Dynamic priority object reaching method and device, equipment, medium and program product

Abstract

The invention relates to a dynamic priority object reaching method and device, equipment, a medium and a program product. The method comprises the following steps: acquiring an initial path priority strategy about a target newly-added object; based on the strategy correction reference data and the to-be-corrected path priority strategy, a current corrected path priority strategy obtained through current iteration is obtained through a strategy correction model, the to-be-corrected path priority strategy obtained through first iteration is an initial path priority strategy, and the to-be-corrected path priority strategy obtained through second iteration is an initial path priority strategy; the to-be-corrected path priority strategy not iterated for the first time is a previous corrected path priority strategy obtained by previous iteration; taking the current corrected path priority strategy as a to-be-corrected path priority strategy of the next iteration, returning to the step of determining the strategy correction reference data of the to-be-corrected path priority strategy, and continuing to execute until an iteration stop condition is met; and the target reaching path is determined based on the last-time path correction priority strategy of the last-time iteration, so that the accuracy of reaching path determination is improved.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method, apparatus, device, medium, and program product for reaching dynamic priority objects. Background Technology

[0002] With the development of computer technology, businesses are increasingly relying on data-driven methods to identify and reach potential customers when expanding their customer base.

[0003] In related technologies, when acquiring customers, it is common practice to use enterprise-related knowledge graphs to determine object description information and identify relationship chains between objects in order to determine the reach path, thereby reaching potential customers and then promoting relevant content to them.

[0004] However, the determination of reach paths in related technologies relies solely on shallow analysis of knowledge graphs, which results in a one-sided and simplistic approach to reach path analysis, leading to low accuracy in reach path determination. Summary of the Invention

[0005] Therefore, it is necessary to provide a dynamic priority object outreach method, apparatus, device, medium, and program product that can improve the accuracy of outreach path determination in response to the above-mentioned technical problems.

[0006] Firstly, this application provides a dynamic priority object access method, including:

[0007] Obtain an initial path priority strategy for the target new object to be reached. The initial path priority strategy indicates the priority order of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding reached objects to the target new object.

[0008] For the current iteration, obtain the path priority strategy to be corrected in the current iteration. The path priority strategy to be corrected corresponding to the first iteration is the initial path priority strategy, and the path priority strategy to be corrected corresponding to non-first iterations is the previous corrected path priority strategy obtained in the previous iteration.

[0009] Determine the policy correction reference data for the path priority strategy to be corrected. Based on the policy correction reference data and the path priority strategy to be corrected, obtain the current corrected path priority strategy obtained in the current iteration through the policy correction model. Use the current corrected path priority strategy as the path priority strategy to be corrected in the next iteration. Return to the step of determining the policy correction reference data for the path priority strategy to be corrected and continue execution until the iteration stops when the iteration stop condition is met.

[0010] Based on the final correction path priority strategy corresponding to the last iteration, the target reach path for the newly added target object is determined from multiple candidate reach paths.

[0011] Secondly, this application also provides a dynamic priority object reach device, comprising:

[0012] The first strategy acquisition module is used to acquire an initial path priority strategy for the target new object to be reached. The initial path priority strategy indicates the priority order of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding reached objects to the target new object.

[0013] The second strategy acquisition module is used to acquire the path priority strategy to be corrected in the current iteration for the current iteration. The path priority strategy to be corrected corresponding to the first iteration is the initial path priority strategy, and the path priority strategy to be corrected corresponding to non-first iterations is the previous corrected path priority strategy obtained in the previous iteration.

[0014] The strategy correction module is used to determine the strategy correction reference data of the path priority strategy to be corrected, and based on the strategy correction reference data and the path priority strategy to be corrected, obtain the current corrected path priority strategy obtained in the current iteration through the strategy correction model, use the current corrected path priority strategy as the path priority strategy to be corrected in the next iteration, return to the step of determining the strategy correction reference data of the path priority strategy to be corrected and continue to execute until the iteration stops when the iteration stop condition is met.

[0015] The reach path determination module is used to determine the target reach path of the newly added target object from multiple candidate reach paths based on the final modified path priority strategy corresponding to the last iteration.

[0016] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0017] Obtain an initial path priority strategy for the target new object to be reached. The initial path priority strategy indicates the priority order of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding reached objects to the target new object.

[0018] For the current iteration, obtain the path priority strategy to be corrected in the current iteration. The path priority strategy to be corrected corresponding to the first iteration is the initial path priority strategy, and the path priority strategy to be corrected corresponding to non-first iterations is the previous corrected path priority strategy obtained in the previous iteration.

[0019] Determine the policy correction reference data for the path priority strategy to be corrected. Based on the policy correction reference data and the path priority strategy to be corrected, obtain the current corrected path priority strategy obtained in the current iteration through the policy correction model. Use the current corrected path priority strategy as the path priority strategy to be corrected in the next iteration. Return to the step of determining the policy correction reference data for the path priority strategy to be corrected and continue execution until the iteration stops when the iteration stop condition is met.

[0020] Based on the final correction path priority strategy corresponding to the last iteration, the target reach path for the newly added target object is determined from multiple candidate reach paths.

[0021] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0022] Obtain an initial path priority strategy for the target new object to be reached. The initial path priority strategy indicates the priority order of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding reached objects to the target new object.

[0023] For the current iteration, obtain the path priority strategy to be corrected in the current iteration. The path priority strategy to be corrected corresponding to the first iteration is the initial path priority strategy, and the path priority strategy to be corrected corresponding to non-first iterations is the previous corrected path priority strategy obtained in the previous iteration.

[0024] Determine the policy correction reference data for the path priority strategy to be corrected. Based on the policy correction reference data and the path priority strategy to be corrected, obtain the current corrected path priority strategy obtained in the current iteration through the policy correction model. Use the current corrected path priority strategy as the path priority strategy to be corrected in the next iteration. Return to the step of determining the policy correction reference data for the path priority strategy to be corrected and continue execution until the iteration stops when the iteration stop condition is met.

[0025] Based on the final correction path priority strategy corresponding to the last iteration, the target reach path for the newly added target object is determined from multiple candidate reach paths.

[0026] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0027] Obtain an initial path priority strategy for the target new object to be reached. The initial path priority strategy indicates the priority order of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding reached objects to the target new object.

[0028] For the current iteration, obtain the path priority strategy to be corrected in the current iteration. The path priority strategy to be corrected corresponding to the first iteration is the initial path priority strategy, and the path priority strategy to be corrected corresponding to non-first iterations is the previous corrected path priority strategy obtained in the previous iteration.

[0029] Determine the policy correction reference data for the path priority strategy to be corrected. Based on the policy correction reference data and the path priority strategy to be corrected, obtain the current corrected path priority strategy obtained in the current iteration through the policy correction model. Use the current corrected path priority strategy as the path priority strategy to be corrected in the next iteration. Return to the step of determining the policy correction reference data for the path priority strategy to be corrected and continue execution until the iteration stops when the iteration stop condition is met.

[0030] Based on the final correction path priority strategy corresponding to the last iteration, the target reach path for the newly added target object is determined from multiple candidate reach paths.

[0031] The aforementioned dynamic priority object reach method, apparatus, device, medium, and program product obtains an initial path priority strategy for the target new object to be reached. This initial path priority strategy indicates the priority ranking of multiple candidate reach paths for the target new object, where each candidate reach path points from a corresponding already reached object to the target new object. For each iteration, the path priority strategy to be corrected in that iteration is obtained. The path priority strategy to be corrected in the first iteration is the initial path priority strategy, and the path priority strategy to be corrected in subsequent iterations is the previous corrected path priority strategy obtained in the previous iteration. Policy correction reference data for the path priority strategy to be corrected is determined. Based on the policy correction reference data and the path priority strategy to be corrected, the path priority strategy to be corrected is dynamically adjusted using a policy correction model to obtain the current corrected path priority strategy for that iteration. This current corrected path priority strategy is then used as the path priority strategy to be corrected in the next iteration. The process returns to the step of determining the policy correction reference data for the path priority strategy to be corrected and continues to execute, performing multiple iterative optimization corrections until the iteration stops when the iteration stopping condition is met, thereby continuously optimizing the priority ranking of each candidate reach path. In this way, based on the final correction path priority strategy corresponding to the last iteration, the target reach path that matches the target new object can be accurately selected from multiple candidate reach paths. In the above process, the priority of each candidate reach path is continuously adjusted through iteration, which can deeply analyze the suitable candidate reach paths that match the target new object, thereby improving the accuracy of reach path determination. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 This is an application environment diagram of the dynamic priority object access method in one embodiment;

[0034] Figure 2 This is a flowchart illustrating a dynamic priority object access method in one embodiment;

[0035] Figure 3 This is a structural block diagram of a dynamic priority object access device in one embodiment;

[0036] Figure 4 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0038] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0039] The dynamic priority object reach method provided in this application embodiment can be applied to, for example, Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104, or it can be located in the cloud or on another network server. The dynamic priority object access method provided in this application embodiment can be executed by terminal 102 or server 104 alone, or it can be executed collaboratively by terminal 102 and server 104.

[0040] In some embodiments, after the server 104 receives an object reach request sent by the terminal 102, the server 104 obtains an initial path priority policy for the target new object to be reached. The initial path priority policy indicates the priority order of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding reached objects to the target new object. For the current iteration, the server obtains the path priority policy to be corrected for the current iteration, wherein the path priority policy to be corrected for the first iteration is the initial path priority policy, and the path priority policy to be corrected for non-first iterations is the previous corrected path priority policy obtained in the previous iteration. The server determines the policy correction reference data for the path priority policy to be corrected. Based on the policy correction reference data and the path priority policy to be corrected, the server obtains the current corrected path priority policy obtained in the current iteration through the policy correction model. The server uses the current corrected path priority policy as the path priority policy to be corrected for the next iteration. The server returns to the step of determining the policy correction reference data for the path priority policy to be corrected and continues to execute until the iteration stops when the iteration stop condition is met. Based on the last corrected path priority policy corresponding to the last iteration, the server determines the target reach path for the target new object from multiple candidate reach paths.

[0041] The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, etc. The server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0042] In one exemplary embodiment, such as Figure 2 As shown, a dynamic priority object access method is provided, which can be applied to computer devices (which can...). Figure 1 Taking terminal 102 or server 104 as an example, the explanation includes the following steps 202 to 208. Wherein:

[0043] Step 202: Obtain the initial path priority strategy for the target new object to be reached. The initial path priority strategy indicates the priority order of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding reached objects to the target new object.

[0044] In this context, the target new objects to be reached are the customer acquisition goals of the current group, i.e., the objects to be expanded, or the potential customers of the current group. The current group is a group with a need to reach out to these objects; it could be a company, a school, etc., for example, a bank. Reached objects are those who have become customers of the current group. If the current group successfully sends its push content to an object, and that object successfully interacts with the product pushed out by the content, then that object can be considered a reached object. Candidate reach paths refer to paths that include target new objects. Further, a candidate reach path is a path from a corresponding reached object to a target new object. Multiple candidate reach paths are different paths, and each candidate reach path contains a path node for a target new object. For example, if a candidate reach path includes ordered path nodes, then the last path node of the candidate reach path is the target new object. The initial path priority strategy is the path priority strategy to be iterated. The path priority strategy reflects the priority ranking of each candidate reach path. Priority ranking is used to reflect the importance of the corresponding candidate reach paths. For example, the more important a candidate reach path is, the higher its priority ranking will be.

[0045] Optionally, if the current object reach conditions are met, the computer device acquires the collective knowledge graph belonging to the current group, and determines the initial path priority strategy for the target new object based on the collective knowledge graph. The collective knowledge graph is determined by integrating structured data such as enterprise qualifications, investment and financing attributes, and transaction records, as well as unstructured data such as industry trends and supply chain documents. For example, if the time between the current moment and the previous reach completion time is a preset duration, it is determined that the current object reach conditions are met. For example, based on the objects involved in the currently acquired collective knowledge graph, the computer device identifies unreached objects, treats all identified unreached objects as target new objects, and determines the initial path priority strategy for the target new objects based on the collective knowledge graph. Image recognition and NLP (Natural Language Processing) technologies can be used to process unstructured data such as industry news and supply chain documents.

[0046] In some embodiments, obtaining an initial path priority strategy for the target new object to be reached includes: obtaining a collective knowledge graph, which includes multiple reach paths, each reach path including multiple objects connected sequentially, and two objects with an association relationship having an edge between them; determining multiple candidate reach paths for the target new object to be reached based on the collective knowledge graph, each candidate reach path including the target new object; determining the priority ranking of each candidate reach path based on the path information corresponding to each candidate reach path; and determining an initial path priority strategy for the target new object to be reached based on the priority ranking of each candidate reach path.

[0047] The collective knowledge graph integrates structured and unstructured data related to a group, transforming entities and multidimensional relationships into a spatiotemporal graph structure that can be computed in real time. For example, this collective knowledge graph is a dynamically scalable, group-level knowledge graph. Structured data includes collective identity credentials (such as business registration information), collective qualifications (enterprise qualifications), investment attributes, transaction records, internal credit lines, etc.; unstructured data includes industry trends, supply chain documents, etc. Entities can be collective entities, such as enterprise entities (including legal entities, key decision-makers, and industry sectors), or single entities, such as individual entities. Multidimensional relationships can include supply chains, equity penetration levels, and cooperation closeness. The collective knowledge graph includes multiple access paths, each containing sequentially connected objects (which can be entities). Objects on an access path can be considered path nodes. Two objects connected by an edge in each access path have a relationship; for example, two connected objects may have a supply chain relationship. The path information for each candidate reach path includes the total number of nodes in the corresponding candidate reach path, the connection variable (total degree of nodes), and the path length (distance from the start path node to the end path node).

[0048] Optionally, the steps for constructing a collective knowledge graph include: acquiring data from various sources (such as databases, documents, web pages, etc.) through a data acquisition unit in a computer device; this data can be structured (such as tables, database records) or unstructured (such as text, images); extracting key information such as entities (such as companies, people), attributes (collective identity credentials, collective qualifications), and relationships (such as transactions, investments) from the collected data through a knowledge extraction unit in the computer device; integrating and unifying information from different sources (extracted information) through a knowledge fusion unit in the computer device, eliminating redundancy and conflicts, and ensuring the consistency and accuracy of knowledge; and constructing the collective knowledge graph in the form of a graph structure through a knowledge storage unit in the computer device, and storing it in a database.

[0049] Optionally, the computer device retrieves the latest collective knowledge graph from the database at the current moment. The retrieved collective knowledge graph includes multiple access paths, and each access path includes multiple objects connected in sequence. There are edges between two objects that have an association relationship.

[0050] Optionally, the computer device identifies unreached candidate objects from the collective knowledge graph, treating each identified candidate object as a target new object. Based on the position of the target new object in the collective knowledge graph, the computer device determines multiple candidate reach paths corresponding to the target new object. For example, based on the position of the target new object in the collective knowledge graph, the computer device determines multiple candidate reach paths containing the target new object, where the object at the end of each candidate reach path is the target new object. Based on the multiple reach paths containing the target new object, multiple candidate reach paths are selected, where at least one of the other objects preceding the target new object in each candidate reach path has already been reached.

[0051] Optionally, for each candidate reach path, the computer device determines the priority ranking of each candidate reach path based on the path information corresponding to that candidate reach path and the business rules of the current collective at the current moment. Based on the priority ranking of each candidate reach path, an initial path priority strategy is determined for the target new object to be reached. For example, for each candidate reach path, the computer device invokes a value quantification model. Based on the value quantification model, and according to the path information and business rules corresponding to that candidate reach path, a corresponding priority score is determined. The priority score reflects the business value of the candidate reach path; the higher the business value, the more important it is, and the higher the priority score. Based on the priority scores of each candidate reach path, the priority ranking of each candidate reach path is determined to determine the initial path priority strategy for the target new object to be reached. The value quantification model is used to evaluate entities and relationships in the knowledge graph, and can also evaluate the business value of reach paths, guiding resource allocation during the construction process and avoiding "average effort."

[0052] In the above embodiments, multiple candidate reach paths related to the target new object can be identified through the graph information of the collective knowledge graph. Based on the path information of each candidate reach path, the importance of each candidate reach path is initially identified, thereby obtaining an initial path priority strategy for the target new object.

[0053] In some embodiments, the priority ranking of each candidate reach path is determined based on the path information corresponding to each candidate reach path, including: for each candidate reach path, determining the scores corresponding to the candidate reach path in the structural dimension and the business dimension respectively based on the path information and business rule information corresponding to the candidate reach path, and merging the scores corresponding to the structural dimension and the business dimension respectively to obtain the priority score corresponding to the candidate reach path; and determining the priority ranking of each candidate reach path based on the priority scores corresponding to each candidate reach path.

[0054] The structural dimension reflects the structural dimensions of the candidate reach path. The business dimension reflects the business dimensions of the candidate reach path. Business rule information is used to indicate the business rules of the corresponding candidate reach path; for example, business rule information includes the number of transactions and shareholding ratio.

[0055] For example, for each candidate reach path, the computer device determines the centrality, page ranking, and path length of the candidate reach path based on its path information. Centrality is defined as node degree (number of connected edges) / total number of nodes. Page ranking is calculated as (1 - d) * (1 / N) + d * Σ [PR(Tj) / C(Tj)], where d is the damping factor (typically 0.85), N is the total number of nodes, Tj is the node pointing to Pi, and C(Tj) is the number of outgoing chains from node Tj. Path length is the shortest path distance from the starting node to the ending node. Based on the centrality, page ranking, path length, and business rule information of the candidate reach path, a value quantification model determines the scores corresponding to the structural dimension and the business dimension. The value quantification model evaluates the centrality, page ranking, and path length of the candidate reach path separately, obtaining corresponding first sub-scores. These first sub-scores are then merged to obtain the score for the structural dimension. The value quantification model scores the business parameters in the business rules information separately, obtaining corresponding second sub-scores. The scores of these second sub-scores are then combined to obtain the business dimension score. For example, based on the number of transactions and the shareholding ratio included in the business rules information, the number of transactions and the shareholding ratio are scored separately. Each indicator is normalized to [0,1], and the scores are weighted and summed to obtain the business dimension score.

[0056] For example, the computer device weights the scores of the candidate reach path according to the respective weights of the structural and business dimensions, respectively, to obtain the priority score of the candidate reach path.

[0057] For example, the computer device determines the priority ranking of each candidate reach path based on its corresponding priority score. The higher the priority score of a candidate reach path, the higher its priority ranking.

[0058] In the above embodiments, the candidate reach path can be adaptively scored in the structural and business dimensions based on the path information and business rule information corresponding to the candidate reach path. This results in a comprehensive score for the candidate reach path, i.e., a priority score, to initially identify the importance of each candidate reach path, thereby obtaining the initial path priority strategy for the target new object to be reached.

[0059] Step 204: For the current iteration, obtain the path priority strategy to be corrected for the current iteration, wherein the path priority strategy to be corrected corresponding to the first iteration is the initial path priority strategy, and the path priority strategy to be corrected corresponding to non-first iterations is the previous corrected path priority strategy obtained in the previous iteration.

[0060] The multiple iterations are used to refine the initial path priority strategy in order to identify the optimal one. The path priority strategy to be refined is the one that needs to be corrected in each iteration. The previously refined path priority strategy refers to the path priority strategy obtained after the strategy refinement in the previous iteration.

[0061] For example, for the current iteration, if the current iteration is the first iteration, the computer device uses the initial path priority policy as the path priority policy to be corrected in the current iteration. If the current iteration is not the first iteration, the computer device uses the previously corrected path priority policy obtained in the previous iteration as the path priority policy to be corrected in the current iteration.

[0062] Step 206: Determine the policy correction reference data for the path priority strategy to be corrected. Based on the policy correction reference data and the path priority strategy to be corrected, obtain the current corrected path priority strategy obtained in the current iteration through the policy correction model. Use the current corrected path priority strategy as the path priority strategy to be corrected in the next iteration. Return to the step of determining the policy correction reference data for the path priority strategy to be corrected and continue execution until the iteration stops when the iteration stop condition is met.

[0063] The strategy correction reference data serves as a reference for correction, assisting the strategy correction model in revising the priority strategy of the path to be corrected. For example, the strategy correction reference data includes at least historical business feedback data, which includes the feedback results of each candidate reach path. The feedback results indicate whether the visit content was successfully sent to the target new object after being sent according to the corresponding candidate reach path. It should be noted that sending the visit content to the target new object does not necessarily mean successful reach, and the target new object is not yet a reached object. In this case, it only indicates that the candidate reach path is valid and belongs to a valid candidate reach path. The visit content can be content without push notifications. Successful reach means that push notifications were sent to the target new object, and the target new object interacted with the push notifications, such as purchasing the product recommended in the push notifications. This is considered a successful reach, and the target new object is considered a reached object. Of course, the strategy correction reference data for each iteration can be the same or different. In some embodiments, when the current iteration i is neither the first nor the second iteration (i.e., i is greater than or equal to 3, for example, i=3), the policy correction reference data for each iteration also includes the corrected path priority policy obtained in the (i-2)th iteration. When i is greater than 3, the policy correction reference data for each iteration also includes the corrected path priority policy obtained in the (i-2)th iteration and the corrected path priority policies obtained in iterations prior to the (i-2)th iteration. The policy correction model is the model used to correct the path priority policy to be corrected; it can be based on a neural network model or a large language model.

[0064] Optionally, the computer device acquires policy correction reference data for the path priority strategy to be corrected, and based on the policy correction reference data and the path priority strategy to be corrected, obtains the current corrected path priority strategy for the current iteration through the policy correction model. For example, when the policy correction model is a large language model, based on the policy correction reference data and the path priority strategy to be corrected, the correction prompt text for the current iteration is determined, and the large language model is called to perform semantic understanding on the correction prompt text to determine the current corrected path priority strategy.

[0065] Optionally, the computer device determines whether the iteration stopping condition is met based on the current iteration count and a preset count. If the current iteration count has not reached the preset count, the iteration stopping condition is determined not to be met; if the current iteration count has reached the preset count, the iteration stopping condition is determined not to be met. Alternatively, the computer device determines whether the iteration stopping condition is met based on business feedback data and the current correction path priority strategy.

[0066] Optionally, if it is determined that the iteration stop condition is not met in the current iteration, the current path priority correction strategy is used as the path priority strategy to be corrected in the next iteration, and the step of determining the strategy correction reference data of the path priority strategy to be corrected is returned to continue execution until the iteration stop condition is met and the iteration stops.

[0067] In some embodiments, determining the policy correction reference data for the path priority strategy to be corrected includes: verifying whether there is updated reach key content in the current iteration; if there is updated reach key content, determining the policy correction reference data for the path priority strategy to be corrected based on business feedback data and the updated reach key content.

[0068] The key content to be reached includes at least one of the collective knowledge graph and business rule information for any candidate reach path. Updated key content to be reached includes at least one of the updated collective knowledge graph and updated business rule information.

[0069] Optionally, it is verified whether updated key reach content exists in the current iteration. If updated key reach content exists, that is, at the time of the current iteration, there is at least one of an updated collective knowledge graph and updated business rule information, the strategy correction reference data for the path priority strategy to be corrected is determined based on the business feedback data and the updated key reach content. In this case, the strategy correction reference data includes the business feedback data and the updated key reach content.

[0070] Optionally, in the absence of updated key content to be reached, business feedback data can be used as reference data for policy correction when determining the priority strategy for paths to be corrected.

[0071] In this embodiment, during each iteration, if an update to the collective knowledge graph or business rule information is detected, policy correction reference data is determined based on the updated key content (including at least one of the updated collective knowledge graph and updated business rule information) and business feedback data to determine the path priority strategy to be corrected. This ensures that the path priority strategy can be adaptively corrected based on the updated collective knowledge graph and updated business rule information during each iteration, ensuring the effectiveness and accuracy of the iteration.

[0072] In some embodiments, after obtaining the current corrected path priority strategy obtained in the current iteration, the method further includes: determining the effective candidate reach path among multiple candidate reach paths that is valid; and determining whether the current iteration meets the iteration stop condition based on the current corrected path priority strategy and the effective candidate reach path.

[0073] For any candidate reach path, if the visit content is successfully sent to the target new object after being sent according to the corresponding candidate reach path, then the candidate reach path is valid and is considered a valid candidate reach path belonging to the valid reach.

[0074] For example, the computer device acquires business feedback data and, based on this data, identifies valid candidate reach paths from among multiple candidate reach paths. The computer device then determines the priority ranking of these valid candidate reach paths based on the current path priority adjustment strategy. Based on the priority ranking of each valid candidate reach path, the computer device determines whether the current iteration meets the iteration stopping condition. For instance, if each valid candidate reach path has a high priority ranking in the current path priority adjustment strategy, then the current iteration meets the iteration stopping condition; otherwise, the iteration stopping condition is not met.

[0075] In the above embodiments, the current iteration is checked for whether the iteration stop condition is met by using the current path priority strategy and the effective candidate reachable paths, so as to accurately evaluate whether the current path priority strategy for the current iteration is optimal.

[0076] In some embodiments, determining whether the current iteration meets the iteration stopping condition based on the current path priority policy and valid candidate reach paths includes: obtaining the number of path thresholds; verifying whether valid candidate reach paths belong to the candidate reach paths with higher priority in the current path priority policy based on the number of path thresholds; if each valid candidate reach path belongs to the candidate reach paths with higher priority in the current path priority policy, then it is determined that the current iteration meets the iteration stopping condition; if at least one valid candidate reach path does not belong to the candidate reach paths with higher priority in the current path priority policy, then it is determined that the current iteration does not meet the iteration stopping condition.

[0077] The path threshold number is used to determine which paths belong to a higher priority ranking. For example, if the path threshold number is 5, then after sorting by priority from largest to smallest, the top 5 paths are those with higher priority rankings. Since the priority ranking of each candidate reach path in the current revised path priority strategy obtained in each iteration is different, the path threshold number can be used to determine the higher priority ranking in the current revised path priority strategy obtained in each iteration.

[0078] For example, the computer device determines a path threshold number based on the number of valid candidate reach paths. For instance, the number of paths may be less than or equal to the path threshold number. Of course, the path threshold number can also be flexibly set according to actual needs.

[0079] For example, the priority ranking of the path priority strategy in the current correction is determined based on the number of path thresholds.

[0080] For example, for each valid candidate reach path, the computer device checks whether the priority ranking of the valid candidate reach path in the current path priority correction strategy is among the top priority rankings. If so, the valid candidate reach path is among the top priority candidate reach paths in the current path priority correction strategy. If not, the valid candidate reach path is not among the top priority candidate reach paths in the current path priority correction strategy. If all valid candidate reach paths are among the top priority candidate reach paths in the current path priority correction strategy, then the current iteration satisfies the iteration stopping condition. If at least one valid candidate reach path is not among the top priority candidate reach paths in the current path priority correction strategy, then the current iteration does not satisfy the iteration stopping condition.

[0081] For example, suppose the valid candidate reach paths are paths 1 to 5, and the number of paths is 5. If paths 1 to 5 are the top 5 in the priority ranking of the current path priority adjustment strategy, then the iteration stopping condition is met. If paths 1 to 5 are ranked 3rd to 7th in the priority ranking of the current path priority adjustment strategy, then paths 4 and 5 are not high priority and the iteration stopping condition is not met.

[0082] Furthermore, assuming the valid candidate reach paths are paths 1 to 5, and the path threshold is 7, if paths 1 to 5 are among the top 5 in the current path priority adjustment strategy, then the iteration stopping condition is satisfied. If paths 1 to 5 rank 3 to 7 in the current path priority adjustment strategy, then the iteration stopping condition is satisfied. If paths 1 to 5 rank 7 to 11 in the current path priority adjustment strategy, then paths 2 to 5 are not among the top priorities, and therefore the iteration stopping condition is not met.

[0083] In the above embodiments, the number of path thresholds is used to verify whether the effective candidate reach path belongs to the candidate reach path with higher priority in the current path priority correction strategy. This can effectively and accurately verify whether the current iteration has stopped, thereby accurately evaluating whether the current path priority correction strategy of the current iteration is optimal.

[0084] Step 208: Based on the final correction path priority strategy corresponding to the last iteration, determine the target reach path for the newly added target object from multiple candidate reach paths.

[0085] For example, the computer device determines the target reach path for the newly added target object with the highest priority ranking in the final revised path priority strategy corresponding to the last iteration. Of course, it can also select a candidate reach path that matches the business needs from multiple candidate reach paths as the target reach path, based on the priority ranking of each candidate reach path indicated by the final revised path priority strategy corresponding to the last iteration, according to actual business needs.

[0086] In the aforementioned dynamic priority object outreach method, an initial path priority strategy for the target new object to be reached is obtained. This initial path priority strategy indicates the priority ranking of multiple candidate outreach paths for the target new object. Each candidate outreach path is a path from a corresponding already reached object to the target new object. For the current iteration, the path priority strategy to be corrected in the current iteration is obtained. The path priority strategy to be corrected in the first iteration is the initial path priority strategy, and the path priority strategy to be corrected in subsequent iterations is the previous corrected path priority strategy obtained in the previous iteration. Policy correction reference data for the path priority strategy to be corrected is determined. Based on the policy correction reference data and the path priority strategy to be corrected, the path priority strategy to be corrected is dynamically adjusted using a policy correction model to obtain the current corrected path priority strategy for the current iteration. This current corrected path priority strategy is then used as the path priority strategy to be corrected in the next iteration. The process returns to the step of determining the policy correction reference data for the path priority strategy to be corrected and continues to execute, performing multiple iterative optimization corrections until the iteration stops when the iteration stopping condition is met, thereby continuously optimizing the priority ranking of each candidate outreach path. In this way, based on the final correction path priority strategy corresponding to the last iteration, the target reach path that matches the target new object can be accurately selected from multiple candidate reach paths. In the above process, the priority of each candidate reach path is continuously adjusted through iteration, which can deeply analyze the suitable candidate reach paths that match the target new object, thereby improving the accuracy of reach path determination.

[0087] In a specific embodiment, the implementation is described using a server as the main body, and the specific steps are as follows:

[0088] Step 1: The server retrieves the collective knowledge graph, which includes multiple reach paths. Each reach path consists of multiple sequentially connected objects, with edges connecting any two related objects. Based on the collective knowledge graph, multiple candidate reach paths are determined for the target new object to be reached. Each candidate reach path includes the target new object.

[0089] Step 2: For each candidate reach path, the server determines the scores corresponding to the candidate reach path in the structural and business dimensions based on the path information and business rule information corresponding to the candidate reach path, and merges the scores corresponding to the structural and business dimensions to obtain the priority score corresponding to the candidate reach path; based on the priority scores corresponding to each candidate reach path, the priority ranking of each candidate reach path is determined; based on the priority ranking of each candidate reach path, the initial path priority strategy for the target new object to be reached is determined.

[0090] Step 3: The server obtains the initial path priority policy for the target new object to be reached. The initial path priority policy indicates the priority order of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding reached objects to the target new object.

[0091] Step 4: For the current iteration, the server obtains the path priority policy to be corrected for the current iteration. The path priority policy to be corrected for the first iteration is the initial path priority policy, and the path priority policy to be corrected for non-first iterations is the previous corrected path priority policy obtained in the previous iteration.

[0092] Step 5: The server verifies whether there is updated key content for reaching the target audience in the current iteration; if there is updated key content for reaching the target audience, the server determines the policy correction reference data for the path priority strategy to be corrected based on the business feedback data and the updated key content for reaching the target audience.

[0093] Step 6: Based on the policy correction reference data and the path priority policy to be corrected, the server obtains the current corrected path priority policy in the current iteration through the policy correction model.

[0094] Step 7: The server determines the valid candidate reach paths among multiple candidate reach paths; obtains the path threshold number; based on the path threshold number, verifies whether the valid candidate reach paths belong to the candidate reach paths with higher priority in the current path priority adjustment strategy; if each valid candidate reach path belongs to the candidate reach paths with higher priority in the current path priority adjustment strategy, then it is determined that the current iteration meets the iteration stopping condition; if there is at least one valid candidate reach path that does not belong to the candidate reach paths with higher priority in the current path priority adjustment strategy, then it is determined that the current iteration does not meet the iteration stopping condition.

[0095] Step 8: If the iteration stopping condition is not met in the current iteration, the server will use the current path priority correction policy as the path priority policy to be corrected in the next iteration. Return to step 5 and continue execution until the iteration stopping condition is met.

[0096] Step 9: The server determines the target reach path for the newly added object from multiple candidate reach paths based on the final correction path priority strategy corresponding to the last iteration.

[0097] In the above embodiments, an initial path priority strategy for the target new object is obtained. This initial path priority strategy indicates the priority ranking of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding already reached objects to the target new object. For the current iteration, the path priority strategy to be corrected in the current iteration is obtained. The path priority strategy to be corrected in the first iteration is the initial path priority strategy, and the path priority strategy to be corrected in subsequent iterations is the previous corrected path priority strategy obtained in the previous iteration. Policy correction reference data for the path priority strategy to be corrected is determined. Based on the policy correction reference data and the path priority strategy to be corrected, the path priority strategy to be corrected is dynamically adjusted through a policy correction model to obtain the current corrected path priority strategy obtained in the current iteration. Then, the current corrected path priority strategy is used as the path priority strategy to be corrected in the next iteration. The step of determining the policy correction reference data for the path priority strategy to be corrected is returned to continue execution to perform multiple iterations of optimization and correction until the iteration stops when the iteration stopping condition is met, so as to continuously optimize the priority ranking of each candidate reach path. In this way, based on the final correction path priority strategy corresponding to the last iteration, the target reach path that matches the target new object can be accurately selected from multiple candidate reach paths. In the above process, the priority of each candidate reach path is continuously adjusted through iteration, which can deeply analyze the suitable candidate reach paths that match the target new object, thereby improving the accuracy of reach path determination.

[0098] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0099] Based on the same inventive concept, this application also provides a dynamic priority object outreach device for implementing the dynamic priority object outreach method described above. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more embodiments of the dynamic priority object outreach device provided below can be found in the limitations of the dynamic priority object outreach method described above, and will not be repeated here.

[0100] In one exemplary embodiment, such as Figure 3 As shown, a dynamic priority object outreach device 300 is provided, including: a first strategy acquisition module 302, a second strategy acquisition module 304, a strategy correction module 306, and an outreach path determination module 308, wherein:

[0101] The first strategy acquisition module 302 is used to acquire the initial path priority strategy for the target new object to be reached. The initial path priority strategy indicates the priority order of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding reached objects to the target new object.

[0102] The second strategy acquisition module 304 is used to acquire the path priority strategy to be corrected in the current iteration for the current iteration, wherein the path priority strategy to be corrected corresponding to the first iteration is the initial path priority strategy, and the path priority strategy to be corrected corresponding to non-first iterations is the previous corrected path priority strategy obtained in the previous iteration.

[0103] The strategy correction module 306 is used to determine the strategy correction reference data of the path priority strategy to be corrected. Based on the strategy correction reference data and the path priority strategy to be corrected, the strategy correction model is used to obtain the current corrected path priority strategy obtained in the current iteration. The current corrected path priority strategy is used as the path priority strategy to be corrected in the next iteration. The step of determining the strategy correction reference data of the path priority strategy to be corrected is returned to continue execution until the iteration stops when the iteration stop condition is met.

[0104] The reach path determination module 308 is used to determine the target reach path of the newly added target object from multiple candidate reach paths based on the final modified path priority strategy corresponding to the last iteration.

[0105] In some embodiments, the first strategy acquisition module 302 is used to acquire a collective knowledge graph, which includes multiple reach paths, each reach path including multiple objects connected in sequence, and there is an edge between two objects that have an association relationship; based on the collective knowledge graph, multiple candidate reach paths are determined for the target new object to be reached, each candidate reach path including the target new object; based on the path information corresponding to each candidate reach path, the priority ranking of each candidate reach path is determined; based on the priority ranking of each candidate reach path, an initial path priority strategy for the target new object to be reached is determined.

[0106] In some embodiments, the first strategy acquisition module 302 is used to determine, for each candidate reach path, the scores corresponding to the candidate reach path in the structural dimension and the business dimension respectively based on the path information and business rule information corresponding to the candidate reach path, and to merge the scores corresponding to the structural dimension and the business dimension respectively to obtain the priority score corresponding to the candidate reach path; and to determine the priority ranking of each candidate reach path based on the priority scores corresponding to each candidate reach path.

[0107] In some embodiments, the strategy correction module 306 is used to verify whether there is updated key content for reaching the target audience in the current iteration; if there is updated key content for reaching the target audience, the strategy correction reference data for the path priority strategy to be corrected is determined based on the business feedback data and the updated key content for reaching the target audience.

[0108] In some embodiments, the apparatus further includes an iteration stop determination module, configured to determine a valid candidate reach path among multiple candidate reach paths that is validly reachable; and to determine whether the current iteration meets the iteration stop condition based on the current path priority policy and the valid candidate reach path.

[0109] In some embodiments, the iteration stop determination module is used to obtain the number of path thresholds, and based on the number of path thresholds, to verify whether the valid candidate reach paths belong to the candidate reach paths with higher priority in the current path priority correction strategy; if each valid candidate reach path belongs to the candidate reach paths with higher priority in the current path priority correction strategy, then it is determined that the current iteration meets the iteration stop condition; if there is at least one valid candidate reach path that does not belong to the candidate reach paths with higher priority in the current path priority correction strategy, then it is determined that the current iteration does not meet the iteration stop condition.

[0110] Each module in the aforementioned dynamic priority object access device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can invoke and execute the operations corresponding to each module.

[0111] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 4 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and databases. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media to run. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a dynamic priority object access method.

[0112] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0113] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0114] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0115] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0116] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0117] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0118] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0119] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A dynamic priority object reach method, characterized by, The method includes: Obtain an initial path priority strategy for the target new object to be reached. The initial path priority strategy indicates the priority order of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding reached objects to the target new object. For the current iteration, obtain the path priority strategy to be corrected in the current iteration. The path priority strategy to be corrected corresponding to the first iteration is the initial path priority strategy, and the path priority strategy to be corrected corresponding to non-first iterations is the previous corrected path priority strategy obtained in the previous iteration. Determine the policy correction reference data for the path priority strategy to be corrected. Based on the policy correction reference data and the path priority strategy to be corrected, obtain the current corrected path priority strategy obtained in the current iteration through the policy correction model. Use the current corrected path priority strategy as the path priority strategy to be corrected in the next iteration. Return to the step of determining the policy correction reference data for the path priority strategy to be corrected and continue execution until the iteration stops when the iteration stop condition is met. Based on the final correction path priority strategy corresponding to the last iteration, the target reach path for the newly added target object is determined from multiple candidate reach paths.

2. The method of claim 1, wherein, The strategy for obtaining the initial path priority of the newly added target object to be reached includes: Obtain a collective knowledge graph, which includes multiple access paths. Each access path includes multiple objects connected in sequence, and there are edges between two objects that have an association relationship. Based on the collective knowledge graph, multiple candidate reach paths are determined for the target new object to be reached, and each candidate reach path includes the target new object; Based on the path information corresponding to each candidate reach path, the priority ranking of each candidate reach path is determined. Based on the priority ranking of each candidate reach path, the initial path priority strategy for the target new object to be reached is determined.

3. The method according to claim 2, characterized in that, The process of determining the priority ranking of each candidate reach path based on its corresponding path information includes: For each candidate reach path, based on the path information and business rule information corresponding to the candidate reach path, the scores corresponding to the candidate reach path in the structural dimension and business dimension are determined respectively, and the scores corresponding to the structural dimension and business dimension are merged to obtain the priority score corresponding to the candidate reach path. Based on the priority score corresponding to each candidate reach path, the priority ranking of each candidate reach path is determined.

4. The method according to claim 1, characterized in that, The policy correction reference data for determining the priority policy of the path to be corrected includes: Verify whether there are any updated key content in the current iteration; If updated key content for reaching out exists, the policy correction reference data for the path priority strategy to be corrected is determined based on business feedback data and the updated key content for reaching out.

5. The method according to claim 1, characterized in that, After obtaining the current corrected path priority strategy obtained in the current iteration, the method further includes: Identify the valid candidate reach paths among multiple candidate reach paths; Based on the current path priority strategy and the effective candidate reach paths, determine whether the current iteration meets the iteration stop condition.

6. The method according to claim 5, characterized in that, The step of determining whether the current iteration meets the iteration stopping condition based on the current path priority strategy and valid candidate reach paths includes: Obtain the number of path thresholds, and based on the number of path thresholds, verify whether the valid candidate reach path belongs to the candidate reach path with the highest priority in the current path priority correction strategy. If each valid candidate access path belongs to the candidate access path with the highest priority in the current path priority strategy, then the current iteration is determined to meet the iteration stop condition. If at least one valid candidate access path does not belong to the candidate access path with the highest priority in the current path priority strategy, then it is determined that the current iteration does not meet the iteration stop condition.

7. A dynamic priority object reach device, characterized in that, The device includes: The first strategy acquisition module is used to acquire an initial path priority strategy for the target new object to be reached. The initial path priority strategy indicates the priority order of multiple candidate reach paths for the target new object. The candidate reach paths are paths from the corresponding reached objects to the target new object. The second strategy acquisition module is used to acquire the path priority strategy to be corrected in the current iteration for the current iteration. The path priority strategy to be corrected corresponding to the first iteration is the initial path priority strategy, and the path priority strategy to be corrected corresponding to non-first iterations is the previous corrected path priority strategy obtained in the previous iteration. The strategy correction module is used to determine the strategy correction reference data of the path priority strategy to be corrected, and based on the strategy correction reference data and the path priority strategy to be corrected, obtain the current corrected path priority strategy obtained in the current iteration through the strategy correction model, use the current corrected path priority strategy as the path priority strategy to be corrected in the next iteration, return to the step of determining the strategy correction reference data of the path priority strategy to be corrected and continue to execute until the iteration stops when the iteration stop condition is met. The reach path determination module is used to determine the target reach path of the newly added target object from multiple candidate reach paths based on the final modified path priority strategy corresponding to the last iteration.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images