Elevator operation and maintenance retrieval enhanced question and answer method, device, equipment and storage medium

Abstract

This invention discloses an elevator maintenance retrieval enhancement question-answering method, apparatus, device, and storage medium. By introducing parameterized hyperedges, this invention overcomes the limitations of binary relations in ordinary knowledge graphs, enabling it to adapt to elevator maintenance retrieval enhancement question-answering scenarios with multiple entities and constraints, thus improving the accuracy of elevator maintenance retrieval enhancement question-answering and elevator maintenance work orders. Furthermore, this invention integrates semantic similarity and consistency scores to obtain a comprehensive score for each parameterized hyperedge in the candidate hyperedge set, and determines the top K target parameterized hyperedges with the highest comprehensive scores. This ensures that the elevator maintenance consultation text is semantically similar to the text entries of the target parameterized hyperedges, and that the on-site observation parameters match the working conditions represented by the constraint set of the target parameterized hyperedges, thereby improving the accuracy of the generated elevator maintenance work orders.

Description

Technical Field

[0001] This invention relates to the field of intelligent retrieval and knowledge engineering, and particularly to a method, apparatus, device, and storage medium for enhancing elevator operation and maintenance retrieval and question answering. Background Technology

[0002] The mainstream approach for engineering knowledge-based question answering and maintenance decision-making still centers on "Retrieval Enhanced Generation (RAG)": documents are segmented, vector representations are calculated, similarity-based recall is performed, and then a large model generates the answer. While this paradigm is easily implemented in general question answering, it exposes three main problems in elevator maintenance scenarios. First, segment-level representation ignores the structural and constraint relationships between items. For example, handling the same fault often requires simultaneously meeting multiple conditions such as "part model, operating conditions, construction time window, personnel qualifications, and applicable clauses." Relying solely on similarity to match text often results in false recalls of "semantically similar but with inconsistent operating conditions," affecting the first-time repair rate. Second, segment boundaries lead to fragmented facts; thresholds, intervals, and logical conditions are easily disconnected from the steps, and the generation process is "self-assembled" by the model, resulting in insufficient stability and traceability. Third, the output is mostly descriptive suggestions, lacking directly executable steps and acceptance criteria for "how to determine if it's repaired," making it difficult to serve as a basis for on-site delivery.

[0003] To compensate for structural deficiencies, some solutions introduce knowledge graphs or graph-based RAGs to enhance context reconstruction through entity and relation modeling. While knowledge graphs have advantages in entity linking and path reasoning, they mostly use binary relations. When faced with complex facts involving multiple entities and numerical conditions, they have to be broken down into multiple edges and then "pieced back together" by paths, which can easily lead to information loss and dilution of confidence. Summary of the Invention

[0004] This invention provides an elevator operation and maintenance retrieval enhancement question-and-answer method, device, equipment, and storage medium to adapt to elevator operation and maintenance retrieval enhancement question-and-answer scenarios with multiple entities and multiple constraints, thereby improving the accuracy of elevator operation and maintenance retrieval enhancement questions and answers and elevator maintenance work orders.

[0005] In a first aspect, the present invention provides an enhanced question-and-answer method for elevator operation and maintenance retrieval, comprising:

[0006] Obtain elevator maintenance consultation text input by the user, and extract on-site observation parameters from the elevator maintenance consultation text;

[0007] The parameterized hyperedges that match the elevator maintenance consultation text are searched from the elevator operation and maintenance knowledge hypergraph as a candidate set of hyperedges. The elevator operation and maintenance knowledge hypergraph is constructed based on the multi-dimensional relationships between entities in elevator operation and maintenance related documents. The elevator operation and maintenance knowledge hypergraph includes multiple parameterized hyperedges used to describe text entries in the elevator operation and maintenance related documents. Each parameterized hyperedge includes a first entity set composed of multiple associated entities and a constraint set used to represent the relationships between the entities.

[0008] Calculate the semantic similarity score between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set;

[0009] Calculate the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the candidate hyperedge set;

[0010] By combining the semantic similarity score and the consistency score, a comprehensive score is obtained for each parameterized hyperedge in the hyperedge candidate set, and the top K target parameterized hyperedges with the highest comprehensive scores are determined.

[0011] For each of the target parameterized hyperedges, the parameters of the constraint set in the target parameterized hyperedge are mapped to the elevator maintenance work order template to obtain the elevator maintenance work order.

[0012] Optionally, the parameterized hyperedges matching the elevator maintenance consultation text are searched from the elevator operation and maintenance knowledge hypergraph as a candidate set of hyperedges, including:

[0013] Extract a second entity set consisting of multiple entities from the elevator maintenance consultation text;

[0014] The parameterized hyperedges corresponding to the first entity set that matches the second entity set are used as the first hyperedge candidate set from the elevator operation and maintenance knowledge hypergraph by keyword matching.

[0015] Calculate the semantic similarity between the elevator maintenance consultation text and the text entries of each parameterized hyperedge in the elevator operation and maintenance knowledge hypergraph, and take the parameterized hyperedges with the highest semantic similarity as the second hyperedge candidate set;

[0016] The first and second hyperedge candidate sets are merged to obtain the hyperedge candidate set.

[0017] Optionally, the constraint set includes soft constraints, which are measurable parameters. The consistency score between the observed parameters and the constraint set of each parameterized hyperedge in the candidate hyperedge set is calculated, including:

[0018] For each parameterized hyperedge in the candidate hyperedge set, calculate the absolute value of the difference between each field observation parameter and the corresponding measurable parameter;

[0019] The degree of deviation between the absolute value and the measurement scale of the field observation parameter is obtained by calculating the ratio between the field observation parameter and the measurable parameter.

[0020] The weighted violation degree is obtained by multiplying the violation degree by the corresponding weight.

[0021] Calculate the sum of the weighted violation degrees corresponding to all the field observation parameters to obtain the total violation degree of the constraint set of the field observation parameters and the parameterized hyperedge;

[0022] The consistency score is calculated by taking the value of a natural exponential function with the negative of the sum of the aforementioned violations as the exponent.

[0023] Optionally, the constraint set includes hard constraints, and after obtaining the elevator maintenance consultation text input by the user and extracting the field observation parameters from the elevator maintenance consultation text, it also includes:

[0024] For each of the field observation parameters, determine whether the field observation parameter satisfies the hard constraint of the parameterized hyperedge;

[0025] If not, the parameterized hyperedge is excluded from the hyperedge candidate set.

[0026] Optionally, the parameterized hyperedge also includes metadata describing the source and update time of the text entry. After searching the elevator operation and maintenance knowledge hypergraph for the parameterized hyperedge matching the elevator maintenance consultation text as a candidate set of hyperedges, the method further includes:

[0027] The source confidence score of each parameterized hyperedge in the candidate hyperedge set is calculated based on the source and update time of the text entry.

[0028] Optionally, by combining the semantic similarity score and the consistency score, a comprehensive score is obtained for each parameterized hyperedge in the hyperedge candidate set, including:

[0029] The comprehensive score of each parameterized hyperedge in the candidate hyperedge set is obtained by multiplying the semantic similarity score, the consistency score, and the source confidence score.

[0030] Optionally, for each of the target parameterized hyperedges, the parameters of the constraint set in the target parameterized hyperedge are mapped to the elevator maintenance work order template to obtain an elevator maintenance work order, including:

[0031] The elevator maintenance work order template containing the skeleton of the standard operating procedure for elevator maintenance is found using the first entity set in the target parameterized hyperedge.

[0032] By using field mapping, the parameters of the constraint set in the target parameterized hyperedge are mapped to the corresponding fields in the elevator maintenance work order template to obtain the elevator maintenance work order.

[0033] Secondly, the present invention also provides an elevator operation and maintenance retrieval enhanced question-and-answer device, comprising:

[0034] The observation parameter extraction module is used to obtain the elevator maintenance consultation text input by the user and extract the on-site observation parameters from the elevator maintenance consultation text;

[0035] The hyperedge candidate set determination module is used to find the parameterized hyperedges that match the elevator maintenance consultation text from the elevator operation and maintenance knowledge hypergraph as the hyperedge candidate set. The elevator operation and maintenance knowledge hypergraph is constructed based on the multi-dimensional relationships between entities in elevator operation and maintenance related documents. The elevator operation and maintenance knowledge hypergraph includes multiple parameterized hyperedges used to describe the text entries in the elevator operation and maintenance related documents. The parameterized hyperedges include a first entity set composed of multiple associated entities and a constraint set used to represent the relationships between the entities.

[0036] The similarity score calculation module is used to calculate the semantic similarity score between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set;

[0037] The consistency score calculation module is used to calculate the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set;

[0038] The scoring and integration module is used to integrate the semantic similarity score and the consistency score to obtain the comprehensive score of each parameterized hyperedge in the hyperedge candidate set, and to determine the top K target parameterized hyperedges with the highest comprehensive scores.

[0039] The maintenance work order generation module is used to map the parameters of the constraint set in each of the target parameterized hyperedges to the elevator maintenance work order template to obtain the elevator maintenance work order.

[0040] Thirdly, the present invention also provides an electronic device, comprising:

[0041] One or more processors;

[0042] Storage device for storing one or more programs;

[0043] When the one or more programs are executed by the one or more processors, the one or more processors implement the elevator operation and maintenance retrieval enhanced question-answering method as described in the first aspect of the present invention.

[0044] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the elevator operation and maintenance retrieval enhanced question-and-answer method as described in the first aspect of the present invention.

[0045] The elevator maintenance retrieval enhancement question-answering method provided in this embodiment of the invention, after obtaining the elevator maintenance consultation text input by the user, extracts on-site observation parameters from the elevator maintenance consultation text, and searches for parameterized hyperedges matching the elevator maintenance consultation text in the elevator maintenance knowledge hypergraph as a hyperedge candidate set. The elevator maintenance knowledge hypergraph is constructed based on the multi-dimensional relationships between entities in elevator maintenance-related documents. The elevator maintenance knowledge hypergraph includes multiple parameterized hyperedges describing text entries in elevator maintenance-related documents. Each parameterized hyperedge includes a first entity set consisting of multiple associated entities and a constraint set representing the relationships between entities. The method calculates the text of the elevator maintenance consultation text and the text of each parameterized hyperedge in the hyperedge candidate set. The semantic similarity score of the entries is calculated, and the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set is calculated. The semantic similarity score and the consistency score are combined to obtain the comprehensive score of each parameterized hyperedge in the hyperedge candidate set, and the top K target parameterized hyperedges with the highest comprehensive scores are determined. For each target parameterized hyperedge, the parameters of the constraint set in the target parameterized hyperedge are mapped to the elevator maintenance work order template to obtain the elevator maintenance work order. This invention, by introducing parameterized hyperedges, solves the limitations of the binary relations of ordinary knowledge graphs, and can adapt to the elevator operation and maintenance retrieval enhanced question answering scenario with multiple entities and multiple constraints, thereby improving the accuracy of elevator operation and maintenance retrieval enhanced question answering and elevator maintenance work orders. Furthermore, in this embodiment of the invention, the semantic similarity score between the elevator maintenance consultation text and the text entries of each parameterized hyperedge in the hyperedge candidate set, and the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set are calculated respectively. The comprehensive score of each parameterized hyperedge in the hyperedge candidate set is obtained by combining the semantic similarity score and the consistency score, and the top K target parameterized hyperedges with the highest comprehensive scores are determined. This ensures that the elevator maintenance consultation text and the text entries of the target parameterized hyperedges are semantically similar, and that the working conditions represented by the field observation parameters and the constraint set of the target parameterized hyperedges are consistent, thereby improving the accuracy of the generated elevator maintenance work orders.

[0046] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 A flowchart of an enhanced question-and-answer method for elevator operation and maintenance retrieval provided by the present invention;

[0049] Figure 2 A flowchart of another elevator operation and maintenance retrieval enhanced question-and-answer method provided by the present invention;

[0050] Figure 3 This is a schematic diagram of the structure of an elevator operation and maintenance retrieval enhancement question and answer device provided by the present invention;

[0051] Figure 4 This is a schematic diagram of the structure of an electronic device provided by the present invention.

[0052] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0053] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0054] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0055] Figure 1 This is a flowchart of an elevator maintenance retrieval enhancement question-and-answer method provided by the present invention. This embodiment implements elevator maintenance retrieval enhancement question-and-answer based on parameterized hyperedges to generate elevator maintenance work orders. This method can be executed by the elevator maintenance retrieval enhancement question-and-answer device provided by the present invention. This device can be implemented in software and / or hardware, and is typically configured in electronic devices, such as... Figure 1 As shown, the elevator operation and maintenance retrieval enhanced question-and-answer method includes the following steps:

[0056] S101. Obtain the elevator maintenance consultation text input by the user, and extract the on-site observation parameters from the elevator maintenance consultation text.

[0057] In this embodiment of the invention, the user can be an elevator maintenance worker. When an elevator malfunction occurs, the user can input a consultation question through a terminal device, generating an elevator maintenance consultation text. After receiving the elevator maintenance consultation text input by the user, on-site observation parameters can be extracted from the elevator maintenance consultation text using methods such as keyword matching or fuzzy matching. On-site observation parameters may include the elevator's door closing time, door operator current, door operation frequency, ambient temperature, elevator load factor, etc., which are not limited herein.

[0058] S102. Find parameterized hyperedges that match the elevator maintenance consultation text from the elevator operation and maintenance knowledge hypergraph as the hyperedge candidate set.

[0059] In this embodiment of the invention, an elevator operation and maintenance knowledge hypergraph can be pre-constructed based on the multivariate relationships between entities in elevator operation and maintenance related documents. The elevator operation and maintenance knowledge hypergraph includes multiple parametric hyperedges (PHEs) describing text entries in the elevator operation and maintenance related documents. Each parametric hyperedge includes a first entity set composed of multiple associated entities and a constraint set representing the relationships between entities. In a typical graph structure, an edge can only connect two points, and an edge is actually a symbol used to represent an adjacency relationship between two points. However, an edge in a hypergraph can associate (or contain) two or more points; that is, all points on the same edge are adjacent, i.e., a hyperedge. In this embodiment of the invention, the hyperedges in the elevator operation and maintenance knowledge hypergraph are parameterized to form parametric hyperedges. Each parametric hyperedge includes a first entity set composed of multiple associated entities (or nodes) and a constraint set representing the relationships between entities. The parametric hyperedge packages multiple related entities and computable constraints in the same record to answer the question, "In these objects / scenarios, when these constraints are satisfied, how should the operation be performed and how should acceptance be carried out?" Entities can include the following types:

[0060] Equipment or components, such as "door operator", "light curtain", "controller", etc.;

[0061] Model / Series: such as "DO-X v2";

[0062] Scenario: such as "residential building ≤ 20 floors", "winter", "strong winds pressuring the entrance floor";

[0063] Qualifications / Resources: such as "Level II Maintenance Qualification" and "Torque Tester";

[0064] Compliance clauses: such as "GB×××× / §5.2.1 (applicable to residential scenarios)".

[0065] The constraints in the constraint set may include measurable parameters (range / threshold / OR logic, such as CloseTime, DoorCurrent, DutyCycle, etc.), time windows, power outages / elevator shutdowns, qualifications / key tools, applicable terms, process preferences, resource optimization, etc., which are not limited in this invention.

[0066] After extracting field observation parameters from elevator maintenance consultation texts, parameterized hyperedges matching the elevator maintenance consultation texts are searched in the elevator operation and maintenance knowledge hypergraph as a candidate set of hyperedges. Matching methods can include semantic similarity matching, keyword matching, etc., and this invention does not limit the specific methods used.

[0067] S103. Calculate the semantic similarity score between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set.

[0068] In this embodiment of the invention, a semantic similarity score is calculated between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set. A higher semantic similarity score indicates a higher semantic similarity between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set. For example, the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set are vectorized separately, and then their cosine similarity is calculated as the semantic similarity score.

[0069] S104. Calculate the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set.

[0070] In this embodiment of the invention, a consistency score is calculated between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set. The consistency score represents the degree of agreement between the field observation parameters and the working condition represented by the constraint set of each parameterized hyperedge in the hyperedge candidate set; a higher consistency score indicates a higher degree of agreement between the field observation parameters and the working condition represented by the constraint set.

[0071] S105. By combining semantic similarity score and consistency score, a comprehensive score is obtained for each parameterized hyperedge in the candidate hyperedge set, and the top K target parameterized hyperedges with the highest comprehensive scores are determined.

[0072] In this embodiment of the invention, a comprehensive score is obtained for each parameterized hyperedge in the candidate hyperedge set by combining semantic similarity score and consistency score, and the top K target parameterized hyperedges with the highest comprehensive scores are determined. For example, the semantic similarity score and consistency score can be multiplied to highlight the weakest link effect, that is, if either score is low, the comprehensive score will inevitably be lowered, which can effectively avoid the situation of "semantic similarity but inconsistent working conditions, or consistent working conditions but dissimilar semantics".

[0073] In this embodiment of the invention, the semantic similarity score between the elevator maintenance consultation text and the text entries of each parameterized hyperedge in the hyperedge candidate set, and the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set are calculated respectively. The comprehensive score of each parameterized hyperedge in the hyperedge candidate set is obtained by combining the semantic similarity score and the consistency score, and the top K target parameterized hyperedges with the highest comprehensive scores are determined. This ensures that the elevator maintenance consultation text and the text entries of the target parameterized hyperedges are semantically similar, and that the working conditions represented by the field observation parameters and the constraint set of the target parameterized hyperedges are consistent, thereby improving the accuracy of the generated elevator maintenance work orders.

[0074] S106. For each target parameterized hyperedge, map the parameters of the constraint set in the target parameterized hyperedge to the elevator maintenance work order template to obtain the elevator maintenance work order.

[0075] In this embodiment of the invention, for each target parameterized hyperedge, the parameters of the constraint set in the target parameterized hyperedge are mapped to the elevator maintenance work order template to obtain an elevator maintenance work order. For example, a field mapping method can be used to map the parameters of the corresponding fields in the constraint set of the target parameterized hyperedge to the corresponding fields in the elevator maintenance work order template to obtain an elevator maintenance work order. In this way, K elevator maintenance work orders can be generated. For example, the K elevator maintenance work orders can be sorted from high to low according to their comprehensive scores, and the elevator maintenance work orders with higher rankings can be pushed to the user first.

[0076] The elevator maintenance retrieval enhancement question-answering method provided in this embodiment of the invention, after obtaining the elevator maintenance consultation text input by the user, extracts on-site observation parameters from the elevator maintenance consultation text, and searches for parameterized hyperedges matching the elevator maintenance consultation text in the elevator maintenance knowledge hypergraph as a hyperedge candidate set. The elevator maintenance knowledge hypergraph is constructed based on the multi-dimensional relationships between entities in elevator maintenance-related documents. The elevator maintenance knowledge hypergraph includes multiple parameterized hyperedges describing text entries in elevator maintenance-related documents. Each parameterized hyperedge includes a first entity set consisting of multiple associated entities and a constraint set representing the relationships between entities. The method calculates the text of the elevator maintenance consultation text and the text of each parameterized hyperedge in the hyperedge candidate set. The semantic similarity score of the entries is calculated, and the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set is calculated. The semantic similarity score and the consistency score are combined to obtain the comprehensive score of each parameterized hyperedge in the hyperedge candidate set, and the top K target parameterized hyperedges with the highest comprehensive scores are determined. For each target parameterized hyperedge, the parameters of the constraint set in the target parameterized hyperedge are mapped to the elevator maintenance work order template to obtain the elevator maintenance work order. This invention, by introducing parameterized hyperedges, solves the limitations of the binary relations of ordinary knowledge graphs, and can adapt to the elevator operation and maintenance retrieval enhanced question answering scenario with multiple entities and multiple constraints, thereby improving the accuracy of elevator operation and maintenance retrieval enhanced question answering and elevator maintenance work orders. Furthermore, in this embodiment of the invention, the semantic similarity score between the elevator maintenance consultation text and the text entries of each parameterized hyperedge in the hyperedge candidate set, and the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set are calculated respectively. The comprehensive score of each parameterized hyperedge in the hyperedge candidate set is obtained by combining the semantic similarity score and the consistency score, and the top K target parameterized hyperedges with the highest comprehensive scores are determined. This ensures that the elevator maintenance consultation text and the text entries of the target parameterized hyperedges are semantically similar, and that the working conditions represented by the field observation parameters and the constraint set of the target parameterized hyperedges are consistent, thereby improving the accuracy of the generated elevator maintenance work orders.

[0077] Figure 2 A flowchart of another elevator operation and maintenance retrieval enhanced question-and-answer method provided by the present invention is shown below. Figure 2 As shown, the elevator operation and maintenance retrieval enhanced question-and-answer method includes the following steps:

[0078] S201. Obtain the elevator maintenance consultation text input by the user, and extract the on-site observation parameters and a second entity set consisting of multiple entities from the elevator maintenance consultation text.

[0079] S202. Find parameterized hyperedges that match the elevator maintenance consultation text from the elevator operation and maintenance knowledge hypergraph as the hyperedge candidate set.

[0080] In this embodiment of the invention, an elevator operation and maintenance knowledge hypergraph can be pre-constructed based on the multivariate relationships between entities in elevator operation and maintenance related documents. The elevator operation and maintenance knowledge hypergraph includes multiple parametric hyperedges (PHEs) describing text entries in the elevator operation and maintenance related documents. Each parametric hyperedge includes a first entity set composed of multiple associated entities and a constraint set representing the relationships between entities. In a typical graph structure, an edge can only connect two points, and an edge is actually a symbol used to represent an adjacency relationship between two points. However, an edge in a hypergraph can associate (or contain) two or more points; that is, all points on the same edge are adjacent, i.e., a hyperedge. In this embodiment of the invention, the hyperedges in the elevator operation and maintenance knowledge hypergraph are parameterized to form parametric hyperedges. Each parametric hyperedge includes a first entity set composed of multiple associated entities (or nodes) and a constraint set representing the relationships between entities. The parametric hyperedge packages multiple related entities and computable constraints in the same record to answer the question, "In these objects / scenarios, when these constraints are satisfied, how should the operation be performed and how should acceptance be carried out?" Entities can include the following types:

[0081] Equipment or components, such as "door operator", "light curtain", "controller", etc.;

[0082] Model / Series: such as "DO-X v2";

[0083] Scenario: such as "residential building ≤ 20 floors", "winter", "strong winds pressuring the entrance floor";

[0084] Qualifications / Resources: such as "Level II Maintenance Qualification" and "Torque Tester";

[0085] Compliance clauses: such as "GB×××× / §5.2.1 (applicable to residential scenarios)".

[0086] The constraints in the constraint set may include measurable parameters (range / threshold / OR logic, such as CloseTime, DoorCurrent, DutyCycle, etc.), time windows, power outages / elevator shutdowns, qualifications / key tools, applicable terms, process preferences, resource optimization, etc., which are not limited in this invention.

[0087] For example, the constraint set can be divided into hard constraints and soft constraints. Hard constraints may include legal provisions, personnel qualifications, whether power needs to be cut off and elevators need to be stopped, prohibition of live work, construction time windows, etc., while soft constraints may include measurable parameters, optimized processes, resource availability, etc.

[0088] In some embodiments of the present invention, searching for parameterized hyperedges that match elevator maintenance consultation text from the elevator operation and maintenance knowledge hypergraph as a candidate set of hyperedges includes the following sub-steps:

[0089] S2021. For each field observation parameter, determine whether the field observation parameter satisfies the hard constraint of the parameterized hyperedge.

[0090] In this embodiment of the invention, for each field observation parameter, it is determined whether the field observation parameter satisfies the hard constraint of the parameterized hyperedge.

[0091] S2022. If not, then parameterized hyperedges are excluded from the hyperedge candidate set.

[0092] In this embodiment of the invention, if any field observation parameter does not meet the corresponding hard constraints, such as the time window not being covered, the elevator having to be stopped but not being able to be stopped on site, or the lack of key qualifications / tools, the parameterized hyperedge is excluded from the hyperedge candidate set, thereby reducing the computational workload of subsequent processes.

[0093] S2023. Use keyword matching to find the parameterized hyperedges corresponding to the first entity set that match the second entity set from the elevator operation and maintenance knowledge hypergraph as the first hyperedge candidate set.

[0094] In this embodiment of the invention, a keyword matching method is used to find the parameterized hyperedges corresponding to the first entity set that match the second entity set from the elevator operation and maintenance knowledge hypergraph as the first hyperedge candidate set.

[0095] S2024. Calculate the semantic similarity between the elevator maintenance consultation text and the text entries of each parameterized hyperedge in the elevator operation and maintenance knowledge hypergraph, and take the parameterized hyperedges with the highest semantic similarity as the second hyperedge candidate set.

[0096] In this embodiment of the invention, the semantic similarity between the elevator maintenance consultation text and the text entries of each parameterized hyperedge in the elevator operation and maintenance knowledge hypergraph is calculated, and the parameterized hyperedges with the highest semantic similarity are selected as the second set of candidate hyperedges. For example, the elevator maintenance consultation text and the text entries of the parameterized hyperedges in the elevator operation and maintenance knowledge hypergraph can be vectorized separately, and then the cosine similarity between the two can be calculated as the semantic similarity.

[0097] S2025. Merge the first and second superedge candidate sets to obtain the superedge candidate set.

[0098] In this embodiment of the invention, the first and second hyperedge candidate sets are merged to obtain a new hyperedge candidate set. By merging the first and second hyperedge candidate sets, the scope of the hyperedge candidate set is expanded, avoiding omissions.

[0099] S203. Calculate the semantic similarity score between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set.

[0100] In this embodiment of the invention, a semantic similarity score is calculated between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set. A higher semantic similarity score indicates a higher semantic similarity between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set. For example, the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set are vectorized separately, and then their cosine similarity is calculated as the semantic similarity score.

[0101] S204. Calculate the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set.

[0102] In this embodiment of the invention, a consistency score is calculated between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set. The consistency score represents the degree of agreement between the field observation parameters and the working condition represented by the constraint set of each parameterized hyperedge in the hyperedge candidate set; a higher consistency score indicates a higher degree of agreement between the field observation parameters and the working condition represented by the constraint set.

[0103] In some embodiments of the present invention, calculating the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set includes the following sub-steps:

[0104] S2041. For each parameterized hyperedge in the candidate hyperedge set, calculate the absolute value of the difference between each field observation parameter and the corresponding measurable parameter.

[0105] In this embodiment of the invention, for each parameterized hyperedge in the candidate hyperedge set, the absolute value of the difference between each field observation parameter and the corresponding measurable parameter is calculated.

[0106] S2042. Calculate the quotient of the absolute value and the measurement scale of the field observation parameter to obtain the degree of deviation between the field observation parameter and the measurable parameter.

[0107] In this embodiment of the invention, the quotient of the absolute value and the measurement scale of the field observation parameter is calculated to obtain the degree of deviation between the field observation parameter and the measurable parameter. The measurement scale of the field observation parameter can be preset.

[0108] S2043. Calculate the product of the violation degree and the corresponding weight to obtain the weighted violation degree.

[0109] In this embodiment of the invention, a corresponding weight is set for each field observation parameter, and the product of the violation degree and the corresponding weight is calculated to obtain the weighted violation degree.

[0110] S2044. Calculate the sum of the weighted violations corresponding to all field observation parameters to obtain the total violation of the constraint set of the field observation parameters and parameterized hyperedges.

[0111] In this embodiment of the invention, the sum of the weighted violation degrees corresponding to all field observation parameters is calculated to obtain the total violation degree of the constraint set of the field observation parameters and parameterized hyperedges.

[0112] S2045. Calculate the value of the natural exponential function with the negative of the sum of the degrees of violation as the index, and use it as the consistency score.

[0113] In this embodiment of the invention, the function value of a natural exponential function with the negative of the sum of the degrees of violation as the exponent is calculated as the consistency score.

[0114] For example, the formula for calculating the consistency score is as follows:

[0115]

[0116] in, A consistency score is given between the field observation parameters and the constraint set of parameterized hyperedges in the candidate hyperedge set. The weight corresponding to the i-th field observation parameter, For the i-th field observation parameter With the corresponding i-th measurable parameter The degree of violation.

[0117] S205. Calculate the source confidence score of each parameterized hyperedge in the hyperedge candidate set based on the source and update time of the text entry.

[0118] For example, the parameterized hyperedge also includes metadata describing the source and update time of the text entry. The source of the text entry can be regulations, type tests, manufacturer manuals, maintenance summaries, personal experience, etc. Different sources correspond to different source confidence scores, with regulations having the highest source confidence score, and the others decreasing in descending order. The update time represents the time elapsed since the text entry's most recent update; the longer the time elapsed, the lower the corresponding source confidence score. In this embodiment of the invention, the source confidence score of the parameterized hyperedge can be obtained by multiplying the score corresponding to the text entry's source by the score corresponding to the update time.

[0119] S206. Calculate the product of semantic similarity score, consistency score and source confidence score to obtain the comprehensive score of each parameterized hyperedge in the hyperedge candidate set, and determine the top K target parameterized hyperedges with the highest comprehensive scores.

[0120] In this embodiment of the invention, the product of semantic similarity score, consistency score, and source confidence score is calculated to obtain the comprehensive score of each parameterized hyperedge in the candidate hyperedge set. This highlights the weakest link effect, meaning that a low score in any one of these categories will inevitably lower the comprehensive score, effectively avoiding situations where "semantics are similar but the working conditions do not match, or the working conditions match but the semantics are not similar." Then, the top K target parameterized hyperedges with the highest comprehensive scores are determined.

[0121] This invention calculates the product of semantic similarity score, consistency score, and source confidence score to obtain a comprehensive score for each parameterized hyperedge in the candidate hyperedge set. It then determines the top K target parameterized hyperedges based on their comprehensive scores, ensuring semantic similarity between the elevator maintenance consultation text and the text entries of the target parameterized hyperedges, and that the on-site observation parameters match the working conditions represented by the constraint set of the target parameterized hyperedges. Simultaneously, it ensures the reliability of the source of the text entries of the target parameterized hyperedges, thereby improving the accuracy of the generated elevator maintenance work orders.

[0122] S207. For each target parameterized hyperedge, map the parameters of the constraint set in the target parameterized hyperedge to the elevator maintenance work order template to obtain the elevator maintenance work order.

[0123] In this embodiment of the invention, for each target parameterized hyperedge, the parameters of the constraint set in the target parameterized hyperedge are mapped to the elevator maintenance work order template to obtain the elevator maintenance work order. For example, a field mapping method can be used to map the parameters of the corresponding fields in the constraint set of the target parameterized hyperedge to the corresponding fields in the elevator maintenance work order template to obtain the elevator maintenance work order.

[0124] In some embodiments of the present invention, an elevator maintenance work order template containing the skeleton of the standard operating procedure for elevator maintenance is found using the first entity set in the target parameterized hyperedge. A field mapping method is used to map the parameters of the constraint set in the target parameterized hyperedge to the corresponding fields in the elevator maintenance work order template, resulting in an elevator maintenance work order containing maintenance steps and quantitative acceptance standards. In this way, K elevator maintenance work orders can be generated. For example, the K elevator maintenance work orders can be sorted from high to low according to their comprehensive scores, and the elevator maintenance work orders with the highest scores are prioritized and pushed to the user.

[0125] For example, in some embodiments of the present invention, a list of hit / conflict constraints, source categories, and update times are listed along with the results, and these are attached as evidence to the elevator maintenance work order to meet audit and review requirements.

[0126] Figure 3 This is a schematic diagram of the structure of an elevator operation and maintenance retrieval enhancement question-and-answer device provided by the present invention, as shown below. Figure 3 As shown, the elevator operation and maintenance retrieval enhancement question and answer device includes:

[0127] The observation parameter extraction module 101 is used to obtain the elevator maintenance consultation text input by the user and extract the on-site observation parameters from the elevator maintenance consultation text.

[0128] The hyperedge candidate set determination module 102 is used to search for the parameterized hyperedges that match the elevator maintenance consultation text from the elevator operation and maintenance knowledge hypergraph as the hyperedge candidate set. The elevator operation and maintenance knowledge hypergraph is constructed based on the multi-dimensional relationships between entities in elevator operation and maintenance related documents. The elevator operation and maintenance knowledge hypergraph includes multiple parameterized hyperedges used to describe the text entries in the elevator operation and maintenance related documents. The parameterized hyperedges include a first entity set composed of multiple associated entities and a constraint set used to represent the relationships between the entities.

[0129] The similarity score calculation module 103 is used to calculate the semantic similarity score between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set;

[0130] The consistency score calculation module 104 is used to calculate the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set;

[0131] The scoring module 105 is used to integrate the semantic similarity score and the consistency score to obtain the comprehensive score of each parameterized hyperedge in the hyperedge candidate set, and to determine the top K target parameterized hyperedges with the highest comprehensive scores.

[0132] The maintenance work order generation module 106 is used to map the parameters of the constraint set in each of the target parameterized hyperedges to the elevator maintenance work order template to obtain an elevator maintenance work order.

[0133] In some embodiments of the present invention, the hyperedge candidate set determination module 102 includes:

[0134] The entity extraction submodule is used to extract a second entity set consisting of multiple entities from the elevator maintenance consultation text.

[0135] The first candidate set determination submodule is used to search for the parameterized hyperedges corresponding to the first entity set that matches the second entity set from the elevator operation and maintenance knowledge hypergraph using keyword matching as the first hyperedge candidate set.

[0136] The second candidate set determination submodule is used to calculate the semantic similarity between the elevator maintenance consultation text and the text entries of each parameterized hyperedge in the elevator operation and maintenance knowledge hypergraph, and to select the parameterized hyperedges with the highest semantic similarity as the second hyperedge candidate set.

[0137] The candidate set merging submodule is used to merge the first hyperedge candidate set and the second hyperedge candidate set to obtain the hyperedge candidate set.

[0138] In some embodiments of the present invention, the constraint set includes soft constraints, which are measurable parameters, and the consistency score calculation module 104 includes:

[0139] The absolute value calculation submodule is used to calculate the absolute value of the difference between each of the field observation parameters and the corresponding measurable parameters for each of the parameterized hyperedges in the hyperedge candidate set.

[0140] The violation calculation submodule is used to calculate the quotient of the absolute value and the measurement scale of the field observation parameter to obtain the violation degree between the field observation parameter and the measurable parameter;

[0141] The violation weighting submodule is used to calculate the product of the violation degree and the corresponding weight to obtain the weighted violation degree;

[0142] The violation sum calculation submodule is used to calculate the sum of the weighted violation degrees corresponding to all the field observation parameters, so as to obtain the sum of the violation degrees of the constraint sets of the field observation parameters and the parameterized hyperedges;

[0143] The consistency score calculation submodule is used to calculate the function value of a natural exponential function with the negative of the sum of the violations as the index, as the consistency score.

[0144] In some embodiments of the present invention, the constraint set includes hard constraints, and the elevator operation and maintenance retrieval enhanced question-and-answer device further includes:

[0145] The judgment module is used to, after obtaining the elevator maintenance consultation text input by the user and extracting the field observation parameters from the elevator maintenance consultation text, determine whether the field observation parameter satisfies the hard constraint of the parameterized hyperedge for each field observation parameter.

[0146] The exclusion module is used to exclude the parameterized hyperedge from the hyperedge candidate set when the field observation parameters do not meet the hard constraints of the parameterized hyperedge.

[0147] In some embodiments of the present invention, the parameterized hyperedge further includes metadata describing the source and update time of the text entry, and the elevator operation and maintenance retrieval enhanced question-and-answer device further includes:

[0148] The source confidence score calculation module is used to calculate the source confidence score of each parameterized hyperedge in the candidate hyperedge set based on the source and update time of the text entry after finding the parameterized hyperedge that matches the elevator maintenance consultation text from the elevator operation and maintenance knowledge hypergraph.

[0149] In some embodiments of the present invention, the scoring and integration module 105 includes:

[0150] The comprehensive scoring submodule is used to calculate the product of the semantic similarity score, the consistency score, and the source confidence score to obtain the comprehensive score of each parameterized hyperedge in the hyperedge candidate set.

[0151] In some embodiments of the present invention, the maintenance work order generation module 106 includes:

[0152] The template determination submodule is used to search for an elevator maintenance work order template that includes the skeleton of the standard operating procedure for elevator maintenance using the first entity set in the target parameterized hyperedge.

[0153] The field mapping submodule is used to map the parameters of the constraint set in the target parameterized hyperedge to the corresponding fields in the elevator maintenance work order template, thereby obtaining the elevator maintenance work order.

[0154] The elevator operation and maintenance retrieval enhancement question and answer device described above can execute the elevator operation and maintenance retrieval enhancement question and answer method provided in the foregoing embodiments of the present invention, and has the corresponding functional modules and beneficial effects of executing the elevator operation and maintenance retrieval enhancement question and answer method.

[0155] Figure 4 This is a schematic diagram of an electronic device provided by the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0156] like Figure 4As shown, the electronic device includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer programs stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0157] Multiple components in the electronic device are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, optical disk, etc.; and a communication unit 19, such as a network card, modem, wireless transceiver, etc. The communication unit 19 allows the electronic device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0158] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, central processing unit (CPU), graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the elevator operation and maintenance retrieval enhanced question answering method.

[0159] In some embodiments, the elevator maintenance retrieval enhanced question-and-answer method can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on an electronic device via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the elevator maintenance retrieval enhanced question-and-answer method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to execute the elevator maintenance retrieval enhanced question-and-answer method by any other suitable means (e.g., by means of firmware).

[0160] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0161] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0162] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0163] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0164] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0165] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0166] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the elevator operation and maintenance retrieval enhanced question-and-answer method as provided in any embodiment of this application.

[0167] In implementing the computer program product, computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0168] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0169] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. An elevator maintenance search enhanced question and answer method, characterized by, include: Obtain elevator maintenance consultation text input by the user, and extract on-site observation parameters from the elevator maintenance consultation text; The elevator operation and maintenance knowledge hypergraph searches for parameterized hyperedges that match the elevator maintenance consultation text as a candidate set of hyperedges. The elevator operation and maintenance knowledge hypergraph is constructed based on the multi-dimensional relationships between entities in elevator operation and maintenance related documents. The elevator operation and maintenance knowledge hypergraph includes multiple parameterized hyperedges used to describe text entries in the elevator operation and maintenance related documents. Each parameterized hyperedge includes a first entity set composed of multiple associated entities and a constraint set used to represent the relationships between the entities. Calculate the semantic similarity score between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set; Calculate the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the candidate hyperedge set; By combining the semantic similarity score and the consistency score, a comprehensive score is obtained for each parameterized hyperedge in the hyperedge candidate set, and the top K target parameterized hyperedges with the highest comprehensive scores are determined. For each of the target parameterized hyperedges, the parameters of the constraint set in the target parameterized hyperedge are mapped to the elevator maintenance work order template to obtain the elevator maintenance work order.

2. The elevator operation and maintenance retrieval enhanced question-and-answer method according to claim 1, characterized in that, The parameterized hyperedges that match the elevator maintenance consultation text are selected from the elevator operation and maintenance knowledge hypergraph as a candidate set of hyperedges, including: Extract a second entity set consisting of multiple entities from the elevator maintenance consultation text; The parameterized hyperedges corresponding to the first entity set that matches the second entity set are used as the first hyperedge candidate set from the elevator operation and maintenance knowledge hypergraph by keyword matching. Calculate the semantic similarity between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the elevator operation and maintenance knowledge hypergraph, and take the parameterized hyperedges with the highest semantic similarity as the second hyperedge candidate set; The first and second hyperedge candidate sets are merged to obtain the hyperedge candidate set.

3. The elevator operation and maintenance retrieval enhanced question-and-answer method according to claim 1 or 2, characterized in that, The constraint set includes soft constraints, which are measurable parameters. The consistency score between the observed parameters and the constraint set of each parameterized hyperedge in the candidate hyperedge set is calculated, including: For each parameterized hyperedge in the candidate hyperedge set, calculate the absolute value of the difference between each field observation parameter and the corresponding measurable parameter; The degree of deviation between the absolute value and the measurement scale of the field observation parameter is obtained by calculating the ratio between the field observation parameter and the measurable parameter. The weighted violation degree is obtained by multiplying the violation degree by the corresponding weight. Calculate the sum of the weighted violation degrees corresponding to all the field observation parameters to obtain the total violation degree of the constraint set of the field observation parameters and the parameterized hyperedge; The consistency score is calculated by taking the value of a natural exponential function with the negative of the sum of the aforementioned violations as the exponent.

4. The elevator operation and maintenance retrieval enhanced question-and-answer method according to claim 1 or 2, characterized in that, The constraint set includes hard constraints, and after obtaining the elevator maintenance consultation text input by the user and extracting the field observation parameters from the elevator maintenance consultation text, it also includes: For each of the field observation parameters, determine whether the field observation parameter satisfies the hard constraint of the parameterized hyperedge; If not, the parameterized hyperedge is excluded from the hyperedge candidate set.

5. The elevator operation and maintenance retrieval enhanced question-and-answer method according to claim 1 or 2, characterized in that, The parameterized hyperedge also includes metadata describing the source and update time of the text entry. After searching the elevator operation and maintenance knowledge hypergraph for the parameterized hyperedge that matches the elevator maintenance consultation text as a candidate set of hyperedges, the method further includes: The source confidence score of each parameterized hyperedge in the candidate hyperedge set is calculated based on the source and update time of the text entry.

6. The elevator operation and maintenance retrieval enhanced question-and-answer method according to claim 5, characterized in that, By combining the semantic similarity score and the consistency score, a comprehensive score is obtained for each parameterized hyperedge in the hyperedge candidate set, including: The comprehensive score of each parameterized hyperedge in the candidate hyperedge set is obtained by multiplying the semantic similarity score, the consistency score, and the source confidence score.

7. The elevator operation and maintenance retrieval enhanced question-and-answer method according to claim 1 or 2, characterized in that, For each of the aforementioned target parameterized hyperedges, the parameters of the constraint set in the target parameterized hyperedge are mapped to the elevator maintenance work order template to obtain the elevator maintenance work order, including: The elevator maintenance work order template containing the skeleton of the standard operating procedure for elevator maintenance is found using the first entity set in the target parameterized hyperedge. By using field mapping, the parameters of the constraint set in the target parameterized hyperedge are mapped to the corresponding fields in the elevator maintenance work order template to obtain the elevator maintenance work order.

8. An elevator operation and maintenance retrieval enhanced question-and-answer device, characterized in that, include: The observation parameter extraction module is used to obtain the elevator maintenance consultation text input by the user and extract the on-site observation parameters from the elevator maintenance consultation text; The hyperedge candidate set determination module is used to find parameterized hyperedges that match the elevator maintenance consultation text from the elevator operation and maintenance knowledge hypergraph as hyperedge candidate sets. The elevator operation and maintenance knowledge hypergraph is constructed based on the multi-dimensional relationships between entities in elevator operation and maintenance related documents. The elevator operation and maintenance knowledge hypergraph includes multiple parameterized hyperedges used to describe text entries in the elevator operation and maintenance related documents. Each parameterized hyperedge includes a first entity set composed of multiple associated entities and a constraint set used to represent the relationships between the entities. The similarity score calculation module is used to calculate the semantic similarity score between the elevator maintenance consultation text and the text entry of each parameterized hyperedge in the hyperedge candidate set; The consistency score calculation module is used to calculate the consistency score between the field observation parameters and the constraint set of each parameterized hyperedge in the hyperedge candidate set; The scoring and integration module is used to integrate the semantic similarity score and the consistency score to obtain the comprehensive score of each parameterized hyperedge in the hyperedge candidate set, and to determine the top K target parameterized hyperedges with the highest comprehensive scores. The maintenance work order generation module is used to map the parameters of the constraint set in each of the target parameterized hyperedges to the elevator maintenance work order template to obtain the elevator maintenance work order.

9. An electronic device, characterized in that, include: One or more processors; Storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the elevator operation and maintenance retrieval enhanced question answering method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the elevator operation and maintenance retrieval enhanced question-and-answer method as described in any one of claims 1-7.

Images