Method and device for constructing a buried point knowledge graph, equipment, medium and product

Abstract

This disclosure relates to the field of computer technology and discloses a method, apparatus, device, medium, and product for constructing a tracking knowledge graph. The method includes: acquiring structured data corresponding to tracking points and tracking events corresponding to tracking points; annotating key parameters of tracking points corresponding to tracking events and adding the annotation information of the key parameters to the structured data to obtain structured annotated data; extracting tracking entities corresponding to tracking events, the relationships between tracking entities, and the attribute information corresponding to tracking entities from the structured annotated data; and constructing a tracking knowledge graph according to the tracking entities, relationships, and attribute information. By implementing the technical solution of this disclosure, the dependencies of each tracking point can be clearly defined, accurately capturing the contextual dependencies between tracking entities, avoiding erroneous knowledge associations, improving the accuracy of tracking knowledge graph construction, and saving the construction and storage costs of large-scale tracking knowledge graphs.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, specifically to a method, apparatus, device, medium, and product for constructing a knowledge graph with embedded data points. Background Technology

[0002] In the current process of event tracking testing, the business team mainly relies on manual testing and regression testing to gradually verify the test case reporting of each core event tracking point. However, this approach has low testing efficiency, resulting in high construction costs for the knowledge graph. Although retrieval enhancement generation methods that rely on large models can improve the efficiency of knowledge retrieval, they mainly depend on cutting documents into independent text blocks, leading to the loss of contextual relationships and affecting the accuracy of knowledge graph construction. Summary of the Invention

[0003] In view of this, this disclosure provides a method, apparatus, equipment, medium and product for constructing a knowledge graph based on event tracking, in order to solve the problem of poor construction effect of the knowledge graph based on event tracking.

[0004] Firstly, this disclosure provides a method for constructing a tracking knowledge graph, including: acquiring structured data corresponding to the tracking points and tracking events corresponding to the tracking points; annotating the key parameters of the tracking points corresponding to the tracking events and adding the annotation information of the key parameters to the structured data to obtain structured annotated data; extracting the tracking entities corresponding to the tracking events, the relationships between tracking entities, and the attribute information corresponding to the tracking entities from the structured annotated data; and constructing a tracking knowledge graph according to the tracking entities, relationships, and attribute information.

[0005] Secondly, this disclosure provides a device for constructing a tracking knowledge graph, comprising: an acquisition module for acquiring structured data corresponding to tracking points and tracking events corresponding to tracking points; an annotation module for annotating key parameters of tracking points corresponding to tracking events and adding the annotation information of key parameters of tracking points to structured data to obtain structured annotated data; an extraction module for extracting tracking entities corresponding to tracking events, the relationships between tracking entities, and the attribute information corresponding to tracking entities from the structured annotated data; and a graph construction module for constructing a tracking knowledge graph according to tracking entities, relationships, and attribute information.

[0006] Thirdly, this disclosure provides a computer device, including: a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to perform the method for constructing a knowledge graph based on the first aspect or any corresponding embodiment described above.

[0007] Fourthly, this disclosure provides a computer-readable storage medium storing computer instructions for causing a computer to execute the method for constructing a knowledge graph of embedded points according to the first aspect or any corresponding embodiment described above.

[0008] Fifthly, this disclosure provides a computer program product, including computer instructions for causing a computer to execute the method for constructing a knowledge graph of embedded points as described in the first aspect or any corresponding embodiment.

[0009] The method, apparatus, equipment, medium, and product for constructing a tracking knowledge graph disclosed herein acquire structured data corresponding to tracking points and annotate key parameters of tracking events. The annotation information of these key parameters is added to the structured data to obtain structured annotated data, enabling multi-dimensional tagging of tracking points for different business needs. Relying on the structured data of the tracking points themselves and the annotation method of key parameters, it is easy to cluster similar tracking entities based on semantics and business knowledge, facilitating accurate perception or link analysis of reused tracking entities or parameters, significantly reducing the inference cost of tracking linking. By extracting tracking entities corresponding to tracking events, the relationships between tracking entities, and the attribute information corresponding to tracking entities from the structured annotated data, a tracking knowledge graph is constructed according to the tracking entities, relationships, and attribute information. This clarifies the dependencies of each tracking point, accurately captures the contextual dependencies between tracking entities, avoids erroneous knowledge associations, improves the accuracy of tracking knowledge graph construction, and saves on the construction and storage costs of large-scale tracking knowledge graphs. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the specific embodiments or related technologies of this disclosure, the accompanying drawings used in the description of the specific embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a flowchart illustrating the method for constructing a knowledge graph based on an embodiment of this disclosure;

[0012] Figure 2 This is a schematic diagram illustrating the setting of association relationships according to embodiments of this disclosure;

[0013] Figure 3 This is a flowchart illustrating another method for constructing a knowledge graph based on an embodiment of this disclosure;

[0014] Figure 4This is a flowchart illustrating another method for constructing a knowledge graph based on an embodiment of the present disclosure;

[0015] Figure 5 This is a schematic diagram of a knowledge graph based on an embodiment of the present disclosure;

[0016] Figure 6 This is an abstract schematic diagram of a knowledge graph based on an embodiment of this disclosure;

[0017] Figure 7 This is a structural block diagram of an apparatus for constructing a knowledge graph based on an embodiment of the present disclosure;

[0018] Figure 8 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present disclosure. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0020] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0021] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.

[0022] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0023] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0024] It is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0025] In current practices, event tracking testing primarily relies on business QA teams to manually test and perform regression testing to progressively verify the reporting status of each core event tracking test case. This approach is inefficient, requiring significant weekly investment of manpower and time for event tracking testing; furthermore, it fails to detect or trace potentially reusable event tracking points or parameters, making it a black-box rather than a white-box test, thus failing to eliminate potential risks.

[0026] With the development of large model technology, the Retrieval Augmentation (RAG) method has improved the efficiency of knowledge retrieval. However, its reliance on splitting documents into independent text chunks leads to the loss of contextual relationships, affecting accuracy. To address this issue, GraphRAG introduces knowledge graphs, extracting entities and relationships through LLM to construct a graph structure, enhancing the semantic understanding and visualization of query results. However, GraphRAG still relies on semantic segmentation to define entities, which may result in semantic gaps. In enterprise database construction, the failure to accurately capture the contextual dependencies between entities during semantic segmentation can easily lead to erroneous knowledge associations, and the cost of building and storing large-scale graphs is very high.

[0027] Neither current purely manual methods for ensuring the quality of event tracking nor machine-based GraphRAG knowledge graph technology can effectively address two key pain points in event tracking testing: ensuring efficiency and identifying potential data risks. Therefore, it is necessary to address the issues of erroneous associations caused by semantic segmentation in knowledge graph construction using LLM and the excessively high cost of building knowledge graphs.

[0028] Based on this, the disclosed technical solution utilizes the structured data inherent in the event tracking data, combined with a semantic lexicon and tagging methods, to enable LLM to train a tagging model for different event tracking entities based on pre-existing business knowledge. This allows for tagging from the dimensions of key event tracking parameters (business scenario, entry point, requirement). Then, by establishing event tracking entities (end-entry point, requirement, event tracking, parameter, dashboard, business, monitoring, test case) corresponding to the event tracking events, and using these entities as entity nodes, relationship edges are established through identical event tracking names, parameter names, and tagging results, generating an event tracking knowledge graph. This allows for the rapid identification of the dependencies of each event tracking point on business scenarios, test cases, and core metrics, thus clarifying the importance of the event tracking points and potential pathways.

[0029] Furthermore, data analysts can use the knowledge graph derived from event tracking to quickly find historical event tracking points with similar requirements and reuse them, reducing the time cost of repetitive design and conceptualization. Testers can analyze the call chain, impact, and priority of event tracking points based on their dependencies, reducing redundancy and repetitive testing costs. This saves on event tracking design and testing costs.

[0030] Furthermore, compared to GraphRAG's approach of defining entities through semantic segmentation, relying on the structured data of the event tracking points themselves and using annotation models to label key parameters of the tracking points significantly reduces the overhead of event tracking inference. Moreover, limiting the number of relational edges through code logic reduces the storage cost of the event tracking knowledge graph. Simultaneously, using annotation models to label key parameters of the tracking points facilitates the clustering of similar event tracking entities based on semantics and business knowledge, improving the accuracy of the event tracking knowledge graph.

[0031] Furthermore, the constructed event tracking knowledge graph clearly demonstrates the dependencies between each event tracking point and business scenarios, test cases, and key metrics. This dependency structure prioritizes identifying event tracking points that significantly impact core metrics, ensuring their accuracy and stability. Simultaneously, the event tracking knowledge graph supports dynamic updates, automatically adjusting based on real-time monitored event tracking trigger frequency, error rate, and other data. This optimizes event tracking design, improves the accuracy of event tracking data, and enhances business decision support capabilities.

[0032] According to the embodiments of this disclosure, an embodiment of a method for constructing a knowledge graph based on data points is provided. It should be noted that the steps shown in the flowcharts of 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 flowcharts, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0033] This embodiment provides a method for constructing a knowledge graph based on data tracking points, which can be used in computer devices such as computers and servers. Figure 1 This is a flowchart of a method for constructing a knowledge graph based on embodiments of this disclosure, such as... Figure 1 As shown, the process includes the following steps:

[0034] Step S101: Obtain the structured data corresponding to the tracking point and the tracking event corresponding to the tracking point.

[0035] Structured data is data stored using a specific, predefined data structure, such as JSON format. Event tracking refers to setting specific tracking points in a webpage or application through code to track and record specific user interactions. For example, event tracking could be a user clicking a button or performing another action that quantifies user interaction.

[0036] Specifically, corresponding tracking events are constructed for business scenarios requiring tracking, such as setting tracking events in scenarios where a user clicks a button. For each tracking event, a unified data structure is defined, and code is written in the corresponding location in the application to implement the tracking based on the designed data structure. Simultaneously, computer devices can collect tracking-related data generated for the tracking events through communication interfaces and store this data in the corresponding database according to the defined data structure, thus achieving the structuring of tracking-related data and obtaining structured data for each tracking event. The data structure may include fields such as event name, event timestamp, and page URL; these are not specifically limited here, and those skilled in the art can determine them according to actual needs.

[0037] Step S102: Annotate the key parameters of the tracking point corresponding to the tracking point event, and add the annotation information of the key parameters of the tracking point to the structured data to obtain the structured annotation data.

[0038] Key event tracking parameters are the relevant parameters required to build an event tracking knowledge graph, specifically including requirements, entry points, and business types. Annotation information consists of tags generated for these key event tracking parameters; different key event tracking parameters correspond to different annotation information.

[0039] Since different event tracking points have different tracking parameters, a pre-trained annotation model is used to annotate these parameters, generating corresponding annotation information on the key tracking parameters. This annotation information is then added to the relevant positions in the structured data, resulting in structured annotated data with added annotations.

[0040] The annotation model is pre-trained using business tracking data. This annotation model can be trained based on a large model architecture or a neural network architecture. There is no limitation here, as long as it can achieve the parameter annotation function.

[0041] Step S103: Extract the tracking entities corresponding to the tracking events, the relationships between tracking entities, and the attribute information corresponding to the tracking entities from the structured annotation data.

[0042] An event tracking entity is a specific object or interaction that is tracked during the event tracking process; the relationships between event tracking entities represent the relationships between different event tracking entities, such as dependency relationships, inclusion relationships, monitoring relationships, etc. These relationships are pre-defined before building the event tracking knowledge graph, such as... Figure 2 As shown; the attribute information corresponding to the embedded entity is descriptive information for the embedded entity.

[0043] Specifically, for different business scenarios, each event tracking point is pre-defined with corresponding entity objects, such as business logic, endpoint entry point, requirement, tracking point, parameter, dashboard, monitoring, and test cases. Each entity corresponding to a tracking point event has been structured according to a pre-defined data structure. The system identifies tracking point entities within the structured labeled data, converting each piece of structured labeled data into a corresponding tracking point entity. For example, each tracking point uses its name as its ID, its type is set to "event," and other core properties are set as attributes. This enables the identification of tracking point entities and their corresponding attribute information for structured labeled data in different business scenarios. Simultaneously, the relationships between different tracking points in the structured labeled data are analyzed to determine the associations between different tracking point entities.

[0044] Step S104: Construct a knowledge graph of the tracking points according to the tracking point entities, relationships, and attribute information.

[0045] Each tracked entity extracted from the structured labeled data is used as a node. Relationship edges between tracked entities are generated according to the triplet of "tracked entity-relationship-tracked entity" and the pair of tracked entity-attribute information. All tracked entities are connected according to these relationship edges to generate a tracked knowledge graph.

[0046] The method for constructing a tracking knowledge graph provided in this embodiment obtains structured data corresponding to tracking points and annotates key parameters of the tracking events. The annotation information of these key parameters is added to the structured data to obtain structured annotated data, enabling multi-dimensional tagging of tracking points for different business needs. Relying on the structured data of the tracking points themselves and the annotation method of key parameters, it is easy to cluster similar tracking entities based on semantics and business knowledge. This allows for accurate perception or link analysis of reused tracking entities or parameters, significantly reducing the inference cost of tracking links. By extracting tracking entities corresponding to tracking events, the relationships between tracking entities, and the attribute information corresponding to tracking entities from the structured annotated data, a tracking knowledge graph is constructed according to the tracking entities, relationships, and attribute information. This clarifies the dependencies of each tracking point, accurately captures the contextual dependencies between tracking entities, avoids erroneous knowledge associations, improves the accuracy of tracking knowledge graph construction, and saves on the construction and storage costs of large-scale tracking knowledge graphs.

[0047] This embodiment provides a method for constructing a knowledge graph based on data tracking points, which can be used in computer devices such as computers and servers. Figure 3 This is a flowchart of a method for constructing a knowledge graph based on embodiments of this disclosure, such as... Figure 3 As shown, the process includes the following steps:

[0048] Step S201: Obtain the structured data corresponding to the tracking point and the tracking event corresponding to the tracking point. For details, please refer to the relevant descriptions of the corresponding steps in the above embodiments, which will not be repeated here.

[0049] Step S202: Annotate the key parameters of the tracking points corresponding to the tracking events, and add the annotation information of the key parameters of the tracking points to the structured data to obtain structured annotation data.

[0050] Specifically, step S202 includes:

[0051] Step S2021: Detect whether the event parameters corresponding to the tracking event carry tracking key parameters, and / or semantically related parameters of the tracking key parameters.

[0052] Event parameters are parameters related to the event being tracked, and may include the event name, endpoint, event requirements, and event business logic. Semantic parameters are parameters that are semantically related to the key event parameters.

[0053] When annotating the key parameters corresponding to a tracking event, it is necessary to parse all event parameters of the tracking event to determine whether it carries the key tracking parameters. Simultaneously, it is necessary to detect whether it carries semantically related parameters that are semantically related to the key tracking parameters.

[0054] If the event parameters carry key tracking parameters and / or semantically related parameters, then proceed to step S2022; otherwise, proceed to step S2023.

[0055] Step S2022: If the event parameters carry key tracking parameters and / or semantically related parameters, then the pre-trained annotation model is used to annotate the key tracking parameters or semantically related parameters.

[0056] When the event parameters carry key tracking parameters, the key tracking parameters in the event parameters are labeled by calling a pre-trained annotation model, thereby achieving the labeling processing of key tracking parameters.

[0057] When the event parameters do not carry key event tracking parameters, but semantically related parameters are detected, the semantically related parameters in the event parameters are labeled by calling a pre-trained annotation model, thereby achieving the labeling processing of semantically related parameters.

[0058] When the event parameters are detected to carry key tracking parameters and semantically related parameters, the key tracking parameters are prioritized. The key tracking parameters in the event parameters are labeled by calling a pre-trained annotation model, thereby achieving the labeling processing of key tracking parameters.

[0059] Step S2023: If the event parameters do not carry key tracking parameters and semantically related parameters, then the event parameters are vectorized to obtain multiple first parameter vectors corresponding to the event parameters.

[0060] If the event parameters do not include key tracking parameters and semantically related parameters, annotation will fail when calling the annotation model. For tracking events that fail to be annotated, their corresponding event parameters are vectorized to obtain the first parameter vector for each event parameter.

[0061] Step S2024: Determine the second parameter vector that matches the key parameters of the embedding point from multiple first parameter vectors.

[0062] Each key parameter of the event tracking point has a corresponding vectorized representation, which is stored in a vectorized database. When the event parameters do not include key parameters of the event tracking point or semantically related parameters, the annotation model is used to calculate the similarity between multiple first parameter vectors corresponding to the event parameters and key parameter vectors of the event tracking point in the vectorized database. This allows the selection of the parameter vector that is most similar to the key parameter vector of the event tracking point from the multiple first parameter vectors, which is the second parameter vector that matches the key parameter of the event tracking point.

[0063] Step S2025: The embedding parameters corresponding to the second parameter vector are used as key embedding parameters and labeled.

[0064] When the second parameter vector that matches the key parameters of the tracking point is retrieved, the target event parameter corresponding to the second parameter vector is determined, and the target event parameter is used as the key parameter of the tracking point event. The key parameter of the tracking point is then annotated using the annotation model to obtain the corresponding annotation information.

[0065] Step S2026: Add the key parameters of the marked tracking points to the structured data to obtain structured marked data. For details, please refer to the relevant descriptions of the corresponding steps in the above embodiments; they will not be repeated here.

[0066] Step S203 involves extracting the tracking entities corresponding to the tracking events, the relationships between tracking entities, and the attribute information corresponding to the tracking entities from the structured annotation data. For details, please refer to the relevant descriptions of the steps in the embodiments shown above; they will not be repeated here.

[0067] Step S204: Construct a knowledge graph of the tracking points based on the tracking entities, relationships, and attribute information. For details, please refer to the relevant descriptions of the corresponding steps in the above embodiments; they will not be repeated here.

[0068] The method for constructing a knowledge graph for event tracking provided in this embodiment can directly use a pre-trained annotation model for annotation when the event parameters are detected to carry key event tracking parameters and / or semantically related parameters. When the event parameters are detected to not carry key event tracking parameters and semantically related parameters, vector matching is used to determine the parameter vector that matches the key event tracking parameters, and the annotation model is used to annotate the event parameters corresponding to the parameter vector. This achieves the annotation processing of key event tracking parameters or semantically related parameters, making it easier to determine key event tracking information in the event tracking through annotation and reducing the reasoning cost of event tracking entities.

[0069] This embodiment provides a method for constructing a knowledge graph based on data tracking points, which can be used in computer devices such as computers and servers. Figure 4 This is a flowchart of a method for constructing a knowledge graph based on embodiments of this disclosure, such as... Figure 4 As shown, the process includes the following steps:

[0070] Step S301: Obtain the structured data corresponding to the tracking point and the tracking event corresponding to the tracking point. For details, please refer to the relevant descriptions of the corresponding steps in the above embodiments, which will not be repeated here.

[0071] Step S302: Annotate the key parameters of the tracking points corresponding to the tracking events, and add the annotation information of the key parameters to the structured data to obtain structured annotation data. For details, please refer to the relevant descriptions of the corresponding steps in the above embodiments, which will not be repeated here.

[0072] Step S303: Extract the tracking entities corresponding to the tracking events, the relationships between tracking entities, and the attribute information corresponding to the tracking entities from the structured annotation data. For details, please refer to the relevant descriptions of the corresponding steps in the above embodiments; they will not be repeated here.

[0073] Step S304: Construct a knowledge graph of the tracking points according to the tracking point entities, relationships, and attribute information.

[0074] Specifically, step S304 includes:

[0075] Step S3041: Treat each embedded entity as an entity node, and associate the attribute information corresponding to each embedded entity with the entity node.

[0076] Entity nodes are the graph nodes used to construct the tracking knowledge graph. Specifically, an entity data table is generated for each identified tracking entity and stored in the graph database. During the construction of the tracking knowledge graph, corresponding entity nodes are created sequentially according to the entity data table. Since each tracking entity has related attribute information, a data structure is created to associate the attribute information with the corresponding entity node for subsequent queries.

[0077] Step S3042: Generate relationship edges between the embedded entities according to the association between them.

[0078] Relationships are used to characterize the relationships between tracked entities, such as dependency, monitoring, inclusion, and requirement relationships. Based on the relationships extracted from structured annotation data, relationship edges are sequentially added between different tracked entities according to these relationships. This generates relationship edges pointing from one tracked entity to another, and these edges are stored in a graph database. Simultaneously, corresponding relationship description information is generated for each edge, such as... Figure 5 As shown.

[0079] Step S3043: Connect each entity node based on the relationship edges to generate a knowledge graph of embedded points.

[0080] Each entity node is traversed sequentially. It is determined whether a corresponding relation edge exists for the currently traversed entity node. If no corresponding relation edge exists, the traversal continues to the next entity node. If a corresponding relation edge exists, other entity nodes corresponding to that relation edge are identified, and the currently traversed entity node is connected to other entity nodes according to the relation edges. This yields an abstract graph constructed from the entity nodes and relation edges, such as... Figure 6 As shown. Then, based on this abstract graph, the corresponding tracking entities for each entity node are determined, and a tracking knowledge graph composed of different tracking entities is generated.

[0081] In some optional implementations, the above method further includes: analyzing the dependencies between the relation edges in the tracking knowledge graph, and determining the hidden entity nodes in the tracking knowledge graph based on the dependencies.

[0082] The dependencies represented by the edges in the knowledge graph are analyzed to determine whether there is a conjugate dependency between the entity nodes. That is, if an entity node (A) is associated with another entity node (B), and entity node B is associated with a third entity node (C), then there may be some kind of relationship between entity nodes A and C. In the knowledge graph, the relationship between these three entity nodes can be represented by a triangle (ABC), where B is the conjugate dependency edge.

[0083] Specifically, to discover conjugate dependency edges in the knowledge graph, all entity nodes and all relation edges in the knowledge graph are traversed to determine whether conjugate dependencies can be formed between the relation edges. If a conjugate dependency is found, the hidden entity node that should exist can be inferred from the conjugate dependency edge, and the hidden entity node is used to fill in the knowledge graph.

[0084] When a new requirement arises, it can be detected whether there are similar characteristics between the new requirement's tracking entity and existing tracking entities. If there are similar characteristics between the new requirement's tracking entity and existing tracking entities, it indicates that there is a lineage relationship between the new requirement's tracking entity and existing tracking entities. At this time, a conjugate dependency edge can be constructed based on this lineage relationship, and the tracking entity of the new requirement can be updated to the tracking knowledge graph according to the conjugate dependency edge.

[0085] In the above implementation, by analyzing the dependencies between each relation edge, the hidden entity nodes in the tracking knowledge graph are determined. The tracking knowledge graph is then updated using these hidden entity nodes, ensuring the completeness of the nodes in the tracking knowledge graph, improving the accuracy of the tracking knowledge graph construction, facilitating the identification of potential tracking links, ensuring accurate associations between each tracking entity, and improving the efficiency and accuracy of subsequent tracking testing.

[0086] In some optional implementations, prior to step S304, the method further includes:

[0087] Step a1: Obtain the scene description information corresponding to the embedded entity.

[0088] Step a2: Align the embedded entities according to the scene description information and attribute information to obtain the target aligned entity.

[0089] Step a3: Perform entity fusion based on the relevant attribute information and relationships of the target aligned entities.

[0090] Scenario description information describes the business scenario in which the tracked entity exists, such as a live-streaming product scenario or a webpage browsing scenario. Specifically, the scenario description information is generated for each business scenario and is pre-stored in a graph database associated with the tracked entity. Before constructing the tracked entity knowledge graph, the scenario description information corresponding to each tracked entity is extracted from the graph database.

[0091] Due to the diversity of data sources, different people have different focuses and recording perspectives on the tracked entities. Therefore, different scene description information may exist for the same tracked entity. Hence, it is necessary to perform semantic similarity detection on the scene description information corresponding to each tracked entity to determine the scene description similarity of each tracked entity, and then unify the scene description information corresponding to the tracked entity according to the scene description similarity.

[0092] Subsequently, entity alignment is performed on each tracking entity based on the similarity of scene descriptions and attribute information to ensure the uniqueness of tracking entities used to construct the tracking knowledge graph, generating corresponding target aligned entities. Simultaneously, entity fusion is performed on the tracking entities in the tracking knowledge graph using the relevant attribute information and relationships of the target aligned entities to merge and remove duplicate tracking entities, further ensuring the uniqueness of each tracking entity and improving the representation quality of tracking entities in the tracking knowledge graph.

[0093] In some optional implementations, after step S304, the method further includes:

[0094] Step b1: When new data points are generated, determine the new data point entity corresponding to the new data point.

[0095] Step b2: If the newly added tracking entity is in the tracking knowledge graph, then the newly added tracking entity will be merged with the existing tracking entity corresponding to the newly added tracking entity.

[0096] Step b3: If the newly added entity is not in the knowledge graph, add the new entity node corresponding to the newly added entity in the knowledge graph, as well as the relationship edge between the new entity node and the existing entity node.

[0097] Real-time monitoring of event tracking data is performed to determine if new event tracking data has been generated. When new event tracking data is detected, a corresponding new event tracking entity is constructed for that data and stored in the entity data table of the graph database. The new event tracking entity is compared with existing event tracking entities in the entity data table. If the new event tracking entity exists in the entity data table, it means that the new event tracking entity is in the event tracking knowledge graph. At this time, the new event tracking entity is merged with its corresponding existing event tracking entity to avoid duplication of event tracking entities, reduce redundant event tracking entities in the event tracking knowledge graph, save the cost of building and storing the event tracking knowledge graph, and reduce the cost of repeated testing of event tracking.

[0098] If the newly added event tracking entity does not exist in the entity data table, it means that the new event tracking entity is not in the event tracking knowledge graph. In this case, it is necessary to create a corresponding new entity node for the new event tracking entity, determine the association between the new entity node and existing entity nodes, and generate relationship edges between the new entity node and existing entity nodes according to the association. Subsequently, the new entity node and the relationship edges between the new entity node and existing entity nodes are added to the event tracking knowledge graph to update the event tracking knowledge graph. This ensures that the event tracking knowledge graph can be updated in real time, ensuring the correctness of the dependencies between its event tracking entities. In this way, it is possible to identify the core event tracking points with significant impact in the event tracking knowledge graph by combining the dependencies.

[0099] In some optional implementations, the above method further includes:

[0100] Step c1: Obtain the number of key relationship edges of different types in the knowledge graph.

[0101] Step c2: Determine the importance of the tracking points based on the number of key relationship edges of different types and the first preset weight corresponding to each type.

[0102] Step c3: Classify the tracking points in the tracking point knowledge graph according to their importance and generate tracking point classification results.

[0103] As described above, relationship edges represent all the relationships between single-item entities bound together in the event tracking knowledge graph, such as business dashboards, requirement references, quality monitoring, and regression test cases. Each binding relationship is represented as an edge. Key relationship edges are used to measure the importance of event tracking to the business, specifically including business requirements and dashboard relationship edges bound to each single event tracking / parameter, such as product dashboard relationship edges, data monitoring dashboard relationship edges, and source database dashboard relationship edges. The number of key relationship edges of different types in the event tracking knowledge graph is counted to determine the number of key relationship edges for each type.

[0104] The first preset weight is a pre-defined weight used to characterize the importance of the business. The number of key relationship edges of different types and their corresponding first preset weights are weighted and summed to obtain a weighted value, which is then used to evaluate the importance of the tracking point. The larger the weighted value, the higher the importance of the tracking point.

[0105] The tracking points in the tracking knowledge graph are classified into core tracking points and non-core tracking points based on their importance, resulting in corresponding tracking point classification results. Specifically, the weighted value used to represent the importance of the tracking points is compared with a preset value (a pre-defined tracking point classification value, such as 40, 50, etc.). When the weighted value is greater than the preset value, it indicates that the tracking entity corresponding to the entity node is a core tracking entity; when the weighted value is less than the preset value, it indicates that the tracking entity corresponding to the entity node is a non-core tracking entity.

[0106] The knowledge graph of tracking points can also be used to export the score of the importance of tracking points with one click, saving the time of manually reviewing core tracking point entities. At the same time, the knowledge graph of tracking points can be used to identify non-core tracking point entities and simplify tracking points in regression, saving manual regression time or regression automation time.

[0107] In the above implementation, the importance of the tracking points is calculated, and the core tracking points and non-core tracking points are classified according to the importance of the tracking points. This makes it easier to identify the core tracking points based on the classification results, improves the identification efficiency of the core tracking points, and helps to improve the efficiency of tracking point testing.

[0108] In some optional implementations, the above method further includes:

[0109] Step d1: Obtain the number of quality assurance edges for different types of data points in the knowledge graph.

[0110] Step d2: Determine the quality assurance level of the data points based on the number of quality assurance edges for different types and the second preset weights corresponding to different types.

[0111] Step d3: Determine the prevention and control level of the embedded points based on their importance and quality assurance.

[0112] Quality assurance edges are relational edges used to measure the quality assurance level of event tracking. Specifically, they include relational edges representing the quality assurance level bound to each individual event tracking point / parameter, such as regression, monitoring, automation, and event tracking design review. The number of quality assurance edges of different types in the event tracking knowledge graph is counted to determine the number of quality assurance edges for each type.

[0113] The second preset weight is a pre-defined weight used to characterize the quality assurance level. Different types of quality assurance edge counts and their corresponding second preset weights are weighted and summed to obtain a corresponding quality assurance weighted value. This quality assurance weighted value is then used to evaluate the quality assurance level of the tracking points. The larger the quality assurance weighted value, the higher the quality assurance level of the tracking points.

[0114] By combining the quality assurance level and importance of the embedded points, the embedded points in the knowledge graph are evaluated to obtain the corresponding embedded point evaluation results, and the embedded point prevention and control level corresponding to the evaluation results is determined. Specifically, the quality assurance weighted value used to characterize the quality assurance level of the event tracking is compared with the preset quality assurance value (a pre-set value used for quality assurance classification, such as 10, 20, etc.). When the quality assurance weighted value is greater than the preset value, it indicates that the quality assurance measures for the event tracking entity are sufficient. Specifically, quality assurance can be distinguished into pre-event, during-event, and post-event stages. The quality assurance weighted value is the sum of the weighted values ​​of the three stages. Pre-event quality assurance measures include event tracking design review and event tracking self-testing, during-event quality assurance measures include manual requirement / regression testing, automated testing, and code release, and post-event quality assurance measures include manual release, diff monitoring, reporting delay monitoring, and rule monitoring. When the quality assurance weighted value is less than the preset value, it indicates that the quality assurance measures for the event tracking entity are average, and when the quality assurance weighted value is 0, it indicates that the quality assurance measures for the event tracking entity are lacking.

[0115] By combining the quality assurance level and importance of the event tracking points, the number of core event tracking points that need to be maintained in the current event tracking knowledge graph is determined. When the scores for the importance level of the event tracking points calculated by the event tracking knowledge graph are all greater than the preset value, and the scores for the quality assurance level are all greater than the preset value, it indicates that the current business has sufficient quality assurance measures for the existing core event tracking points (regression coverage xx%, diff monitoring coverage xx%), and no new quality assurance measures are needed for these existing core event tracking points for the time being.

[0116] If, among all the event entities / parameters corresponding to the current event knowledge graph, N event entities / parameters have average or missing quality assurance measures, and X event entities / parameters have event importance scores greater than the preset value, it indicates that the current business has insufficient quality assurance measures for Y core event points. Quality assurance measures such as monitoring, automation, and regression should be added. It is recommended to prioritize automation, monitoring, and other in-process / post-process measures to improve the event quality assurance score, thereby reducing risk control while maximizing manpower savings.

[0117] In the above implementation, the quality assurance level of the data points is calculated, and the data point control level is determined by combining the importance of the data points with the quality assurance level. This facilitates the optimization of data point design using the data point control level, improves the accuracy of data point data, and supports business decision-making.

[0118] In some optional implementations, the above method further includes:

[0119] Step e1: Obtain the knowledge retrieval request initiated for the knowledge graph of the tracking points.

[0120] Step e2: Search the tracking knowledge graph according to the knowledge retrieval request to obtain the tracking knowledge that matches the knowledge retrieval request.

[0121] A knowledge retrieval request is a user-initiated request to retrieve event tracking data for the current business context, such as "Please check if there are any event tracking data displayed on the image and text detail page." When a user initiates a knowledge retrieval request, the computer device can obtain the request and retrieve event tracking entities from the event tracking knowledge graph according to the request. By leveraging the dependencies in the event tracking knowledge graph, the device can accurately retrieve event tracking knowledge information related to the business scenario in question, thus improving the retrieval efficiency and accuracy of event tracking knowledge information.

[0122] The method for constructing a tracking knowledge graph provided in this embodiment uses each tracking entity as an entity node, associates corresponding attribute information in each entity node, and generates corresponding relationship edges by utilizing the relationship between tracking entities. The relationship edges are used to connect each tracking entity to form a tracking knowledge graph, realizing accurate dependencies between tracking entities. This makes it easy to sort out the tracking knowledge associations based on the dependency relationship edges. At the same time, it can identify the call chain of each tracking entity in the tracking knowledge graph and its impact on the entire business scenario, determine its priority, and improve the efficiency of tracking testing.

[0123] This embodiment also provides a device for constructing a knowledge graph based on data points. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0124] This embodiment provides a device for constructing a knowledge graph based on data tracking, such as... Figure 7 As shown, it includes:

[0125] The acquisition module 401 is used to acquire the structured data corresponding to the tracking point and the tracking event corresponding to the tracking point.

[0126] The annotation module 402 is used to annotate the key parameters of the tracking point corresponding to the tracking point event, and add the annotation information of the key parameters of the tracking point to the structured data to obtain structured annotation data.

[0127] The extraction module 403 is used to extract the tracking entities corresponding to the tracking events, the relationships between tracking entities, and the attribute information corresponding to the tracking entities from the structured annotation data.

[0128] The graph construction module 404 is used to construct a knowledge graph of tracking points according to the tracking point entities, relationships, and attribute information.

[0129] In some alternative implementations, the annotation module 402 includes:

[0130] The parameter detection unit is used to detect whether the event parameters corresponding to the tracking event carry tracking key parameters, and / or semantically related parameters of the tracking key parameters.

[0131] The first annotation unit is used to annotate the key tracking parameters or semantically related parameters using a pre-trained annotation model if the event parameters carry key tracking parameters and / or semantically related parameters.

[0132] In some alternative implementations, the annotation module 402 further includes:

[0133] The vectorization unit is used to vectorize the event parameters if the event parameters do not carry key tracking parameters and semantically related parameters, so as to obtain multiple first parameter vectors corresponding to the event parameters.

[0134] The vector matching unit is used to determine the second parameter vector that matches the key parameters of the embedded point from multiple first parameter vectors.

[0135] The second annotation unit is used to annotate the target event parameters corresponding to the second parameter vector as key parameters for the embedding point.

[0136] In some alternative implementations, the map construction module 404 includes:

[0137] The node determination unit is used to treat each embedded entity as an entity node and associate the attribute information corresponding to each embedded entity with the entity node.

[0138] The edge generation unit is used to generate relationship edges between embedded entities according to the association relationship between them.

[0139] The connection unit is used to connect various entity nodes based on relation edges to generate a knowledge graph of embedded points.

[0140] In some alternative embodiments, the above-described apparatus further includes:

[0141] The dependency analysis module is used to analyze the dependency relationships between the edges of each relationship in the knowledge graph and to determine the hidden entity nodes in the knowledge graph based on the dependency relationships.

[0142] In some alternative embodiments, the above-described apparatus further includes:

[0143] The scene description module is used to obtain scene description information corresponding to the embedded entities.

[0144] The entity alignment module is used to align the embedded entities according to the scene description information and attribute information to obtain the target aligned entity.

[0145] The entity fusion module is used to perform entity fusion based on the relevant attribute information and relationships of the target aligned entities.

[0146] In some alternative embodiments, the above-described apparatus further includes:

[0147] A new tracking entity generation module has been added to determine the new tracking entity corresponding to the new tracking data when new tracking data is generated.

[0148] The fusion module is used to fuse the newly added tracking entity with the existing tracking entity corresponding to the newly added tracking entity if the new tracking entity is in the tracking knowledge graph.

[0149] Add a module to add a new entity node corresponding to the newly added entity and the relationship edges between the new entity node and existing entity nodes in the event tracking knowledge graph if the newly added entity is not in the event tracking knowledge graph.

[0150] In some alternative embodiments, the above-described apparatus further includes:

[0151] The first edge count acquisition module is used to obtain the edge count of different types of key relationships in the knowledge graph.

[0152] The importance determination module is used to determine the importance of the tracking points based on the number of key relationship edges of different types and the first preset weight corresponding to each type.

[0153] The classification module is used to classify the tracking points in the tracking point knowledge graph according to their importance and generate tracking point classification results.

[0154] In some alternative embodiments, the above-described apparatus further includes:

[0155] The second edge count acquisition module is used to obtain the quality assurance edge count of different types in the knowledge graph of the tracking points.

[0156] The quality assurance level determination module is used to determine the quality assurance level of the data points based on the number of quality assurance edges of different types and the second preset weight corresponding to different types.

[0157] The prevention and control level determination module is used to determine the prevention and control level of the embedded points based on their importance and quality assurance.

[0158] In some alternative embodiments, the above-described apparatus further includes:

[0159] The retrieval module is used to retrieve knowledge retrieval requests initiated against the knowledge graph of tracking points.

[0160] The retrieval module is used to retrieve the tracking knowledge graph according to the knowledge retrieval request and obtain the tracking knowledge that matches the knowledge retrieval request.

[0161] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0162] In this embodiment, the device for constructing the knowledge graph of embedded points is presented in the form of functional units. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0163] The device for constructing a tracking knowledge graph provided in this embodiment acquires structured data corresponding to tracking points and annotates key parameters of tracking events. The annotation information of these key parameters is added to the structured data to obtain structured annotated data. This allows for multi-dimensional tagging of tracking points based on different business needs. Relying on the structured data of the tracking points themselves and the annotation method of key parameters, it is easy to cluster similar tracking entities based on semantics and business knowledge. This facilitates accurate perception or link analysis of reused tracking entities or parameters, significantly reducing the inference cost of tracking links. By extracting tracking entities corresponding to tracking events, the relationships between tracking entities, and the attribute information corresponding to tracking entities from the structured annotated data, a tracking knowledge graph is constructed according to the tracking entities, relationships, and attribute information. This clarifies the dependencies of each tracking point, accurately captures the contextual dependencies between tracking entities, avoids erroneous knowledge associations, improves the accuracy of tracking knowledge graph construction, and saves on the construction and storage costs of large-scale tracking knowledge graphs.

[0164] This disclosure also provides a computer device having the above-described features. Figure 7 The device shown is for constructing a knowledge graph based on embedded points.

[0165] Please see Figure 8 , Figure 8 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of this disclosure, such as... Figure 8As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 8 Take a processor 10 as an example.

[0166] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0167] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0168] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0169] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0170] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0171] This disclosure also provides a computer-readable storage medium in which the methods described in this disclosure can be implemented in hardware or firmware, or implemented as recordable on a storage medium, or implemented as computer code originally stored on a remote storage medium or a non-transitory machine-readable storage medium and subsequently stored on a local storage medium after being downloaded over a network. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium may be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium may also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code that, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0172] A portion of this disclosure can be applied to computer program products, such as computer program instructions, which, when executed by a computer, can invoke or provide methods and / or technical solutions according to this disclosure through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, and installation package files. Accordingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions; the computer compiling the instructions and then executing the corresponding compiled program; the computer reading and executing the instructions; or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0173] Although embodiments of the present disclosure have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present disclosure, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for constructing a knowledge graph based on data tracking, characterized in that, The method includes: Obtain the structured data corresponding to the tracking point and the tracking event corresponding to the tracking point; The key parameters corresponding to the event are labeled, and the labeling information of the key parameters is added to the structured data to obtain structured labeled data. Extract the tracking entities corresponding to the tracking events, the relationships between the tracking entities, and the attribute information corresponding to the tracking entities from the structured annotation data; Construct a knowledge graph of tracking points based on the tracking entities, the relationships, and the attribute information.

2. The method according to claim 1, characterized in that, The annotation of the key parameters corresponding to the event includes: Detect whether the event parameters corresponding to the event point carry the key parameters of the event point, and / or the semantically related parameters of the key parameters of the event point; If the event parameters carry the key tracking parameters and / or the semantically related parameters, then a pre-trained annotation model is used to annotate the key tracking parameters or the semantically related parameters.

3. The method according to claim 2, characterized in that, Also includes: If the event parameters do not carry the key parameters of the tracking point and the semantic related parameters, then the event parameters are vectorized to obtain multiple first parameter vectors corresponding to the event parameters; A second parameter vector matching the key parameters of the embedding point is determined from multiple first parameter vectors; The target event parameters corresponding to the second parameter vector are used as the key parameters of the embedding point and labeled.

4. The method according to claim 1, characterized in that, The step of constructing a knowledge graph based on the event tracking entities, the relationships, and the attribute information includes: Each of the aforementioned tracking points is treated as an entity node, and the attribute information corresponding to each of the aforementioned tracking points is associated with the entity node; Based on the association relationships between the embedded entities, relationship edges are generated between the embedded entities; The entity nodes are connected based on the relationship edges to generate the embedded knowledge graph.

5. The method according to claim 4, characterized in that, Also includes: Analyze the dependencies between the edges of each relationship in the knowledge graph, and determine the hidden entity nodes in the knowledge graph based on the dependencies.

6. The method according to claim 1, characterized in that, Before constructing the tracking knowledge graph according to the tracking entities, the relationships, and the attribute information, the following steps are also included: Obtain the scene description information corresponding to the embedded entity; The embedded entity is aligned according to the scene description information and the attribute information to obtain the target aligned entity; Entity fusion is performed based on the relevant attribute information and relationships of the target aligned entities.

7. The method according to claim 1 or 6, characterized in that, After constructing the tracking knowledge graph according to the tracking entities, the relationships, and the attribute information, the method further includes: When new data points are generated, the new data point entity corresponding to the new data point is determined. If the newly added tracking entity is in the tracking knowledge graph, then the newly added tracking entity and the existing tracking entity corresponding to the newly added tracking entity will be merged. If the newly added tracking entity is not in the tracking knowledge graph, then add a new entity node corresponding to the newly added tracking entity, as well as the relationship edge between the new entity node and the existing entity node, to the tracking knowledge graph.

8. The method according to claim 1, characterized in that, Also includes: Obtain the number of key relationship edges of different types in the aforementioned knowledge graph; The importance of the tracking points is determined based on the number of edges of the key relationships of different types and the first preset weight corresponding to each type. The tracking points in the tracking point knowledge graph are classified according to their importance, and tracking point classification results are generated.

9. The method according to claim 8, characterized in that, Also includes: Obtain the quality assurance edge count of different types in the aforementioned knowledge graph; The quality assurance level of the embedded points is determined based on the number of quality assurance edges of different types and the second preset weight corresponding to different types. The level of security and control of the embedded points is determined based on their importance and quality assurance.

10. The method according to claim 1, characterized in that, Also includes: Obtain the knowledge retrieval request initiated for the aforementioned knowledge graph; The knowledge graph of tracking points is retrieved according to the knowledge retrieval request to obtain tracking point knowledge that matches the knowledge retrieval request.

11. A device for constructing a knowledge graph based on embedded data points, characterized in that, The device includes: The acquisition module is used to acquire the structured data corresponding to the tracking point and the tracking event corresponding to the tracking point; The annotation module is used to annotate the key parameters of the tracking point corresponding to the tracking point event, and add the annotation information of the key parameters of the tracking point to the structured data to obtain structured annotated data; The extraction module is used to extract the tracking entity corresponding to the tracking event, the association relationship between the tracking entities, and the attribute information corresponding to the tracking entity from the structured annotation data. The knowledge graph construction module is used to construct a knowledge graph of the tracking points according to the tracking point entities, the relationships, and the attribute information.

12. A computer device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the method for constructing a knowledge graph based on any one of claims 1 to 10.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the method for constructing the knowledge graph of the tracking points as described in any one of claims 1 to 10.

14. A computer program product, characterized in that, It includes computer instructions for causing a computer to execute the method for constructing a knowledge graph based on any one of claims 1 to 10.

Images