by the first instance entity in the simplified subgraph contains logical semantic contradictions. Based on the search processing results and the semantic contradiction detection results, the logical conflict check results are determined.

6. The method for automatic logical consistency checking of landslide survey forms based on knowledge graphs according to claim 5, characterized in that, Based on the standard pattern subgraph in the landslide knowledge graph, a subgraph isomorphic search process is performed on the simplified subgraph to determine the search results, including: Based on the target attributes of the target entity in the standard pattern subgraph and the instance attributes of the first instance entity in the simplified subgraph, the node matching likelihood between the target entity in the standard pattern subgraph and the first instance entity in the simplified subgraph is determined. Based on the node matching likelihood of the target entity in the standard pattern subgraph and the first instance entity in the simplified subgraph, the overall matching likelihood of the simplified subgraph and the standard pattern subgraph is determined; the overall matching likelihood is used to measure whether there is a subgraph structure in the simplified subgraph that is isomorphic to the standard pattern subgraph, and the instance attribute values ​​of each first instance entity in the subgraph structure satisfy the constraints of the standard pattern subgraph. The search processing result is determined based on the overall matching likelihood and the likelihood threshold.

7. The method for automatic logical consistency checking of landslide survey forms based on knowledge graphs according to claim 6, characterized in that, Based on the entity attributes of the target entity in the standard pattern subgraph and the instance attributes of the first instance entity in the simplified subgraph, the node matching likelihood between the target entity in the standard pattern subgraph and the first instance entity in the simplified subgraph is determined, including: The node matching likelihood is determined using the following formula: ； in, This represents the node matching likelihood between the target entity u and the first instance entity v in the standard pattern subgraph t. This represents the attribute similarity function between the target entity u and the first instance entity v; This represents the weight of the k-th target attribute of the target entity u; This represents the preset value of the k-th target attribute of target entity u in the standard pattern subgraph t; This represents the actual value of the instance attribute corresponding to the first instance entity v in the simplified subgraph.

8. The method for automatic logical consistency checking of landslide survey forms based on knowledge graphs according to claim 5, characterized in that, Based on the meta-path templates in the landslide knowledge graph, semantic contradiction detection is performed on the simplified subgraph to determine the semantic contradiction detection results of the simplified subgraph, including: Based on the meta-path template in the landslide knowledge graph, the simplified subgraph is searched to determine at least one instance path in the simplified subgraph; the instance path represents the sequence of instance entities formed by the first instance entity in the simplified subgraph. Based on the consistency score and score threshold of the instance path, the instance path containing semantic logical contradictions is identified, and the instance path is identified as the semantic contradiction detection result.

9. The method for automatic logical consistency checking of landslide survey forms based on knowledge graphs according to claim 8, characterized in that, The consistency score of the instance path is obtained by the following formula: ； in, This represents the consistency score of the instance path P obtained based on the current metapath template r. This represents two adjacent first instance entities in instance path P. The degree of matching; Represents the global consistency factor; , They represent the weights, and ; n represents the number of entities in the first instance.

10. An automatic logical consistency check system for landslide survey forms based on knowledge graphs, characterized in that, include: The first construction module is used to construct a landslide knowledge graph; the landslide knowledge graph contains multiple different types of target entities, target attributes of each type of target entity, and constraint relationships between different target entities; each target entity corresponds to a target node in the landslide knowledge graph. The second construction module is used to map the landslide investigation form to be inspected based on the landslide knowledge graph to obtain the initial instantiation subgraph of the landslide investigation form to be inspected. The initial instantiated subgraph contains various types of instance entities in the landslide investigation table, instance attributes corresponding to each instance entity, and association edges between different instance entities; each instance entity corresponds to an instance node in the initial instantiated subgraph. The processing module is used to perform logical checks on the initial instantiated subgraph based on a preset rule engine, a preset graph matching algorithm, and the constraint relationships in the landslide knowledge graph, so as to obtain the logical conflict check results of the landslide investigation form to be checked. Based on the logical conflict check results and the preset prompt template library, a structured report is generated, which includes the location of the logical conflict, the type of logical conflict, and modification suggestions.

Images