Airline order analysis method and device, electronic equipment and storage medium

Abstract

The invention discloses an aviation order analysis method and device, electronic equipment and a storage medium, and relates to the technical field of aviation information or other related fields, and the method comprises the steps: receiving to-be-analyzed target order data, and verifying the target order data based on a preset verification strategy; under the condition that verification is passed, service entity information is extracted from the target order data based on a preset extraction strategy, and the preset extraction strategy is used for specifying an extraction field and a field association relationship; a time sequence correlation model of the target order is constructed based on the business entity information by adopting a time sequence analysis algorithm, and the time sequence correlation model is used for recording all historical business operations of the target order according to a time sequence; and converting the time sequence correlation model into a business flow chart, and rendering the business flow chart to a display area. According to the invention, the technical problem of low data analysis precision caused by poor process data integration capability of an aviation system for complex orders in the prior art is solved.

Description

Technical Field

[0001] This invention relates to the field of aviation information technology, and more specifically, to a method and apparatus for parsing airline orders, an electronic device, and a storage medium. Background Technology

[0002] Against the backdrop of digital transformation in the aviation industry, processing complex order data for airlines' B2C and B2B sales operations has become a major challenge. Existing order data structures, especially models based on SPNR (Standard Product Notification Resource), while designed to integrate product definitions and service change records, have seen a sharp increase in the correlation and complexity of order data due to the diversification of business scenarios, leading to the following main problems:

[0003] 1. Difficulty in data traceability: The order lifecycle involves various operations such as flight adjustments, service additions, and changes in payment status. Traditional data storage methods are difficult to clearly record and display these dynamic change paths. Especially after an order has undergone multiple modifications or status changes, it becomes extremely difficult to find the cause and impact of a specific event.

[0004] 2. Low analysis efficiency: Faced with massive amounts of information for a single order, it is necessary to rely on manual analysis of the order text and judge logical nodes based on personal experience and intuition. Especially when dealing with complex situations such as online and offline operations being synchronized, product item splitting, and fund flow, it is not only time-consuming and laborious, but also prone to differences and omissions in interpretation, affecting accuracy.

[0005] 3. Lack of visualization capabilities: Current analysis tools are limited to text-based displays, making it difficult to convert the complex change logic in order information into intuitive graphical representations, which significantly increases communication costs in cross-departmental collaboration;

[0006] 4. Challenges in handling multi-version data: Frequent updates to the order data model generate multiple versions of the SPNR data structure, requiring the system to flexibly adapt to different versions of the schema when processing orders. However, existing technologies often struggle to achieve real-time version adaptation and data parsing, thus limiting the system's response speed and accuracy.

[0007] 5. Lagging Business Rules and Knowledge Updates: As business rules continue to evolve, the existing system's knowledge base update mechanism lags behind and cannot reflect the latest business scenarios and rules in a timely manner, which may lead to business gaps and misjudgments in the processing of complex orders.

[0008] While some technologies have attempted to address individual aspects of the above issues, there is still no comprehensive solution that can fully cover intelligent analysis of airline order data, business semantic visualization mapping, and knowledge-driven anomaly detection closed loop. Summary of the Invention

[0009] This invention provides a method, apparatus, electronic device, and storage medium for parsing airline orders, to at least solve the technical problem in the related art of poor process data integration capability of aviation systems for complex orders, resulting in low data parsing accuracy.

[0010] According to one aspect of the present invention, a method for parsing airline orders is provided, comprising: receiving target order data to be parsed, and verifying the target order data based on a preset verification strategy, wherein the preset verification strategy includes at least: structure verification and format verification; if the verification passes, extracting business entity information from the target order data based on a preset extraction strategy, wherein the preset extraction strategy is used to specify extraction fields and field relationships; constructing a time-series association model of the target order based on the business entity information using a time-series analysis algorithm, wherein the time-series association model is used to record all historical business operations of the target order in time sequence; converting the time-series association model into a business process diagram, and rendering the business process diagram to a display area.

[0011] Furthermore, the step of verifying the target order data based on a preset verification strategy includes: obtaining the latest version of the data specification from a metadata repository, wherein the metadata repository is a set of data specifications that are updated periodically, and the data specifications are used to define structural standards, semantic constraints, and business rules; and performing layered verification on the target order data based on the structural standards, the semantic constraints, and the business rules to obtain the verification results.

[0012] Further, the step of performing layered verification on the target order data based on the structural standard, the semantic constraints, and the business rules to obtain verification results includes: verifying all fields in the target order data based on the path expression defined in the structural standard to obtain a first verification sub-result, wherein the path expression at least standardizes the following: root node, namespace, and nested field elements; verifying all fields in the target order data based on the business thesaurus defined in the semantic constraints to obtain a second verification sub-result, wherein the semantic constraints at least standardize the following: field naming and field semantic consistency; verifying business fields in the target order data based on the business rules defined in the business rules to obtain a third verification sub-result, wherein the business rules at least standardize the following: the value range of numeric fields and the operational compliance of non-numeric fields; and integrating the first verification sub-result, the second verification sub-result, and the third verification sub-result to obtain the verification result.

[0013] Further, the step of extracting business entity information from the target order data based on a preset extraction strategy includes: obtaining an entity field mapping table, and locating the extraction field specified by the preset extraction strategy in the target order data based on the entity field mapping table to obtain field location information as the location result; traversing the target order data based on the field association relationship, and performing layered parsing of the target order data based on business dimension, transaction dimension, and operation record dimension during the traversal process to obtain parsing results; and performing feature extraction on the target order data based on the location result and parsing result to obtain the business entity information.

[0014] Furthermore, the step of performing layered parsing of the target order data based on business dimension, transaction dimension, and operation record dimension during the traversal process to obtain the parsing result includes: performing subject parsing on the target order data in the business dimension to obtain an order subject set; performing feature parsing on the target order data in the transaction dimension to obtain a transaction feature set; performing source tracing parsing on the target order data in the operation record dimension to obtain an operation source tracing set, wherein the operation source tracing set is used to record operation events and corresponding operation records; and generating the parsing result based on the order subject set, the transaction feature set, and the operation source tracing set.

[0015] Furthermore, the step of constructing a time-series association model of the target order based on the business entity information using a time-series analysis algorithm includes: performing time-series analysis on the business entity information to obtain time-series records of all historical business operations; generating an initial model framework based on the time-series records and the operation traceability set in the business entity information, wherein the initial model framework contains operation nodes corresponding to the historical business operations arranged in time sequence; generating subject nodes based on the order subject set in the business entity information and generating node labels based on the transaction feature set; supplementing the subject nodes and the node labels to the initial model framework and establishing association lines between the subject nodes and the operation nodes to obtain the time-series association model.

[0016] Further, the step of establishing the association lines between the main nodes and the operation nodes to obtain the temporal association model includes: for each operation event corresponding to the operation node, analyzing the impact of the operation event on the state of the target order, wherein the state impact includes: the sub-influence of the operation event on the state of each main node, and the changes in transaction characteristics caused by the operation event; generating a first type of association line between the operation nodes based on the sub-influence of the state; constructing a second type of association line between the node label and the model node based on the changes in transaction characteristics; and generating the operation causal chain of the target order by analyzing the first type of association line and the second type of association line.

[0017] Furthermore, after constructing a time-series correlation model of the target order based on the business entity information, the method further includes: integrating the business entity information, transaction features, and operation records extracted from the target order data to obtain business summary information, and generating a business report corresponding to the target order data based on the business summary information; comparing and analyzing the time-series correlation model and the business report to obtain the mapping relationship between model elements and the business summary information, wherein the model elements refer to model nodes and correlation lines; generating an order analysis chart based on the mapping relationship and the business report, and rendering the order analysis chart to the display area.

[0018] According to another aspect of the present invention, an apparatus for parsing airline orders is also provided, comprising: a verification unit, configured to receive target order data to be parsed and verify the target order data based on a preset verification strategy, wherein the preset verification strategy includes at least: structure verification and format verification; an extraction unit, configured to extract business entity information from the target order data based on a preset extraction strategy if the verification passes, wherein the preset extraction strategy is used to specify extraction fields and field relationships; a construction unit, configured to construct a time-series association model of the target order based on the business entity information using a time-series analysis algorithm, wherein the time-series association model is used to record all historical business operations of the target order in time sequence; and a display unit, configured to convert the time-series association model into a business process diagram and render the business process diagram to a display area.

[0019] Furthermore, the verification unit includes: an acquisition module, used to acquire the latest version of the data specification from a metadata repository, wherein the metadata repository is a set of data specifications that are updated periodically, and the data specifications are used to define structural standards, semantic constraints, and business rules; and a hierarchical verification module, used to perform hierarchical verification on the target order data based on the structural standards, the semantic constraints, and the business rules, and obtain verification results.

[0020] Further, the hierarchical verification module includes: a first verification submodule, used to verify all fields in the target order data based on the path expression defined in the structural standard, to obtain a first verification sub-result, wherein the path expression at least specifies the following: root node, namespace, and nested field elements; a second verification submodule, used to verify all fields in the target order data based on the business thesaurus defined in the semantic constraints, to obtain a second verification sub-result, wherein the semantic constraints at least specify the following: field naming and field semantic consistency; a third verification submodule, used to verify the business fields in the target order data based on the business rules defined in the business rules, to obtain a third verification sub-result, wherein the business rules at least specify the following: the field value range of numeric fields, and the operational compliance of non-numeric fields; and an integration submodule, used to integrate the first verification sub-result, the second verification sub-result, and the third verification sub-result to obtain the verification result.

[0021] Further, the extraction unit includes: a positioning module, used to obtain an entity field mapping table, and locate the extraction field specified by the preset extraction strategy in the target order data based on the entity field mapping table, to obtain field location information as the positioning result; a hierarchical parsing module, used to traverse the target order data based on the field association relationship, and perform hierarchical parsing of the target order data based on business dimension, transaction dimension and operation record dimension during the traversal process, to obtain parsing results; and a feature extraction module, used to extract features from the target order data based on the positioning result and the parsing result, to obtain the business entity information.

[0022] Further, the hierarchical parsing module includes: a first parsing submodule, used to perform subject parsing on the target order data in the business dimension to obtain an order subject set; a second parsing submodule, used to perform feature parsing on the target order data in the transaction dimension to obtain a transaction feature set; a third parsing submodule, used to perform source tracing parsing on the target order data in the operation record dimension to obtain an operation source tracing set, wherein the operation source tracing set is used to record operation events and corresponding operation records; and a first generation submodule, used to generate the parsing result based on the order subject set, the transaction feature set, and the operation source tracing set.

[0023] Further, the construction unit includes: an analysis module, used to perform time-series analysis on the business entity information to obtain time-series records of all historical business operations; a first generation module, used to generate an initial model framework based on the time-series records and the operation traceability set in the business entity information, wherein the initial model framework contains operation nodes corresponding to the historical business operations arranged in time sequence; a second generation module, used to generate subject nodes based on the order subject set in the business entity information and generate node labels based on the transaction feature set; and an establishment module, used to supplement the subject nodes and the node labels to the initial model framework and establish the association lines between the subject nodes and the operation nodes to obtain the time-series association model.

[0024] Further, the establishment module includes: an analysis submodule, used to analyze the impact of the operation event corresponding to each operation node on the state of the target order, wherein the state impact includes: the sub-influence of the operation event on the state of each main node, and the changes in transaction characteristics caused by the operation event; a second generation submodule, used to generate a first type of correlation line between the operation nodes based on the sub-influence of the state; a construction submodule, used to construct a second type of correlation line between the node label and the model node based on the changes in transaction characteristics; and a third generation submodule, used to generate the operation causal chain of the target order by analyzing the first type of correlation line and the second type of correlation line.

[0025] Furthermore, the airline order parsing device further includes: an integration module, used to integrate the business entity information, transaction features, and operation records extracted from the target order data after constructing a time-series correlation model of the target order based on the business entity information, to obtain business summary information, and to generate a business report corresponding to the target order data based on the business summary information; a comparison and analysis module, used to compare and analyze the time-series correlation model and the business report to obtain the mapping relationship between model elements and the business summary information, wherein the model elements refer to model nodes and correlation lines; and a rendering module, used to generate an order analysis chart based on the mapping relationship and the business report, and to render the order analysis chart to the display area.

[0026] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute any of the above-described methods for parsing air order.

[0027] According to another aspect of the present invention, an electronic device is also provided, including one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement any of the above-described methods for parsing airfare orders.

[0028] This invention proposes a method for parsing airline orders. First, it receives the target order data to be parsed and verifies it based on a preset verification strategy. This preset verification strategy includes at least structural and format verification. Then, if the verification passes, it extracts business entity information from the target order data based on a preset extraction strategy, which specifies the extraction fields and their relationships. Next, a time-series analysis algorithm is used to construct a time-series association model of the target order based on the business entity information. This time-series association model records all historical business operations of the target order in chronological order. Finally, the time-series association model is converted into a business process diagram, which is then rendered to the display area.

[0029] This invention employs a multi-layered data verification approach, utilizing a dynamic rule engine and metadata repository to ensure the consistency of order data structure and semantics, thereby improving the accuracy and efficiency of data parsing. Specifically, this invention combines entity field mapping tables and hierarchical parsing algorithms to achieve in-depth mining and precise positioning of business entity information, effectively integrating complex data from three dimensions: order subject, transaction characteristics, and operation records, ensuring data comprehensiveness and accuracy. Subsequently, a time-series association model is constructed using a time-series analysis algorithm. This model accurately records the change trajectory of the entire order lifecycle according to the operation time sequence, aiding in in-depth analysis and anomaly localization in dynamic business scenarios. Finally, the time-series association model is converted into a business process diagram, using visualization technology to transform the abstract order data change process into an intuitive process display, enhancing the readability and understandability of the information. Overall, this invention, through a four-stage architecture of intelligent verification, precise extraction, time-series analysis, and interactive visualization, significantly improves the automation level of aviation order data parsing and the intelligence of problem diagnosis, providing airlines with an efficient and accurate order management and analysis solution. This solves the technical problem of poor data parsing accuracy caused by the poor integration capability of complex order process data in related technologies within aviation systems. Attached Figure Description

[0030] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0031] Figure 1This is a structural diagram of an optional full-process analysis system for airline orders according to an embodiment of the present invention;

[0032] Figure 2 This is a flowchart of an optional method for parsing airfare orders according to an embodiment of the present invention;

[0033] Figure 3 This is a flowchart of an optional data verification module according to an embodiment of the present invention;

[0034] Figure 4 This is a flowchart of an optional data extraction module according to an embodiment of the present invention;

[0035] Figure 5 This is a flowchart of an optional intelligent analysis module according to an embodiment of the present invention;

[0036] Figure 6 This is a flowchart of an optional visualization module according to an embodiment of the present invention;

[0037] Figure 7 This is a schematic diagram of an optional air ticket parsing device according to an embodiment of the present invention;

[0038] Figure 8 This is a structural block diagram of an electronic device for executing an airline order parsing method according to an embodiment of the present invention. Detailed Implementation

[0039] 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.

[0040] 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.

[0041] To facilitate understanding of the present invention by those skilled in the art, some terms or nouns involved in the various embodiments of the present invention are explained below:

[0042] SPNR, Structured Product Naming and Resolution, is an industry-standard data model based on online travel agencies, used for airline order data storage and exchange. Its characteristics include a highly flexible data structure, capable of carrying a large amount of complex additional service and product information as well as rich status change records.

[0043] The TRP platform, Travel Retail Platform, is an online travel sales platform designed specifically for airlines. It provides a range of tools and services to help airlines manage product definition, order processing, payment processes, and customer service, enabling data integration and analysis under the new retail strategy.

[0044] Schema validation is the process of validating XML files based on the structure and type rules defined in the XML Schema. The Schema describes the structure and element types of the data, ensuring the correctness of the data format and semantic consistency. In this invention, it is used to verify the integrity and compliance of SPNR order data.

[0045] A rules engine is a programming component that allows for the dynamic configuration and execution of runtime decision logic. In this invention, the rules engine is used for dynamic strategy configuration when parsing order data, enabling it to adapt to different business scenarios and improve the flexibility and accuracy of order parsing.

[0046] Domain ontology mapping maps concepts, entities, and relationships within a business domain to a form that computers can understand, and is typically used to construct knowledge graphs. In this invention, domain ontology mapping technology is used to extract specific business entity information from order data, such as product definitions, payment details, and operation records, enabling refined data processing and structured storage of high-dimensional information.

[0047] The time-series graph storage model is a special data structure used to store and represent relational data with time attributes. In this invention, the time-series graph storage model is used to store and maintain the time-series association model of airline orders, which includes the time-series records of all historical business operations of the orders, facilitating subsequent causal chain reconstruction and business process tracing.

[0048] The Graphviz engine, a graph layout tool, can automatically transform graph-structured data into visual charts. In this invention, Graphviz is used to transform a time-series correlation model into a business process diagram, intuitively displaying the order operation process and lowering the barrier to understanding the business.

[0049] LLM, or Large Language Model, possesses powerful natural language understanding and generation capabilities. In this invention, LLM assists in the implementation of context-aware reasoning in the intelligent analysis module, used to identify and predict potential anomalies and business logic breakpoints in the order process.

[0050] A knowledge graph, a structured database centered on entities, attributes, and relationships, is used to store and represent domain knowledge. The knowledge graph library in this invention stores compliance paths for typical business scenarios, serving as a crucial basis for decision support in the intelligent analysis module, ensuring the accuracy of anomaly detection and the rationality of business interpretation.

[0051] The following embodiments of the present invention can be applied to various systems / applications / devices that require refined analysis of airline orders and full-process visual traceability, enabling airline order data management functions based on multi-dimensional verification and intelligent analysis. The present invention uses a dynamic rule engine to verify the structural and semantic consistency of order data, and then employs domain ontology mapping and hierarchical parsing algorithms to extract business entity information, thereby better constructing a temporal correlation model of orders and achieving automatic generation and rendering of business process diagrams.

[0052] Through the above process, this invention greatly simplifies airlines' understanding and operation of complex order data, providing powerful data analysis and visualization tools for problem identification, decision support, and business optimization.

[0053] The present invention will now be described in detail with reference to various embodiments.

[0054] Example 1

[0055] According to an embodiment of the present invention, an embodiment of an air order parsing method is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0056] Examples of embodiments of the present invention Figure 2The method for parsing airline orders shown is implemented using an airline order data visualization and analysis system. Combining time-series graph construction and intelligent analysis technologies, it is applied to new retail business scenarios in the aviation industry. Specifically addressing the logical coupling of multi-version order data and the bottlenecks of manual parsing, it utilizes a dynamic rule engine, domain ontology mapping, time-series analysis algorithms, and the Graphviz visualization engine. The specific steps include receiving, verifying, extracting, analyzing, and visualizing the data to improve the accuracy of order data parsing and the transparency of business processes. This system can significantly improve the efficiency and accuracy of airlines in processing complex orders, providing passengers with a better service experience, while reducing operating costs and enhancing the intelligence of business decision-making.

[0057] Figure 1 This is a structural diagram of an optional end-to-end parsing system for airline orders according to an embodiment of the present invention, as shown below. Figure 1 As shown, the full-process analysis system includes: a data verification module, a data extraction module, an intelligent analysis module, and a visualization module.

[0058] The data verification module, based on a dynamic schema adaptation engine, utilizes a multi-version order XSD schema library to achieve adaptation and compliance verification of order data across multiple versions. It employs a three-layer funnel-style verification: the structural layer verifies root node consistency, tag closure, and redundant fields; the semantic layer enforces verification of the completeness of required fields; and the business layer dynamically monitors rules such as amount thresholds. A tiered fault-tolerance mechanism is implemented, triggering process interruption alarms for blocking errors (such as missing critical fields) and generating visual risk warnings for non-blocking anomalies (such as amounts approaching the upper limit), achieving a balance between verification efficiency and business flexibility.

[0059] Data extraction module: By adopting a rule engine and domain ontology mapping technology framework, the module realizes the general structured parsing and temporal knowledge graph construction of multi-version order data. The structured parsing stage is based on three mechanisms: (1) accurate semantic mapping between XPath expressions and business feature library; (2) dynamic parsing strategy configuration driven by declarative rules; (3) multi-level entity penetration algorithm to realize cross-layer traversal of core entities such as Product (product definition data), Payment (payment order entity data), Audit (operation record), Remark (refund order entity data), etc., to fully extract the three dimensions of data: order subject (product), transaction features (payment status / amount verification, etc.) and operation record (change trajectory / refund traceability, etc.), ensuring cross-entity semantic consistency. The temporal modeling stage implements dual-path construction: based on the event flow segmentation of operation record sequence, a state transition probability matrix is ​​constructed to quantify operation correlation; multiple types of core entity data master nodes (Product / Payment / Audit / Remark) are defined to establish a dual correlation dimension of operation traceability chain (entity change → operation record trigger) and causal correlation chain (product iteration → payment status migration). The final output is a knowledge graph storage model with time-series tags. Through a hierarchical entity mapping mechanism, it ensures the full retention of business elements and provides structural support for end-to-end operation tracing and causal reasoning.

[0060] Intelligent Analysis Module: This module is based on a three-layer architecture to achieve knowledge-driven decision-making and visualization collaboration: 1) The knowledge graph construction layer uses deep semantic feature extraction based on BERT, integrates edit distance disambiguation and constraint rule verification mechanisms from a predefined thesaurus, and accurately identifies structured relationships such as "function-entity" and "function-constraint" in requirements and business documents. 2) The RAG enhanced analysis layer uses a hybrid index to dynamically retrieve internal and external knowledge bases, integrates LLM through a bidirectional attention mechanism to generate multi-hop reasoning paths, and outputs structured reports with confidence labels. 3) The business decision layer deploys a multimodal classifier to distinguish between anomalies and business reports. Anomalies are blocked from output by a three-level logical verification chain, while business reports activate a parameterized template engine to generate an adaptive topology graph. The visualization mapping logic is optimized through semantic routing and reinforcement learning feedback to form a closed loop of knowledge-decision-expression.

[0061] Visualization Module: This module builds order operation flowcharts based on Graphviz, using a top-bottom split layout (static business report sub-chart at the top, dynamic main flowchart at the bottom). It achieves composite visualization of business reports and multi-source operation flowcharts through five parts: structured layout initialization, dynamic data node generation (23-field parsing + color encoding), three-level nested sub-chart architecture (operation record / product / remark layering), time-series connection optimization (ID order + operation action label), and DOT file merging (gvpack tool), generating a hierarchical operation view.

[0062] Figure 2 This is a flowchart of an optional method for parsing airfare orders according to an embodiment of the present invention, such as... Figure 2 As shown, the method includes the following steps:

[0063] Step S201: Receive the target order data to be parsed and verify the target order data based on a preset verification strategy, wherein the preset verification strategy includes at least: structure verification and format verification.

[0064] Data validation is the initial step in this embodiment. Its main goal is to ensure that the order data's structure and format conform to predefined specifications before further analysis, avoiding deviations or failures in subsequent analyses due to data quality issues. Specifically, structure validation checks the completeness and validity of the XML document's structure, including confirming whether the document contains the correct root node, whether all necessary elements exist and are closed, and whether the relationships between elements conform to the preset schema (data structure model). Format validation focuses on verifying the correctness of data element formats; for example, checking whether field types (such as date and amount) meet expectations, whether numerical ranges are reasonable, and whether the use of special characters is appropriate.

[0065] In this invention, the structural and format validation strategies are implemented through an intelligent metadata-driven engine. This engine can access a distributed versioned metadata repository, automatically obtain the latest data specifications, and ensure that the validation strategies can adapt to version updates of order data. Furthermore, the high-frequency schema hotspots established through an LRU-K caching strategy significantly improve the response speed of the validation process.

[0066] Optionally, the step of verifying the target order data based on a preset verification strategy includes: obtaining the latest version of the data specification from the metadata repository, wherein the metadata repository is a collection of data specifications that are updated periodically, and the data specifications are used to define structural standards, semantic constraints and business rules; and performing layered verification on the target order data based on the structural standards, semantic constraints and business rules to obtain the verification results.

[0067] The steps described above outline a dynamically layered validation process designed to ensure that the target order data conforms to the latest standards in terms of structure, semantics, and business rules. Specifically, this process involves retrieving the latest version of the data specifications from the metadata repository and performing multi-layered validation on the target order data based on these specifications.

[0068] First, data specifications are proactively retrieved from a metadata repository. This repository, a dynamically updated collection of specifications, includes structural standards, semantic constraints, and business rules. Structural standards define the root node, elements, and attributes of an XML document; semantic constraints focus on the value range, type, and logical relationships between fields; and business rules cover validity checks for specific business scenarios, such as monetary thresholds and operation order. This intelligent retrieval process is optimized through an LRU-K caching strategy, ensuring that frequently accessed specifications respond quickly while adapting to data changes, providing the latest rule basis for the validation process.

[0069] Next, based on the acquired specifications, the target order data undergoes layered validation. This process follows a funnel-like filtering principle. First, a structural validation layer checks the document's structural consistency, including the integrity and closure of root nodes and element nesting levels. Then, a semantic validation layer validates the attribute values ​​and types of elements to ensure the semantic integrity of the data. Finally, a business rule validation layer uses a dynamic threshold calculation engine to monitor sensitive fields such as amounts and exchange rates in real time, combining this with business rules for compliance verification. This layered validation mechanism effectively distinguishes and handles structural, semantic, and business errors, ensuring data quality while improving validation efficiency and accuracy.

[0070] Alternatively, a supplementary error classification and handling operation can be added after validation to differentiate and handle validation errors of different levels. Blocking errors (such as missing critical fields) will trigger process interruption and real-time alarms, while non-blocking errors (such as amounts approaching the limit) will generate risk warnings, be stored in the context environment, and displayed as additional information in subsequent visualization modules. This enables risk gradient management, preventing serious errors while retaining the ability to analyze and alarm for non-serious errors, ensuring system robustness and data integrity.

[0071] Optionally, the step of performing layered validation on the target order data based on structural standards, semantic constraints, and business rules to obtain validation results includes: validating all fields in the target order data based on the path expression defined in the structural standards to obtain a first validation sub-result, wherein the path expression at least specifies the following: root node, namespace, and nested field elements; validating all fields in the target order data based on the business thesaurus defined in the semantic constraints to obtain a second validation sub-result, wherein the semantic constraints at least specify the following: field naming and field semantic consistency; validating the business fields in the target order data based on the business rules defined in the business rules to obtain a third validation sub-result, wherein the business rules at least specify the following: the value range of numeric fields and the operational compliance of non-numeric fields; and integrating the first, second, and third validation sub-results to obtain the validation result.

[0072] The above steps describe the specific hierarchical verification of target order data based on structural standards, semantic constraints, and business rules. These are key steps to ensure high-quality and high-precision processing of order data.

[0073] Structural standard validation involves a comprehensive scan of all fields in the target order data using path expressions (XPath or similar techniques) to ensure the integrity and correctness of the data structure. The checks primarily include: 1. Root Node and Namespace: Ensuring the root element and namespace of the XML document conform to the expected structural standards to avoid data parsing failures due to structural mismatches; 2. Field Element Nesting: Locating the nesting level of elements using XPath expressions to confirm that fields are correctly positioned in the document, the structure is closed, and there are no redundant or missing fields.

[0074] During the semantic constraint verification phase, the field names and semantic consistency of the target order data are verified. Based on the business thesaurus (such as a predefined JSON thesaurus), the matching degree between the field names and the field semantics is checked to ensure that the field names accurately reflect their business meaning and avoid data misreading due to semantic confusion.

[0075] Business rule validation focuses on checking the value range of numeric fields and the business logic compliance of non-numeric fields (such as operation actions and status flags). This check ensures that business fields such as payment amount and operation type conform to preset business rules, such as non-negative amounts and operation order conforming to business processes, thereby guaranteeing the business consistency of order data.

[0076] After verifying the three sub-results mentioned above, an integration mechanism is used to comprehensively analyze the first, second, and third verification sub-results to form the final verification result. This process not only checks the correctness of the data structure and semantics but also ensures the compliance of the business logic.

[0077] Figure 3 This is a flowchart of an optional data verification module according to an embodiment of the present invention, such as... Figure 3 As shown, this module aims to verify data compliance based on user-provided order XML format data through a structured, layered validation mechanism driven by an intelligent metadata engine. The specific process is as follows:

[0078] Multimodal metadata management is implemented, building a distributed, versioned metadata repository that supports multiple versions of order XML schemas. To improve performance, an LRU-K caching strategy is used to establish high-frequency schema hotspots, achieving millisecond-level response times.

[0079] The process implements a layered verification logic. The core logic of this process follows the "funnel" filtering principle, ensuring data compliance through a progressive verification mechanism.

[0080] The structure verification layer performs consistency checks on the document root node and namespace based on the schema version, intercepting request anomalies caused by version conflicts. Subsequently, XPath localization technology is used to perform a deep scan of the element nesting hierarchy, accurately identifying missing or redundant node structures, such as detecting unclosed tags or illegally inserted fields in order data.

[0081] Semantic validation layer: Based on the minOccurs attribute defined in the schema, mandatory integrity checks are performed on required fields to ensure that core data such as order number (OrderID) and payment amount (PaymentAmount) are not missing.

[0082] Business rules layer: Dynamically configures threshold rules (non-negative amounts, exchange rate sensitive fields, etc.), calculates thresholds in real time, and triggers out-of-bounds warnings. It implements multilingual state machine transitions using finite automata and leverages Unicode-standardized forms to semantically unify internationalized fields (such as confirmed / canceled), eliminating multilingual ambiguity. This layered mechanism uses a static / dynamic separation strategy to front-load 90% of the validation load, ultimately performing strong consistency checks only on valid data, balancing efficiency and accuracy.

[0083] Errors are handled in a tiered manner, establishing a mechanism for categorizing errors into disruptive errors (missing critical fields, structural damage) and non-disruptive errors (early warnings for monetary thresholds). The former triggers process interruption and real-time alarms, while the latter generates risk logs and outputs early warning reports through a visualization module, achieving tiered risk control.

[0084] Step S202: If the verification passes, extract business entity information from the target order data based on a preset extraction strategy, wherein the preset extraction strategy is used to specify the extraction fields and the field association relationships.

[0085] Data extraction involves parsing and extracting business entity information from validated order data based on a pre-defined extraction strategy. This process aims to accurately extract various business entities and their associated information from the order data, ensuring the accuracy and completeness of subsequent analysis. The pre-defined extraction strategy not only specifies the specific fields to be extracted but also defines the relationships between these fields to achieve a deep understanding of the order data and accurate identification of business logic.

[0086] In practice, this is mainly achieved through the following steps:

[0087] Structured parsing: Employing a domain ontology mapping framework and a declarative rule definition mechanism, the system accurately locates core entities such as Product, Payment, Audit, and Remark, enabling structured parsing of heterogeneous order data. By leveraging the precise correspondence between XPath expressions and field key-value pairs, it ensures accurate extraction of key information from nested data structures.

[0088] Semantic Recognition and Association: Based on a multi-level node traversal algorithm, this system penetrates multiple levels of business entities to identify and extract key business fields (such as order number, operation type, operation sequence, product type, and payment information). Simultaneously, through temporal feature extraction, a state transition probability matrix is ​​constructed to quantify the temporal correlation between operation records and product changes / payment records, forming a causal chain between product change events and operation records, as well as a cross-entity causal chain between product changes and payment state transitions. This process ensures the comprehensiveness of business entity information and the clarity of logical relationships.

[0089] Entity information fusion: The entity information extracted through the above steps will be integrated to build a multi-dimensional graph storage model that includes product definition, business operation, payment and refund status, etc., forming a structured data model that provides a solid foundation for subsequent intelligent analysis.

[0090] Optionally, the step of extracting business entity information from target order data based on a preset extraction strategy includes: obtaining an entity field mapping table, and locating the extraction fields specified by the preset extraction strategy in the target order data based on the entity field mapping table to obtain field location information as the location result; traversing the target order data based on field relationships, and performing layered parsing of the target order data based on business dimensions, transaction dimensions, and operation record dimensions during the traversal process to obtain parsing results; and extracting features from the target order data based on the location result and parsing result to obtain business entity information.

[0091] The above steps describe the process of extracting business entity information from target order data. Based on a preset extraction strategy, it utilizes techniques such as entity field mapping tables, field relationship traversal, and hierarchical parsing to achieve accurate data location and analysis. A detailed explanation of this process follows.

[0092] The obtained entity field mapping table is a collection of mapping relationships used to guide the location of specific entity fields in the target order data. This mapping table is pre-designed according to business requirements and data structure, and contains the correspondence between XPath expressions and target fields, ensuring that the exact location of the field to be extracted in the XML document can be found accurately.

[0093] Based on the entity field mapping table, XPath or similar techniques are used to locate fields specified by a pre-defined extraction strategy, obtaining their specific positions in the order data and generating location results. This process ensures that the system can quickly and accurately access key data information.

[0094] Next, the entire target order data is traversed based on field relationships, and layered parsing is performed according to business, transaction, and operation record dimensions. This process is achieved through a multi-level node traversal algorithm, which can penetrate multiple levels such as Product, Payment, Audit, and Remark to extract all related field information and build a complete information framework for the business entity.

[0095] After obtaining the location and analysis results, feature extraction is performed, including extracting the values ​​of key fields and the correlation information between fields, ultimately generating business entity information containing the order subject, transaction characteristics, and operation records. This information will serve as the foundational data for subsequent intelligent analysis and visualization.

[0096] Alternatively, a supplementary step of temporal feature extraction can be added after the above steps: after extracting static business entity information, further analyze the temporal relationships of each entity's operations, construct a state transition probability matrix to quantify the temporal correlation between operation records and product changes and payment records, forming a dynamic reconstruction of business scenarios. This not only enhances the in-depth understanding of business entity information but also provides important temporal logic for intelligent analysis.

[0097] Optionally, the steps of performing layered parsing of the target order data based on business dimension, transaction dimension, and operation record dimension during the traversal process to obtain the parsing results include: in the business dimension, performing subject parsing of the target order data to obtain an order subject set; in the transaction dimension, performing feature parsing of the target order data to obtain a transaction feature set; in the operation record dimension, performing source tracing parsing of the target order data to obtain an operation source tracing set, wherein the operation source tracing set is used to record operation events and corresponding operation records; and generating parsing results based on the order subject set, transaction feature set, and operation source tracing set.

[0098] The steps described above, which involve layering and analyzing target order data based on business, transaction, and operation record dimensions, are the core of the entire data analysis process. This aims to gain a deep understanding of order data from multiple perspectives and accurately extract key information. The following is a detailed analysis of this process.

[0099] Order entity parsing is performed at the business level. This level focuses on the core information of the order, including product definitions (such as flight information, additional services, packaged products, etc.), passenger information, fare structure, and lifecycle status. A multi-level node traversal algorithm penetrates the multi-layered structure of order data to extract all key information related to the order entity, forming an order entity set, providing a foundation of core information for business logic analysis.

[0100] Perform transaction feature analysis at the transaction level. Focus on transaction-related features such as payment status, refund status, transaction amount, and payment channel verification. By performing structured analysis of payment and refund information in order data, accurately identify the entity data of payment and refund orders, construct a transaction feature set, and ensure clear traceability of the payment and refund process.

[0101] Operation tracing and analysis are performed at the operation record level. This involves in-depth analysis of order data operation records, documenting each operation event (such as changes, refunds, payments, etc.) and its corresponding record. This process creates an operation tracing set, fully preserving the time sequence and causal chain of operations, providing crucial data support for anomaly detection and process compliance assessment.

[0102] Finally, based on the analysis results of the above three dimensions (order entity set, transaction feature set, and operation traceability set), a comprehensive analysis result is generated, including all key entity information and their relationships, providing a solid data foundation for subsequent intelligent analysis and visualization.

[0103] Figure 4 This is a flowchart of an optional data extraction module according to an embodiment of the present invention, such as... Figure 4 As shown, this module aims to perform structured parsing of heterogeneous order data by using the valid structured files generated in the previous step based on the rule engine and domain ontology mapping framework, and extract the parsing results through temporal features, finally outputting a graph storage structure with temporal correlation.

[0104] Specifically, the structured parsing of heterogeneous order data has three core technical paths: (1) establishing a precise correspondence between XPath expressions and target features through an ontology-field mapping table; (2) implementing dynamic parsing strategy configuration using a declarative rule definition mechanism; and (3) achieving precise positioning of nested data structures based on a multi-level node traversal algorithm, effectively penetrating multi-level business entities such as Product, Payment, Audit, and Remark. The core functions focus on: order subject parsing (product type, flight segment details, passenger information, fare system and lifecycle status); transaction feature extraction (payment status, refund status, transaction amount, payment channel verification); and operation record tracking (product change records, refund flow traceability, etc.). Then, the layered parsing architecture ensures the complete extraction of key business elements and the maintenance of semantic consistency.

[0105] Next, temporal feature extraction is performed, segmenting the data according to the operation sequence and constructing an operation event sequence code. The temporal correlation between operation records and product changes / payment records is quantified using a state transition probability matrix. The entity modeling system includes node definition and association dimensions. First, four core node types are defined: Product, Payment, Audit, and Remark. Second, the association dimensions are designed as dual association dimensions, including: an operation tracing chain (Product change event → trigger → Audit operation record, Payment action → generation → Audit operation record) and a cross-entity causal chain (Product version iteration → result → Payment state transition). This module ensures the complete extraction of business elements through a layered parsing architecture, ultimately outputting a graph storage structure with temporal correlation to support subsequent process tracing.

[0106] Step S203: Using a time-series analysis algorithm, a time-series association model of the target order is constructed based on the business entity information. The time-series association model is used to record all historical business operations of the target order in time sequence.

[0107] This embodiment focuses on using time-series analysis algorithms to construct a time-series correlation model of the target order during the analysis process. The aim is to capture and reconstruct the time-series relationships of all historical business operations in the order data, providing a logically coherent temporal data view for subsequent business decisions and anomaly detection.

[0108] After extracting the business entity information, time-series analysis algorithms can be used to process this information. This involves constructing a state transition graph with time labels, where each node represents a business operation, such as flight change, payment confirmation, or ticket refund, and edges represent the temporal relationships between different operations. Through algorithmic analysis, the causal relationships between different operations can be identified, maintaining clear traceability of the operation process even in complex and ever-changing business scenarios.

[0109] Constructing a temporal relationship model for the target order involves arranging the operation records in the business entity information in chronological order and analyzing the logical relationships between them. The model construction process involves the following key steps:

[0110] 1) Operation event sequence encoding: Convert each business operation into a specific sequence element, including operation type, operation time, and state changes before and after the operation;

[0111] 2) Construction of the state transition probability matrix: Statistical analysis is used to evaluate the probability of occurrence between business operations, thereby quantifying the temporal correlation in the operation sequence, which is key to the model's understanding of the business operation chain.

[0112] 3) Definition of core entity nodes: Define core entity nodes such as Product, Payment, Audit, and Remark as the main components in the time-series association model;

[0113] 4) Design of dual-relationship dimensions: Not only does it consider the direct operational relationship between entities, but it also incorporates cross-entity causal chain analysis, such as how product changes affect payment status, and how changes in payment status trigger new operations.

[0114] This embodiment can achieve a complete record of all business operations in the order lifecycle by constructing a time-series association model. Even when faced with multiple versions of sub-orders and complex operation logic, it can clearly show the triggering reason, scope of impact and subsequent reaction of each operation, forming a logically rigorous and time-series clear business operation history view.

[0115] Optionally, the steps of constructing a time-series association model of the target order based on business entity information using a time-series analysis algorithm include: performing time-series analysis on the business entity information to obtain time-series records of all historical business operations; generating an initial model framework based on the time-series records and the operation traceability set in the business entity information, wherein the initial model framework contains operation nodes corresponding to historical business operations arranged in time sequence; generating subject nodes based on the order subject set in the business entity information, and generating node labels based on the transaction feature set; supplementing the subject nodes and node labels to the initial model framework, and establishing association lines between subject nodes and operation nodes to obtain the time-series association model.

[0116] The above steps involve using time-series analysis algorithms to construct a time-series association model of the target order based on business entity information, which is a key step in this invention for handling the dynamics and complexity of order data.

[0117] First, a time-series analysis is performed on the business entity information extracted from the order data. This process involves analyzing the sequence of operation events in the operation traceability set to identify the specific time of each business operation (such as flight changes, payment confirmations, etc.) and the changes in the state before and after the operation, forming a time-series record of all historical business operations. This record will serve as the timeline for building the model.

[0118] Based on the aforementioned time-series records and operation traceability set, an initial model framework is generated. This framework contains operation nodes corresponding to historical business operations arranged in chronological order. This step ensures the temporal accuracy of each node in the model, laying the foundation for subsequent node association and model construction.

[0119] Building upon the initial framework, entity nodes are further generated based on the order entity set. These nodes represent various business entities within the order, such as flights, passengers, and payments. Simultaneously, based on the transaction feature set, node labels are added to each operation node. These labels contain key information such as the specific type of operation, amount, and status, enhancing the model's semantic richness and business transparency.

[0120] Finally, the main nodes and node labels are added to the initial model framework, and a complete temporal relationship model is formed by establishing the connection lines between the main nodes and the operation nodes. This model not only presents the temporal logic of business operations, but also associates each operation with a specific business entity, providing intuitive visualization support for anomaly detection and process analysis.

[0121] Alternatively, after the above steps, a multimodal semantic adapter and multi-level node penetration algorithm based on dynamically generated paths can be used. This application can accurately extract entity data such as product definition data and business operation information. Furthermore, through causal chain reconstruction and multi-threaded operation logic breakpoint detection, it can further enrich the information dimension and logical depth of the temporal association model when parsing complex order data.

[0122] Optionally, the step of establishing the association lines between the main nodes and the operation nodes to obtain the temporal association model includes: for each operation event corresponding to the operation node, analyzing the impact of the operation event on the state of the target order, wherein the state impact includes: the sub-influence of the operation event on the state of each main node, and the changes in transaction characteristics caused by the operation event; generating a first type of association line between operation nodes based on the sub-influence of the state; constructing a second type of association line between node labels and model nodes based on the changes in transaction characteristics; and generating the operation causal chain of the target order by analyzing the first type of association line and the second type of association line.

[0123] The process of establishing the connection between the main node and the operation node in the above steps is one of the core steps in building the temporal correlation model, which aims to show the causal relationship between business operations and order status.

[0124] For each operation node, the impact of the operation event on the target order status is analyzed in depth. This impact is divided into two levels: first, the sub-impact on the status of each main node (such as flight, passenger, service item, etc.), including the status changes before and after the operation, such as the flight changing from a booked status to a confirmed status, and the passenger changing from pending payment to paid; second, the analysis of changes in overall transaction characteristics, such as the increase or decrease in payment amount, the switching of payment channels, etc., which can reveal deeper business logic.

[0125] Based on the analysis of the impact of operational event status sub-items, a series of first-type correlation lines are generated between operational nodes. These correlation lines represent the direct causal relationship between operational events, such as a flight change operation directly leading to a subsequent payment status adjustment, or the deletion of a service item triggering a refund operation, etc.

[0126] Based on insights into changes in transaction characteristics, a second type of correlation line was constructed between node labels and model nodes. This type of correlation line focuses more on reflecting the relationship between business operations and specific transaction characteristics. For example, by marking the correlation between payment status change nodes and refund process nodes, the logical relationship between payment and refund can be clearly shown.

[0127] By comprehensively analyzing the first and second types of correlation lines, a causal chain of operations for the target order is constructed. This causal chain not only demonstrates the direct sequential relationship between operational events, but also reveals the indirect impact of operations on order status and transaction characteristics, forming a comprehensive and coherent business operation model.

[0128] Alternatively, after the above steps, causal chain reconstruction can be performed based on dynamic graphs. When analyzing breakpoints in multi-threaded operation logic and constructing time-series model data, the construction of dynamic graphs can more accurately track the evolution of order status, capture hidden changes in business rules, and provide richer details and more precise logical support for the construction of causal chains.

[0129] Optionally, after constructing a time-series correlation model of the target order based on business entity information, the method further includes: integrating business entity information, transaction features, and operation records extracted from the target order data to obtain business summary information, and generating a business report corresponding to the target order data based on the business summary information; comparing and analyzing the time-series correlation model and the business report to obtain the mapping relationship between model elements and business summary information, wherein model elements refer to model nodes and correlation lines; generating an order analysis chart based on the mapping relationship and the business report, and rendering the order analysis chart to the display area.

[0130] This embodiment, after completing the construction of the time-series correlation model, can take additional steps to integrate the business entity information, transaction feature information, and operation record information extracted from the target order data into business summary information, thereby generating a detailed business report. The purpose of this process is to provide a higher-level business overview, making it easier for users to quickly understand the business process and key data points of the order, while deepening the analysis of the order operation logic.

[0131] Specifically, business summary information is a concise summary of the main characteristics and business processes of an order, formed by processing and refining the original order data. It includes at least the order status, the main business entities involved (such as flight information and passenger information), key transaction features (payment status, refund status, etc.), and an overview of operation records.

[0132] The business report is a comprehensive document that goes beyond the business summary information and provides further in-depth analysis. It covers the detailed business logic, operation history, and anomaly detection results of the order. It not only reflects the static characteristics of the order, but also records in detail the business operations, transaction dynamics, and change history related to the order.

[0133] The purpose of comparing and analyzing the time-series correlation model and the business report is to ensure that their representation methods and data content are mutually corroborative. Through comparative analysis, potential omissions in the model or inaccuracies in the report can be identified, allowing for correction and improvement. Establishing a mapping relationship ensures that the descriptions in the business report correspond one-to-one with the graph elements (i.e., model nodes and correlation lines) in the time-series correlation model, guaranteeing a clear expression of business logic and a consistent visual presentation.

[0134] Order analysis charts are generated based on mapping relationships and business report content to further enhance the visualization of business processes. Order analysis charts can include the following formats: flowcharts, used to show the transformation process of order status, emphasizing the sequence of operations and status changes; timeline charts, which visually show the time points of each operation, assisting in the analysis of the temporal relationships between operations; Gantt charts, particularly suitable for showing the duration of order operations, helping to identify long-cycle or delayed operations; bar charts, line charts, or pie charts, used to statistically analyze and compare key data in orders, such as payment amount distribution and refund ratios; and heatmaps, used to show the frequency distribution of different order statuses or operations, helping to identify hot or cold spots in business activities.

[0135] Figure 5 This is a flowchart of an optional intelligent analysis module according to an embodiment of the present invention, such as... Figure 5 As shown, this module aims to achieve enhanced reasoning from multi-source data by integrating the topological relationships between time-series graphs and knowledge graphs, based on RAG (Retrieval Enhanced Generation) technology, and output structured decision support reports. Its implementation logic consists of a three-layer architecture:

[0136] 1) Knowledge graph construction layer: Based on requirement documents, solution design documents, bug fix documents and daily business summaries, Python-docx and PyPDF2 are used to parse and analyze multiple document formats. An improved BERT multi-task model (integrating entity recognition and relationship classification dual-head structure) is used to accurately extract functional modules, business entities and constraints. Semantic disambiguation is achieved by combining a predefined JSON thesaurus and edit distance algorithm. Constraint verification is completed by regular expressions based on YAML rule tables (such as verifying the compliance of "adults carrying ≤1 infants"). Finally, a traceable entity relationship network and adaptive business topology graph are built based on Neo4j.

[0137] 2) The RAG enhanced analysis layer uses a hybrid index (Faiss + Elasticsearch) to dynamically retrieve triples from the internal knowledge graph and the external knowledge base. It constructs a context-aware network by integrating the self-attention layer of LLM through a bidirectional attention mechanism, performs multi-hop inference path generation, and triggers pruning strategies based on confidence thresholds. It outputs a structured report with logical chain labels and semantic metadata.

[0138] 3) At the business decision-making layer, a multimodal fusion classifier (text semantic vectors + graph embedding) is deployed to distinguish between anomalies and business reports. Anomaly reports are blocked from graphical output through a three-level logical verification chain (spatiotemporal constraint verification, attribute conflict detection, and rule engine filtering). Business reports, on the other hand, activate the parameterized template engine driven by D3.js to dynamically bind the graph embedding vectors to the schema, generating a topology-adaptive visual graph. Simultaneously, a semantic routing mechanism is used to match the mapping relationship between knowledge subgraphs and visual elements in real time, forming a feedback optimization closed loop based on reinforcement learning. This module achieves deep collaboration between knowledge-driven decision-making and visual representation.

[0139] Step S204: Convert the time-series correlation model into a business process diagram and render the business process diagram to the display area.

[0140] The key to graphical representation is to transform the constructed temporal relationship model into an intuitive business process diagram and render it in the display area to achieve visualization and understandability of business operations.

[0141] An optional one, based on Figure 1 The diagram illustrating the structure of the full-process analysis system for airline orders also includes a visualization module. After constructing the temporal relationship model, this module transforms the nodes and their interrelationships into a more easily understood graphical representation—a business process diagram. This involves mapping operation nodes and main nodes to specific graphical elements in the flowchart, such as rectangles, circles, or arrows, while preserving the temporal sequence and state influence relationships of the operation events. Through this transformation, complex temporal relationships and operational processes are clearly displayed, facilitating business personnel and analysts to quickly locate problems and understand the business logic.

[0142] After the business process diagram is generated, it is rendered to a dedicated display area, which is usually the visual module in the user interface. The rendering process may involve adjusting the position of graphic elements, optimizing visual effects, and ensuring the readability and aesthetics of the flowchart. In addition, different color codes and labels should be used according to the nature of the operation events (such as normal operations, abnormal events, etc.) so that the business process diagram can intuitively reflect changes in order status and potential abnormal situations.

[0143] Figure 6 This is a flowchart of an optional visualization module according to an embodiment of the present invention, such as... Figure 6 As shown, this module aims to map the content extracted from the analysis report in the previous step into graphical elements and generate a directed graph of order operations that reflects the process flow and status changes through interactive rendering, and then embed it into the business report. The Graphviz-based order operation flowchart design adopts a vertical directed graph structure to achieve multi-dimensional operation tracking, constructing a hierarchical operation view through the following five core modules:

[0144] 1. The structured layout is initialized using a top-bottom split-column design (rankdir=LR). The upper fixed area creates an independent sub-graph, cluster_report, to display static business reports (including key content such as flight information, passenger information, payment information, and service information, while also drawing a timeline of process stages). The lower main graph displays the dynamic process. The main graph uses record-shaped nodes (shape='record') to encapsulate composite data, defining hollow arrowheads (arrowhead='empty') to ensure that process symbols conform to ISO standards. This layout strategy effectively maps the time-dimensional evolution of order operations by arranging nodes in a logical order.

[0145] 2. Dynamic data-driven node generation: Based on a large model, operation node parsing is returned, dynamically extracting 23 key fields such as operation ID, operation action, and associated product status / type / flight segment number. A color coding mechanism (e.g., historical data identifiers trigger black / red tones) is adopted to construct standardized node labels {ModularProduct|field key-value pairs}, achieving visualization and dimensionality reduction of complex business data.

[0146] 3. A three-level nested subgraph architecture is used to create a hierarchical structure of cluster_audit_X→cluster_mp_X→cluster_remark_X: The main operation container cluster_audit_X wraps single-step operations with a gray background and enhances visual grouping with a 2px border; the product subgraph cluster_mp_X parses 15+ business fields and processes refund information (generating the {refundTicket|ticket details} node) and payment data information (generating the {payment|payment details) node) through the three-level subgraph cluster_remark_X. The payment / refund subgraphs map payment status (amount / type / historical identifier) ​​and refund information (status / amount / activation identifier) ​​respectively, realizing dynamic coloring driven by field values.

[0147] 4. Optimized temporal connections and output: Node connections are established according to the operation record ID order, and the operationAction is displayed on the edge label. During the rendering stage, the Graphviz engine is used to generate 1024x1024 resolution PNGs, and the preview mode is disabled to adapt to batch generation scenarios. The exception handling module implements SVG format degradation output to ensure high concurrency stability. This solution uses structured DOT script generation technology to integrate multi-source data such as order operations, product changes, and payment trajectories into a unified visual narrative.

[0148] 5. Use Graphviz's gvpack tool to merge multiple DOT files, ultimately generating a composite visual document containing business context and process operation records.

[0149] Through the above steps S201 to S204, the target order data to be parsed can be received first, and the target order data can be verified based on a preset verification strategy. The preset verification strategy includes at least structural verification and format verification. If the verification passes, business entity information can be extracted from the target order data based on a preset extraction strategy. The preset extraction strategy is used to specify the extraction fields and field relationships. Then, a time-series analysis algorithm is used to construct a time-series association model of the target order based on the business entity information. The time-series association model is used to record all historical business operations of the target order in time sequence. Finally, the time-series association model is converted into a business process diagram and rendered to the display area.

[0150] In this embodiment, a multi-layered data verification approach is adopted. Through a dynamic rule engine and metadata repository, the consistency of order data structure and semantics is ensured, thereby improving the accuracy and efficiency of data parsing. Specifically, this embodiment combines entity field mapping tables and hierarchical parsing algorithms to achieve in-depth mining and precise positioning of business entity information. It effectively integrates complex data from three dimensions: order subject, transaction characteristics, and operation records, ensuring the comprehensiveness and accuracy of the data. Subsequently, a time-series association model is constructed using a time-series analysis algorithm. This model can accurately record the change trajectory of the entire order lifecycle according to the operation time sequence, facilitating in-depth analysis and anomaly localization in dynamic business scenarios. Finally, the time-series association model is converted into a business process diagram, and visualization technology is used to transform the abstract order data change process into an intuitive process display, enhancing the readability and understandability of the information. Overall, this embodiment, through a four-stage architecture of intelligent verification, precise extraction, time-series analysis, and interactive visualization, significantly improves the automation level of airline order data parsing and the intelligence level of problem diagnosis. It provides airlines with an efficient and accurate order management and analysis solution, thereby solving the technical problem of poor integration capabilities of complex order process data in related technologies, leading to low data parsing accuracy.

[0151] The invention will now be described in conjunction with another alternative embodiment.

[0152] Example 2

[0153] The air order parsing device provided in this embodiment includes multiple implementation units, each of which corresponds to a specific implementation step in Embodiment 1 above.

[0154] Figure 7 This is a schematic diagram of an optional air ticket parsing device according to an embodiment of the present invention, such as... Figure 7 As shown, the device may include: a verification unit 71, an extraction unit 72, a construction unit 73, and a display unit 74.

[0155] The verification unit 71 is used to receive the target order data to be parsed and to verify the target order data based on a preset verification strategy. The preset verification strategy includes at least: structure verification and format verification.

[0156] Extraction unit 72, if the verification passes, extracts business entity information from the target order data based on a preset extraction strategy, wherein the preset extraction strategy is used to specify the extraction fields and the field association relationships.

[0157] Construction unit 73 is used to construct a time-series association model of the target order based on business entity information using a time-series analysis algorithm. The time-series association model is used to record all historical business operations of the target order in time sequence.

[0158] Display unit 74 is used to convert the time-series correlation model into a business process diagram and render the business process diagram to the display area.

[0159] The aforementioned airline order parsing device can first receive the target order data to be parsed through the verification unit 71 and verify the target order data based on a preset verification strategy. The preset verification strategy includes at least structural verification and format verification. Then, if the verification passes, the extraction unit 72 extracts business entity information from the target order data based on a preset extraction strategy. The preset extraction strategy is used to specify the extraction fields and field relationships. Then, the construction unit 73 uses a time-series analysis algorithm to construct a time-series association model of the target order based on the business entity information. The time-series association model is used to record all historical business operations of the target order in time sequence. Finally, the display unit 74 converts the time-series association model into a business process diagram and renders the business process diagram to the display area.

[0160] In this embodiment, a multi-layered data verification approach is adopted, utilizing a dynamic rule engine and metadata repository to ensure the consistency of order data structure and semantics, thereby improving the accuracy and efficiency of data parsing. Specifically, this embodiment combines entity field mapping tables and hierarchical parsing algorithms to achieve in-depth mining and precise positioning of business entity information, effectively integrating complex data from three dimensions: order subject, transaction characteristics, and operation records, ensuring the comprehensiveness and accuracy of the data. Subsequently, a time-series association model is constructed using a time-series analysis algorithm. This model can accurately record the change trajectory of the entire lifecycle of the order according to the operation time sequence, assisting in in-depth analysis and anomaly localization of dynamic business scenarios. Finally, the time-series association model is converted into a business process diagram, using visualization technology to transform the abstract order data change process into an intuitive process display, enhancing the readability and understandability of the information. Overall, this embodiment, through a four-stage architecture of intelligent verification, precise extraction, time-series analysis, and interactive visualization, significantly improves the automation level of aviation order data parsing and the intelligence level of problem diagnosis, providing airlines with an efficient and accurate order management and analysis solution. This solves the technical problem of poor integration capability of aviation systems for complex order process data, leading to low data parsing accuracy.

[0161] Optionally, the verification unit includes: an acquisition module, used to acquire the latest version of the data specification from the metadata repository, wherein the metadata repository is a collection of data specifications that are updated periodically, and the data specifications are used to define structural standards, semantic constraints and business rules; and a hierarchical verification module, used to perform hierarchical verification on the target order data based on the structural standards, semantic constraints and business rules, and obtain the verification results.

[0162] Optionally, the hierarchical verification module includes: a first verification submodule, used to verify all fields in the target order data based on the path expression defined in the structural standard, to obtain a first verification sub-result, wherein the path expression at least specifies the following: root node, namespace, and nested field elements; a second verification submodule, used to verify all fields in the target order data based on the business thesaurus defined in the semantic constraints, to obtain a second verification sub-result, wherein the semantic constraints at least specify the following: field naming and field semantic consistency; a third verification submodule, used to verify the business fields in the target order data based on the business rules defined in the business rules, to obtain a third verification sub-result, wherein the business rules at least specify the following: the value range of numeric fields, and the operational compliance of non-numeric fields; and an integration submodule, used to integrate the first verification sub-result, the second verification sub-result, and the third verification sub-result to obtain a verification result.

[0163] Optionally, the extraction unit includes: a positioning module, used to obtain an entity field mapping table and locate the extraction fields specified by a preset extraction strategy in the target order data based on the entity field mapping table, and obtain the field location information as the positioning result; a hierarchical parsing module, used to traverse the target order data based on field relationships, and perform hierarchical parsing of the target order data based on business dimensions, transaction dimensions, and operation record dimensions during the traversal process, and obtain the parsing result; and a feature extraction module, used to extract features from the target order data based on the positioning result and the parsing result, and obtain business entity information.

[0164] Optionally, the hierarchical parsing module includes: a first parsing submodule, used to perform subject parsing on the target order data in the business dimension to obtain an order subject set; a second parsing submodule, used to perform feature parsing on the target order data in the transaction dimension to obtain a transaction feature set; a third parsing submodule, used to perform source tracing parsing on the target order data in the operation record dimension to obtain an operation source tracing set, wherein the operation source tracing set is used to record operation events and corresponding operation records; and a first generation submodule, used to generate parsing results based on the order subject set, the transaction feature set, and the operation source tracing set.

[0165] Optionally, the construction unit includes: an analysis module for performing time-series analysis on business entity information to obtain time-series records of all historical business operations; a first generation module for generating an initial model framework based on the time-series records and the operation traceability set in the business entity information, wherein the initial model framework contains operation nodes corresponding to historical business operations arranged in time sequence; a second generation module for generating subject nodes based on the order subject set in the business entity information and generating node labels based on the transaction feature set; and an establishment module for supplementing the subject nodes and node labels into the initial model framework and establishing the association lines between subject nodes and operation nodes to obtain a time-series association model.

[0166] Optionally, the establishment module includes: an analysis submodule, used to analyze the impact of operation events on the state of the target order for each operation node, wherein the state impact includes: the sub-influence of the operation event on the state of each subject node, and the changes in transaction characteristics caused by the operation event; a second generation submodule, used to generate a first type of correlation between operation nodes based on the sub-influence of the state; a construction submodule, used to construct a second type of correlation between node labels and model nodes based on the changes in transaction characteristics; and a third generation submodule, used to generate the operation causal chain of the target order by analyzing the first type of correlation and the second type of correlation.

[0167] Furthermore, the airline order parsing device also includes: an integration module, used to integrate the business entity information, transaction features, and operation records extracted from the target order data after constructing a time-series correlation model of the target order based on business entity information, to obtain business summary information, and to generate a business report corresponding to the target order data based on the business summary information; a comparison and analysis module, used to compare and analyze the time-series correlation model and the business report to obtain the mapping relationship between model elements and business summary information, where model elements refer to model nodes and correlation lines; and a rendering module, used to generate an order analysis chart based on the mapping relationship and the business report, and to render the order analysis chart to the display area.

[0168] The aforementioned air order parsing device may also include a processor and a memory. The aforementioned verification unit 71, extraction unit 72, construction unit 73, display unit 74, etc., are all stored in the memory as program units, and the processor executes the aforementioned program units stored in the memory to realize the corresponding functions.

[0169] The aforementioned processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and by adjusting kernel parameters, a timing analysis algorithm is employed to construct a timing association model of the target order based on business entity information. This timing association model records all historical business operations of the target order in chronological order. The timing association model is then converted into a business process diagram, which is rendered to the display area.

[0170] The aforementioned memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0171] This application also provides a computer program product, which, when executed on a data processing device, is suitable for executing an initialization program with the following method steps: receiving target order data to be parsed, and validating the target order data based on a preset validation strategy, wherein the preset validation strategy includes at least: structure validation and format validation; if the validation passes, extracting business entity information from the target order data based on a preset extraction strategy, wherein the preset extraction strategy is used to specify the extraction fields and field relationships; using a time-series analysis algorithm, constructing a time-series association model of the target order based on the business entity information, wherein the time-series association model is used to record all historical business operations of the target order in time sequence; converting the time-series association model into a business process diagram, and rendering the business process diagram to the display area.

[0172] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the computer-readable storage medium is located to execute the air order parsing method of any one of the above embodiments.

[0173] According to another aspect of the present invention, an electronic device is also provided, including one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the air order parsing method of any one of the above embodiments.

[0174] Figure 8 This is a structural block diagram of an electronic device for executing an airline order parsing method according to an embodiment of the present invention, such as... Figure 8 As shown, the electronic device may include: one or more ( Figure 8 (Only one is shown) processor 802, memory 804, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module and the display.

[0175] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the air order parsing method and apparatus in this embodiment. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the aforementioned air order parsing method. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0176] Those skilled in the art will understand that Figure 8 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 8 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 8 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 8 The different configurations shown.

[0177] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0178] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0179] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0180] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0181] 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 units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0182] Furthermore, 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. The integrated unit can be implemented in hardware or as a software functional unit.

[0183] If the integrated unit 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 the present invention, in essence, or the part that contributes to the prior art, or all or 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 described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0184] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle 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 parsing airline orders, characterized in that, include: Receive the target order data to be parsed, and verify the target order data based on a preset verification strategy, wherein the preset verification strategy includes at least: structure verification and format verification; If the verification passes, business entity information is extracted from the target order data based on a preset extraction strategy, wherein the preset extraction strategy is used to specify the extraction fields and the field association relationships; A time-series analysis algorithm is used to construct a time-series association model of the target order based on the business entity information. The time-series association model is used to record all historical business operations of the target order in time sequence. The time-series correlation model is converted into a business process diagram, and the business process diagram is rendered to the display area.

2. The analytical method according to claim 1, characterized in that, The steps for validating the target order data based on a preset validation strategy include: The latest version of the data specification is obtained from the metadata repository, which is a collection of data specifications that are updated regularly. The data specifications are used to define structural standards, semantic constraints, and business rules. The target order data is subjected to hierarchical verification based on the structural standards, semantic constraints, and business rules to obtain the verification results.

3. The analytical method according to claim 2, characterized in that, The steps for performing layered validation on the target order data based on the structural standards, semantic constraints, and business rules to obtain validation results include: Based on the path expression defined in the structural standard, all fields in the target order data are validated to obtain the first validation sub-result. The path expression at least specifies the following: root node, namespace, and nested field elements. Based on the business thesaurus defined in the semantic constraints, all fields in the target order data are validated to obtain a second validation sub-result. The semantic constraints at least standardize the following: field naming and field semantic consistency. Based on the business rules defined in the business rules, the business fields in the target order data are validated to obtain a third validation sub-result. The business rules at least regulate the following: the value range of numeric fields, and the compliance of operations on non-numeric fields. The verification result is obtained by integrating the first verification result, the second verification result, and the third verification result.

4. The analytical method according to claim 1, characterized in that, The steps for extracting business entity information from the target order data based on a preset extraction strategy include: Obtain the entity field mapping table, and locate the extraction field specified by the preset extraction strategy in the target order data based on the entity field mapping table, and obtain the field location information as the location result; Based on the field relationships, the target order data is traversed, and during the traversal, the target order data is parsed hierarchically based on business dimension, transaction dimension, and operation record dimension to obtain the parsing results; Based on the location and analysis results, feature extraction is performed on the target order data to obtain the business entity information.

5. The analytical method according to claim 4, characterized in that, The steps involved in performing layered parsing of the target order data based on business, transaction, and operation record dimensions during the traversal process to obtain the parsing results include: In the aforementioned business dimension, the target order data is parsed to obtain a set of order subjects; In the transaction dimension, feature parsing is performed on the target order data to obtain a transaction feature set; In the operation record dimension, the target order data is analyzed for traceability to obtain an operation traceability set, wherein the operation traceability set is used to record operation events and corresponding operation records; The parsing result is generated based on the order subject set, the transaction feature set, and the operation traceability set.

6. The analytical method according to claim 1, characterized in that, The steps of constructing a time-series association model of the target order based on the business entity information using a time-series analysis algorithm include: Perform time-series analysis on the business entity information to obtain time-series records of all historical business operations; An initial model framework is generated based on the operation traceability set in the time sequence records and the business entity information, wherein the initial model framework contains operation nodes corresponding to the historical business operations arranged in time sequence; A main node is generated based on the set of order subjects in the business entity information, and a node label is generated based on the set of transaction features. The main node and the node label are added to the initial model framework, and the association lines between the main node and the operation node are established to obtain the temporal association model.

7. The analytical method according to claim 6, characterized in that, The steps of establishing the association lines between the main node and the operation node to obtain the temporal association model include: For each operation event corresponding to the operation node, analyze the impact of the operation event on the state of the target order, wherein the state impact includes: the sub-impact of the operation event on the state of each main node, and the changes in transaction characteristics caused by the operation event; Based on the state sub-influence, a first type of association line is generated between the operation node and the operation node; Based on the changes in the transaction characteristics, a second type of correlation line is constructed between the node labels and the model nodes; By analyzing the first type of correlation line and the second type of correlation line, the operational causal chain of the target order is generated.

8. The analytical method according to claim 1, characterized in that, After constructing a time-series association model of the target order based on the business entity information, the method further includes: The business entity information, transaction characteristics, and operation records extracted from the target order data are integrated to obtain business summary information, and a business report corresponding to the target order data is generated based on the business summary information. By comparing and analyzing the time-series correlation model and the business report, the mapping relationship between model elements and business summary information is obtained, wherein the model elements refer to model nodes and correlation lines; Based on the mapping relationship and the business report, an order analysis chart is generated and rendered to the display area.

9. An apparatus for analyzing airline orders, characterized in that, include: The verification unit is used to receive the target order data to be parsed and to verify the target order data based on a preset verification strategy, wherein the preset verification strategy includes at least: structure verification and format verification. An extraction unit is used to extract business entity information from the target order data based on a preset extraction strategy when the verification passes. The preset extraction strategy is used to specify the extraction fields and the field association relationships. A construction unit is used to construct a time-series association model of the target order based on the business entity information using a time-series analysis algorithm, wherein the time-series association model is used to record all historical business operations of the target order in time sequence; The display unit is used to convert the time-series correlation model into a business process diagram and render the business process diagram to the display area.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform the method for parsing airfare orders as described in any one of claims 1 to 8.

11. An electronic device, characterized in that, It includes one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method for parsing airfare orders as described in any one of claims 1 to 8.

Images