Processing method, system and equipment for automatic rule configuration process and medium

Abstract

The invention relates to the technical field of data processing, in particular to an automatic rule configuration process processing method, system and device and a medium, and the method specifically comprises the following steps: obtaining first delivery rule data, and carrying out semantic analysis and structured extraction on the first delivery rule data to form second delivery rule data; establishing a standard index library, and performing index systematic construction on the second putting rule data according to the standard index library to generate a multi-way tree rule; performing automatic verification on the multi-way tree rule according to a rule verification tool, determining whether the multi-way tree rule is reasonable or not, and if the multi-way tree rule is not reasonable, optimizing the multi-way tree rule; and obtaining historical advertisement putting data, and performing an advertisement putting scene simulation test on the multi-way tree rule according to the historical advertisement putting data. According to the invention, through generation, optimization and application of the multi-way tree rule, intelligent configuration, management and optimization of the advertisement putting rule are realized.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a processing method, system, device and medium for an automated rule configuration process. Background Art

[0002] In today's increasingly competitive gaming industry, the efficiency and flexibility of the game automation delivery system, as a key tool to promote accurate product reach to users and increase market share, are directly related to the success of game operations. However, the challenges faced by the current system in terms of rule configuration and management are becoming increasingly prominent, which not only limits the optimization potential of the system, but also increases operating costs and risks, which are specifically manifested in the following aspects:

[0003] 1. Complexity and inefficiency of rule configuration: Different business teams or individuals have significant differences in naming and understanding of key indicators when configuring delivery rules, which leads to frequent communication barriers and misunderstandings during the rule configuration process, greatly reducing configuration efficiency. At the same time, the existing rule configuration process is highly dependent on manual work. From demand sorting to rule writing, business personnel and technical personnel are required to work closely together, but the differences in professional backgrounds of both parties often lead to poor information transmission, increasing the complexity and error rate of the configuration process.

[0004] 2. Heavy burden of technology conversion: Technical personnel need to manually convert the rules provided by business personnel in natural language or document form into code or configuration format that the system can understand and execute. This process is time-consuming and error-prone, which seriously affects the system's ability to respond quickly. As the scale of games expands and the delivery strategy becomes more complicated, the number of rules that need to be converted increases dramatically, further increasing the workload of the technical team.

[0005] 3. The verification process is cumbersome: After the rule configuration is completed, multiple rounds of manual verification are required, including the correctness of business logic, system compatibility and potential conflicts. This process is time-consuming and prone to missing problems.

[0006] 4. Lack of automated testing: The lack of effective automated verification tools makes it difficult to fully test and optimize newly configured rules before going online, increasing system instability and uncertainty in delivery effects.

[0007] 5. Limited system flexibility and scalability: The current rule configuration method often leads to relatively rigid system rules, which is difficult to quickly adapt to market changes or business adjustments, limiting the innovation and flexibility of the launch strategy. With the enrichment of game features and the diversification of operational needs, the system needs to frequently expand or reconstruct functions, but the limitations of the existing architecture in rule management make this process extremely difficult. Summary of the invention

[0008] The object of the present invention is to provide a method, system, device and medium for processing an automated rule configuration process to solve at least one of the above-mentioned prior art problems.

[0009] In a first aspect, the present invention provides a method for processing an automated rule configuration process, the method specifically comprising:

[0010] Acquire first delivery rule data, perform semantic analysis and structured extraction on the first delivery rule data, and form second delivery rule data;

[0011] Establishing a standard indicator library, and constructing an indicator system for the second delivery rule data according to the standard indicator library to generate a multi-branch tree rule;

[0012] Automatically verify the multi-branch tree rule according to the rule verification tool to determine whether the multi-branch tree rule is reasonable, and if it is unreasonable, optimize the multi-branch tree rule;

[0013] Acquire historical advertisement delivery data, and perform advertisement delivery scenario simulation test on the multi-branch tree rule according to the historical advertisement delivery data;

[0014] An advertisement delivery scheme for each advertisement is determined according to the multi-branch tree rule, and display analysis is performed on advertisements that do not comply with the multi-branch tree rule.

[0015] In a second aspect, the present invention provides a processing system for an automated rule configuration process, the system specifically comprising:

[0016] A first processing module is used to obtain first delivery rule data, perform semantic analysis and structured extraction on the first delivery rule data, and form second delivery rule data;

[0017] The second processing module is used to establish a standard indicator library, and to construct an indicator system for the second delivery rule data according to the standard indicator library to generate a multi-branch tree rule;

[0018] A third processing module is used to automatically check the multi-tree rule according to a rule checking tool to determine whether the multi-tree rule is reasonable, and if it is unreasonable, optimize the multi-tree rule;

[0019] A fourth processing module is used to obtain historical advertising data, and perform advertising scene simulation test on the multi-branch tree rule according to the historical advertising data;

[0020] The fifth processing module is used to determine the advertising scheme for each advertisement according to the multi-branch tree rule, and to perform display analysis on the advertisements that do not comply with the multi-branch tree rule.

[0021] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor and a computer program stored in the memory, and when the computer program is executed on the processor, a processing method for an automated rule configuration process as described in any one of the above methods is implemented.

[0022] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the processing method of the automated rule configuration process as described in any one of the above methods is implemented.

[0023] Compared with the prior art, the present invention has at least one of the following technical effects:

[0024] 1. The present invention realizes the intelligent configuration, management and optimization of advertising delivery rules through the generation, optimization and application of multi-branch tree rules, improves the accuracy and effect of advertising delivery, and provides advertisers with more scientific and efficient delivery decision support.

[0025] 2. Through semantic analysis and structured extraction, complex and ambiguous delivery rule data is transformed into second delivery rule data with clear structure and easy to process, which greatly improves the efficiency and accuracy of rule configuration.

[0026] 3. Establish a standard indicator library to achieve standardization and normalization of the indicator system, reduce configuration errors caused by differences in naming and understanding, and improve the maintainability and scalability of the system.

[0027] 4. Introduce rule verification tools for automated verification to promptly identify and optimize unreasonable rules and ensure the rationality and effectiveness of the delivery rules.

[0028] 5. Use historical advertising data to conduct scenario simulation tests, optimize delivery strategies through machine learning algorithms, and improve the accuracy and effectiveness of advertising delivery. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0030] Figure 1 It is a flowchart of a processing method of an automated rule configuration process provided by an embodiment of the present invention;

[0031] Figure 2 It is a structural diagram of a processing system for an automated rule configuration process provided by an embodiment of the present invention;

[0032] Figure 3 It is a structural schematic diagram of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0033] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0034] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0035] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0036] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0037] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0038] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0039] In the embodiment of the present application, the execution subject of the process includes a terminal device, which includes but is not limited to: a server, a computer, a smart phone, a tablet computer, and other devices capable of executing the method disclosed in the present application. Figure 1 A flowchart of a method for processing an automated rule configuration process disclosed in the first embodiment of the present invention is shown, and is described in detail as follows:

[0040] S101, obtaining first delivery rule data, performing semantic analysis and structured extraction on the first delivery rule data, and forming second delivery rule data.

[0041] In this embodiment, in the game automated delivery system, business personnel use the XMind tool to create a delivery rule file containing multiple levels and complex logical relationships in order to formulate a new advertising delivery strategy. The file describes in detail the advertising display logic and priority under different user groups, advertising channels, delivery time and other conditions. In order to convert these complex delivery rules into a format that the system can recognize, the system needs to perform semantic analysis and structured extraction on the XMind file.

[0042] Specifically, business personnel upload XMind files through the system interface. After the system receives the files, it performs format verification to ensure that the files are complete and not damaged. It uses a special XMind parsing library (such as the xmindparser library in Python) to read the XMind file content, parse it into structured data in JSON format, and extract all topics and subtopics, as well as the hierarchical relationships and attribute information between them (such as titles, notes, etc.). Natural language processing (NLP) is performed on the titles of each topic and subtopic to identify keywords and phrases, such as "user age", "advertising channel", "delivery time", etc. The keywords are further parsed using the predefined vocabulary and rule library to determine their specific business meanings and constraints. In the structured data, the logical relationships between the various rules (such as "and", "or" relationships), conditional judgments (such as "greater than", "less than", etc.) and execution actions (such as "display ad A", "adjust bid", etc.) are clearly expressed. The structured data is stored in the system as the second delivery rule data for subsequent processing.

[0043] In this embodiment, by parsing the XMind file, the system can support business personnel to create complex delivery rules in a graphical way, reducing the threshold and difficulty of rule configuration. The semantic analysis process can automatically identify and parse the keywords and constraints in the delivery rules, reducing configuration errors caused by differences in human understanding. The second delivery rule data after structured extraction has a clear hierarchical structure and clear logical relationships, which is easy for technical personnel to understand and maintain. In general, the automated parsing and extraction process greatly shortens the time for rule configuration and improves work efficiency.

[0044] In some embodiments, in the above step S101, the semantic analysis and structured extraction of the first delivery rule data to form the second delivery rule data specifically includes:

[0045] Performing semantic analysis on the first delivery rule data using natural language processing technology to determine key elements in the first delivery rule data and obtain third delivery rule data;

[0046] Extracting the core semantic information of the third delivery rule data, constructing a knowledge graph through the core semantic information, and vectorizing the knowledge graph using a graph embedding algorithm to obtain a low-dimensional dense vector representation;

[0047] According to the low-dimensional dense vector representation, a clustering algorithm is used to perform semantic similarity calculation and cluster analysis on the third delivery rule data to generate a plurality of rule clusters;

[0048] An association rule mining algorithm is used to identify association relationships and dependency relationships between multiple rule clusters, and second delivery rule data is formed according to the association relationships and dependency relationships.

[0049] In this embodiment, a word segmentation algorithm in natural language processing technology is used to perform word segmentation processing on the first delivery rule data, obtain keywords and phrases in the delivery rule data, and obtain a delivery rule keyword set after word segmentation. The delivery rule keyword set after word segmentation is tagged with parts of speech by a part-of-speech tagging algorithm, and the part of speech of each keyword is determined to obtain a delivery rule keyword set with part-of-speech tags. According to a named entity recognition algorithm, named entity recognition is performed on the delivery rule keyword set with part-of-speech tags to identify key entities in the delivery rule data, such as brand names, product categories, etc., to obtain a delivery rule key entity set. A dependency syntax analysis algorithm is used to perform syntax analysis on the first delivery rule data, obtain the dependency relationship between keywords in the delivery rule data, determine the semantic relationship between keywords, and obtain a semantic dependency relationship graph between delivery rule keywords. A semantic role labeling algorithm is used to label the semantic dependency relationship graph between delivery rule keywords, and the semantic role of each keyword in the delivery rule is determined, such as conditions, target audience, delivery time, etc., to obtain a delivery rule keyword semantic dependency relationship graph with semantic role labels. According to the knowledge base and ontology of the delivery rule business domain, the semantic dependency graph of delivery rule keywords with semantic role annotations is semantically expanded and inferred to obtain the key elements and constraints implicit in the delivery rule data, and obtain the expanded delivery rule key element set. Using the rule template of the delivery rule business domain, the expanded delivery rule key element set is structured to obtain structured third delivery rule data, determine the key elements of the delivery rule and their logical relationships, and form a structured representation of the delivery rule that can be executed by the delivery system.

[0050] According to the third delivery rule data, natural language processing technology is used to obtain the core semantic information of the delivery rule data through semantic analysis and keyword extraction, and the key elements of the delivery rules and the relationships between them are obtained. According to the obtained core semantic information, the knowledge graph construction technology is used to organize the key elements of the delivery rules into a structured knowledge graph by defining entities, attributes and relationships, and obtain a knowledge graph model representing the semantics of the delivery rules. According to the constructed knowledge graph, a graph embedding algorithm is used to vectorize the entities and relationships in the knowledge graph, convert the high-dimensional sparse knowledge graph into a low-dimensional dense vector representation, and obtain the semantic vector representation of the delivery rules.

[0051] According to the low-dimensional dense vector representation, the K-means clustering algorithm is used to perform cluster analysis on the third placement rule data. By calculating the Euclidean distance between vectors as a similarity measure, the rule data with high similarity are divided into the same cluster to obtain a preliminary rule cluster division result. The preliminary rule cluster generated in the previous step is obtained, and the silhouette coefficient evaluation method is used to calculate the average distance a between each rule data point and other points in the cluster and the minimum average distance b between the point and other cluster points. The silhouette coefficient of each data point is obtained by the formula (ba) / max(a, b). The value range of the silhouette coefficient is [-1,1]. The closer to 1, the data point is assigned to the appropriate cluster. According to the silhouette coefficient calculated in the previous step, it is judged whether the current clustering result is reasonable. If there are many data points with low silhouette coefficients, it is considered that the current clustering effect is not good, and it is necessary to adjust the clustering parameters such as the number of clusters K, and re-cluster until a better clustering result is obtained. By analyzing the clustering results, the central vector of each rule cluster is obtained. This vector can represent the average semantic features of the cluster. The central vector of each rule cluster is used as the semantic representation of the cluster to form multiple representative rule clusters. The internal consistency of each rule cluster is evaluated by the average cosine similarity between the central vector of the rule cluster and the rule vectors in the cluster. The higher the cosine similarity, the more similar the semantics of the rules in the cluster, so as to judge the clustering effect. For rule clusters with poor internal consistency, they can be further subdivided or merged to finally obtain rule clusters with high semantic similarity and good discrimination. By analyzing multiple rule clusters, the association and dependency relationships between the rule clusters are determined, and the association and dependency data between the rule clusters are obtained. According to the association and dependency data between the rule clusters, the second delivery rule data is formed to determine the final delivery strategy.

[0052] S102, establishing a standard indicator library, and constructing an indicator system for the second delivery rule data according to the standard indicator library to generate a multi-branch tree rule.

[0053] In this embodiment, in the game advertising delivery system, in order to uniformly manage and optimize the advertising delivery strategy, the system needs to establish a standard indicator library, and according to the library, the second delivery rule data imported by the business personnel is systematically constructed as an indicator, and finally a multi-branch tree rule is generated. These multi-branch tree rules will serve as the basis for advertising delivery decisions and guide the precise delivery of advertisements.

[0054] Specifically, sort out all key indicators involved in the advertising business, such as user behavior indicators (click-through rate, conversion rate), advertising channel indicators (CPC, CPM), delivery time indicators (time period, date), etc. Define and describe each indicator, clarify its calculation method and business meaning. Sort all indicators by category, level and importance to form a standard indicator library, and store the standard indicator library in the database for subsequent query and use.

[0055] Then, read the second delivery rule data, parse its structure and content, and extract the indicator items related to the standard indicator library. According to the classification and hierarchy of the standard indicator library, classify and organize the indicator items in the second delivery rule data. Use decision tree algorithm or similar methods to build a multi-level indicator system based on the logical relationship and dependency relationship between indicators. Each layer of indicators represents a more specific delivery dimension or condition. During the construction process, ensure the integrity and consistency of the indicator system to avoid redundancy and conflict.

[0056] Based on the constructed indicator system, the second delivery rule data is converted into a multi-tree structure. Each node represents an indicator or condition judgment, and the child node represents the further subdivision or execution action under the condition. Use graph theory algorithms or special data structures to represent multi-tree rules to ensure efficient traversal and query. Name and annotate the multi-tree rules so that business and technical personnel can understand and maintain them.

[0057] In this embodiment, the establishment of a standard indicator library provides a unified basis and standard for the formulation and evaluation of advertising delivery strategies, improving the maintainability and scalability of the system. The systematic construction of indicators transforms complex delivery rules into a clear multi-branch tree structure, making the advertising delivery decision process more intuitive and efficient. The multi-branch tree rules can comprehensively consider various delivery conditions and logical relationships to ensure the accuracy and scientificity of advertising delivery decisions. At the same time, since the multi-branch tree rules are configurable, the system can flexibly respond to market changes and business needs, and dynamically adjust and optimize the delivery strategy.

[0058] In some embodiments, in the above step S102, constructing the second delivery rule data into an indicator system according to the standard indicator library to generate a multi-branch tree rule specifically includes:

[0059] Using the hierarchical analysis method to classify and organize the second delivery rule data, and obtain dimension information and attribute information of the second delivery rule data;

[0060] According to the dimension information and attribute information, a multi-level indicator system is constructed by using a decision tree algorithm, and indicator weights are assigned to indicators at each level of the multi-level indicator system by using a fuzzy comprehensive evaluation method;

[0061] According to the standard indicator library, the naming of indicators at each level in the multi-level indicator system is standardized and revised to form a multi-branch tree rule.

[0062] In this embodiment, the second delivery rule data is classified and organized by using the hierarchical analysis method, and data with similar attributes are classified into the same category to form a plurality of data sets of different dimensions, thereby obtaining the dimension information of the second delivery rule data. By analyzing the characteristics of each dimensional data set, key information that can reflect the essential attributes of the dimensional data is extracted, and the attribute information of the second delivery rule data in different dimensions is obtained.

[0063] Obtain relevant dimension information and attribute information based on business scenarios, and determine the construction objectives and scope of the multi-level indicator system by analyzing business needs. Use the decision tree algorithm to process the acquired dimension information and attribute information, and construct a multi-level indicator system that meets business needs through recursive partitioning and feature selection, and obtain the hierarchical structure and logical relationship between indicators at all levels. Based on the constructed multi-level indicator system, use the fuzzy comprehensive evaluation method to assign weights to indicators at all levels, determine the importance of each indicator in the entire indicator system through expert scoring and weight calculation, and obtain the indicator weight distribution that reflects business attributes.

[0064] According to the standard indicator library, indicators at all levels in the multi-level indicator system are obtained to obtain a preliminary indicator naming list. By analyzing the business attributes, the specific business meaning of each indicator is determined, and the corresponding relationship between the business attributes and the indicators is obtained. According to the business attributes, the preliminary indicator naming list is corrected using standardized naming rules to obtain a standardized indicator naming list. By hierarchically dividing the standardized indicator naming list, the hierarchical relationship of indicators at all levels is determined to obtain the preliminary structure of the multi-level indicator system. According to the preliminary structure of the multi-level indicator system, a multi-tree structure is constructed for indicators at all levels to obtain an indicator system with a multi-tree structure. By verifying the business attributes of the indicator system with a multi-tree structure, the naming and structure of the final multi-level indicator system are determined to obtain a standardized multi-level indicator system, namely, a multi-tree rule.

[0065] S103, automatically verifying the multi-branch tree rule according to a rule verification tool to determine whether the multi-branch tree rule is reasonable, and optimizing the multi-branch tree rule if it is unreasonable.

[0066] In this embodiment, in the game advertisement delivery system, in order to ensure the accuracy and effectiveness of the multi-branch tree rules, the system introduces a rule verification tool. The tool can automatically verify the constructed multi-branch tree rules to check whether there are logical errors, data conflicts or problems that do not meet business specifications. Once unreasonableness is found, the system will automatically or prompt the user to optimize the multi-branch tree rules.

[0067] Specifically, design and implement a set of rule verification tools, which should include multiple verification modules to verify different aspects of the polytree rules. For example, the logic verification module checks whether the logical relationship and conditional judgment in the rules are correct; the data verification module checks whether the data used in the rules is valid and consistent; the business specification verification module checks whether the rules comply with established business specifications and standards. Define detailed verification rules and verification algorithms for each verification module to ensure that all aspects of the polytree rules can be fully covered.

[0068] Then, the constructed multi-branch tree rules are input into the rule verification tool. The rule verification tool verifies the multi-branch tree rules item by item according to the preset verification process and verification rules. For each verification item, the tool will output the verification results, including the mark of verification passed or failed, and the specific reason and location of the verification failure. The system receives the verification results of the rule verification tool and analyzes and processes them. If all verification items pass, it means that the multi-branch tree rule is reasonable, and the system can continue with subsequent delivery operations. If there is a verification failure, the system will prompt the user to optimize the multi-branch tree rule according to the specific reason and location of the verification failure. The optimization may include modifying logical judgments, adjusting data input, correcting business specifications, etc. The user optimizes the multi-branch tree rules according to the system prompts. The optimized rules are input into the rule verification tool again for verification to ensure that the problem has been solved and the rules meet all verification requirements.

[0069] In this embodiment, through automated verification, errors and unreasonableness in the multi-tree rules can be discovered and corrected in a timely manner, thereby improving the quality and accuracy of the rules. Automated verification reduces the links of manual intervention and judgment, and reduces the risk of errors caused by human negligence or misunderstanding. Automated verification can quickly complete the verification of a large number of rules, saving time and energy for manual verification and improving work efficiency.

[0070] In some embodiments, in the above step S103, the multi-tree rule is automatically checked according to the rule checking tool to determine whether the multi-tree rule is reasonable, and if it is unreasonable, the multi-tree rule is optimized, which specifically includes:

[0071] Using a depth-first search algorithm to traverse the multi-branch tree rule, obtaining attribute information and child node relationships of each node of the multi-branch tree rule;

[0072] Based on the attribute information of each node and the relationship between its sub-nodes, determine whether the attribute values ​​of each node meet the predefined constraints, and whether the relationship between nodes meets the preset business logic requirements;

[0073] If the attribute value of each node does not meet the predefined constraint conditions or the relationship between the nodes does not meet the preset business logic requirements, a heuristic search algorithm is used to locally adjust the multi-branch tree rules;

[0074] The adjusted multi-branch tree rules are evaluated by using the hierarchical analysis method, and quantitative scoring is performed from multiple dimensions such as structural complexity, business coverage, and redundancy to obtain quantitative scoring results. Based on the quantitative scoring results, it is determined whether the multi-branch tree rules should be re-optimized.

[0075] In this embodiment, according to the root node of the multi-branch tree rule, a depth-first search algorithm is used to obtain the attribute information of the root node, including the node type, node name, node value, etc., to obtain the complete attribute information of the root node. Through the obtained root node information, it is determined whether the root node has child nodes. If there are child nodes, the number of child nodes is obtained to determine the number of child nodes of the root node. A depth-first search algorithm is used to obtain the attribute information of the first child node of the root node in a left-to-right order, including the node type, node name, node value, etc., to obtain the complete attribute information of the first child node. According to the attribute information of the first child node obtained, it is determined whether the child node has child nodes. If so, the depth-first search algorithm is continued to be used to obtain the attribute information of all child nodes of the child node until the leaf node is traversed to obtain the attribute information of all nodes of the subtree with the first child node as the root. Backtracking to the root node, the same method is used to obtain the attribute information of the second child node of the root node and the attribute information of all child nodes of the child node, to obtain the attribute information of all nodes of the subtree with the second child node as the root. By analogy, the depth-first search algorithm is used to obtain the attribute information of all child nodes of the root node and its descendant nodes until the entire multi-branch tree is traversed to obtain the attribute information of all nodes in the multi-branch tree. In the process of depth-first search, by recording the parent node information of each node, the parent-child relationship between nodes is obtained, and the topological structure between nodes in the multi-branch tree is determined. Based on the attribute information and parent-child relationship information of all nodes obtained, the complete structure and node information of the multi-branch tree rules are obtained, providing basic data for subsequent business logic processing.

[0076] According to the attribute information of each node, the attribute value of the node is obtained, and by comparing it with the predefined constraints, it is determined whether the attribute value of the node meets the constraints, and the legitimacy result of the node attribute is obtained. The depth-first search algorithm is used to traverse the parent-child relationship between nodes, obtain the hierarchical structure and association relationship between nodes, and by matching the node relationship with the preset business logic requirements, it is determined whether the relationship between the nodes meets the business logic, and the compliance result of the node relationship is obtained. According to the legitimacy result of the node attribute and the compliance result of the node relationship, logical operations are used to comprehensively evaluate whether the attributes and relationships of the node meet the predefined constraints and business logic requirements at the same time, and the final verification result of the node is obtained.

[0077] According to the obtained node list and node relationship list that do not meet the requirements, a heuristic search algorithm is used to evaluate the impact of node attribute adjustment and node relationship adjustment on the satisfaction of business rules to obtain the optimal node attribute and relationship adjustment scheme. According to the optimal node attribute and relationship adjustment scheme, a local adjustment algorithm of the tree is used to modify the node attribute values ​​that do not meet the requirements and adjust the node relationships that do not meet the requirements to obtain the adjusted multi-branch tree rules. According to the adjusted multi-branch tree rules, a business rule verification algorithm is used to verify whether each node attribute meets the constraint conditions and whether the node relationships meet the business logic one by one to determine whether the adjusted multi-branch tree rules fully meet the business requirements. If the adjusted multi-branch tree rules fully meet the business requirements, the multi-branch tree rules are determined to be the final business rules; if there are still node attributes or node relationships that do not meet the requirements, return to continue local adjustments until a multi-branch tree rule that fully meets the business requirements is obtained.

[0078] According to the adjusted multi-branch tree rule, its structural complexity needs to be evaluated first. By analyzing the number of nodes, level depth and number of branches of the multi-branch tree, the preliminary quantitative score of the structural complexity is obtained. According to the preliminary quantitative score of the structural complexity, the hierarchical analysis method is used to further refine it. By constructing a judgment matrix, the weights of each node and level are obtained to obtain the final quantitative score of the structural complexity. According to the final quantitative score of the structural complexity, the business coverage rate is evaluated next. By analyzing the application of the multi-branch tree rule in actual business scenarios, the preliminary quantitative score of the business coverage rate is obtained. According to the preliminary quantitative score of the business coverage rate, the hierarchical analysis method is used to further refine it. By constructing a judgment matrix, the weights of each business scenario are obtained to obtain the final quantitative score of the business coverage rate. According to the final quantitative score of the business coverage rate, the redundancy is evaluated next. By analyzing the repeated nodes and redundant paths in the multi-branch tree rule, the preliminary quantitative score of the redundancy is obtained. According to the preliminary quantitative score of the redundancy, the hierarchical analysis method is used to further refine it. By constructing a judgment matrix, the weights of each redundant node and path are obtained to obtain the final quantitative score of the redundancy. According to the final quantitative scores of structural complexity, business coverage and redundancy, the analytic hierarchy process is used to conduct a comprehensive evaluation. By constructing a comprehensive judgment matrix, the comprehensive weights of each dimension are obtained to obtain the comprehensive quantitative score of the multi-tree rule. According to the comprehensive quantitative score of the multi-tree rule, it is determined whether it needs to be re-optimized. If the comprehensive quantitative score is lower than the preset threshold, it is determined that the multi-tree rule needs to be optimized; otherwise, it is determined that the current multi-tree rule does not need to be adjusted. Through the above steps, the final optimization decision is obtained to ensure that the multi-tree rule reaches the best state in multiple dimensions such as structural complexity, business coverage and redundancy.

[0079] S104, obtaining historical advertisement delivery data, and performing an advertisement delivery scenario simulation test on the multi-branch tree rule according to the historical advertisement delivery data.

[0080] In this embodiment, in the game advertising delivery system, in order to verify the effectiveness and applicability of the multi-branch tree rules, the system needs to obtain historical advertising delivery data and perform advertising delivery scenario simulation tests on the multi-branch tree rules based on these data. Through simulation tests, the performance of the rules in different scenarios can be evaluated, potential problems can be discovered and optimized.

[0081] Specifically, the system retrieves historical advertising data from the database. This data should include key information such as advertising delivery time, channel, target user group, advertising content, delivery results (such as click-through rate, conversion rate, cost, etc.). The acquired data is cleaned and preprocessed to remove invalid or abnormal data to ensure the accuracy and completeness of the data. An advertising delivery scenario simulation framework is designed, which should be able to simulate different scenarios such as user behavior, market environment, and delivery conditions. The multi-branch tree rule is used as the decision engine of the simulation framework to simulate the advertising delivery process based on the logical relationship and conditional judgment in the rules. During the simulation process, historical advertising data is used as input to simulate the advertising display and click behavior of different user groups at different times and channels.

[0082] Collect the data generated during the simulation test, including the simulated delivery results (such as simulated click-through rate, simulated conversion rate, etc.) and the actual delivery results (obtained from historical data). Evaluate the simulation results, compare the differences between the simulation results and the actual results, and analyze the reasons for the differences. Evaluate the performance of the multi-branch tree rules in different scenarios and identify potential problems or deficiencies in the rules. Based on the results of the simulation test and the evaluation analysis, optimize the multi-branch tree rules, including adjusting logical judgments, modifying conditional parameters, adding or deleting rule nodes, etc. The optimized rules are simulated again to verify the optimization effect.

[0083] In this embodiment, through simulation testing, the effectiveness and applicability of the multi-branch tree rules in different scenarios can be verified to ensure that the rules can accurately guide advertising delivery. By conducting simulation testing before actual delivery, potential problems in the rules can be discovered and corrected in a timely manner, reducing the delivery risk caused by rule errors. The results of the simulation test can provide data support for optimizing the delivery strategy and help the system more accurately locate the target user group and delivery channels.

[0084] In some embodiments, in the above step S104, the advertising delivery scenario simulation test of the multi-branch tree rule according to the historical advertising delivery data specifically includes:

[0085] Performing an advertising delivery scenario simulation on the multi-branch tree rule according to the historical advertising delivery data, and simultaneously modeling the advertising delivery scenario simulation process using a Markov model to obtain an advertising delivery scenario simulation result;

[0086] According to the simulation results of the advertising delivery scenario, the AdaBoost algorithm is used to train the integrated decision tree classifier to obtain the delivery strategy optimization model.

[0087] In this embodiment, based on the historical advertising data, a multi-tree algorithm is used to model the advertising delivery scenario to obtain a multi-tree model; based on the multi-tree model, a Markov model is used to model the process of advertising delivery scenario simulation to obtain a Markov state transfer matrix; based on the Markov state transfer matrix, the advertising delivery scenario is simulated by a Monte Carlo simulation method to obtain an advertising delivery scenario simulation sequence; based on the advertising delivery scenario simulation sequence, a statistical analysis method is used to calculate the probability distribution of each advertising delivery scenario to obtain an advertising delivery scenario probability distribution result; based on the advertising delivery scenario probability distribution result, the delivery effect of each advertising delivery scenario is judged to obtain an advertising delivery scenario effect evaluation result.

[0088] According to the simulation results of the advertising delivery scenario, the historical data of advertising delivery is obtained, including attributes such as advertising content, delivery time, delivery channel, target audience, and effect indicators such as click-through rate and conversion rate of the advertisement, as the data set for training the AdaBoost model. The data preprocessing technology is used to clean, transform and feature engineer the acquired historical data of advertising delivery, remove noise data, convert categorical variables into numerical variables, and select effect indicators such as click-through rate and conversion rate as classification labels to obtain a data set suitable for AdaBoost algorithm training. The preprocessed advertising delivery data set is divided into a training set and a test set by the cross-validation method. The training set is used to train the AdaBoost model, and the test set is used to evaluate the generalization performance of the model. The AdaBoost algorithm is used to set a suitable base classifier (such as a decision tree). Through multiple rounds of iterative training, the training process of the base classifier is adjusted according to the weight of the sample in each iteration, and the sample weight is updated to obtain a series of base classifiers, which are combined to form the final strong classifier, that is, the delivery strategy optimization model. According to the training process of the AdaBoost algorithm, by integrating the prediction results of multiple weak classifiers, the prediction values ​​of the delivery strategy optimization model for the click-through rate, conversion rate and other effect indicators of the new advertising delivery scenario are obtained. The delivery strategy optimization model obtained by training is evaluated using the test set data, and the evaluation indicators such as the accuracy, precision, recall rate, and F1 value of the model are calculated to determine whether the performance of the model meets the needs of advertising delivery optimization. According to the model evaluation results, the delivery strategy optimization model is tuned, and the performance of the model is improved by adjusting the hyperparameters of the AdaBoost algorithm (such as the number of iterations, learning rate, etc.) and selecting a more appropriate base classifier, so as to obtain the optimal delivery strategy optimization model. The tuned delivery strategy optimization model is applied to the actual advertising delivery scenario. According to the prediction results of the model, the advertising delivery strategy is automatically adjusted, such as delivery time, delivery channel, target audience, etc., to obtain higher advertising click-through rate and conversion rate, and to achieve advertising delivery optimization.

[0089] S105, determining an advertisement placement scheme for each advertisement according to the multi-branch tree rule, and performing display analysis on advertisements that do not comply with the multi-branch tree rule.

[0090] In this embodiment, in the game advertisement delivery system, according to the established and verified multi-tree rules, the system needs to determine a suitable delivery scheme for each advertisement to be delivered. At the same time, the system also needs to identify and process those delivered advertisements that do not comply with the multi-tree rules for further analysis and optimization.

[0091] Specifically, the system traverses the multi-branch tree rules and determines a matching advertising plan for each ad to be placed based on the conditions and logical judgments of each node. The advertising plan includes key information such as the advertising time, distribution channel, target user group, budget allocation, etc. When determining the advertising plan, the system needs to comprehensively consider factors such as the content of the advertisement, the attributes of the target users, and changes in the market environment.

[0092] In the process of determining the advertising plan, the system also checks whether each advertisement to be delivered meets all the conditions of the multi-tree rules. If it is found that a certain advertisement cannot find a matching delivery path in the multi-tree rules or does not meet the conditions of a certain node, the system will mark the advertisement as a delivery advertisement that does not comply with the rules. For delivery advertisements that do not comply with the rules, the system collects them for display analysis. The display analysis includes the basic information of the advertisement (such as advertising ID, advertising content, target user group, etc.), the specific reasons for non-compliance with the rules (such as lack of necessary attributes, target user group does not match the rules, etc.), and possible solutions (such as modifying the advertising content, adjusting the target user group, etc.). The results of the display analysis can be used as a reference for advertising delivery personnel so that they can carry out subsequent optimization processing.

[0093] In this embodiment, by automatically determining the advertising plan based on the multi-branch tree rules, the system can quickly allocate appropriate delivery resources to each advertisement and improve delivery efficiency. Through strict rule verification and delivery plan determination process, the system can reduce delivery errors caused by human misjudgment or negligence and reduce the error rate. Display analysis of ads that do not meet the rules helps advertising personnel to promptly discover problems and deficiencies in ads, and perform targeted optimization to improve the quality of ads.

[0094] In some embodiments, in the above step S105, determining the advertisement placement scheme for each advertisement according to the multi-branch tree rule specifically includes:

[0095] Constructing an advertising placement decision tree model according to the multi-branch tree rule, and using a depth-first search algorithm to traverse the tree structure of the advertising placement decision tree model to obtain multiple advertising placement plans;

[0096] For each advertising plan, a comprehensive score of each advertising plan is calculated by setting multiple evaluation indicators;

[0097] Sorting all advertising delivery plans according to the comprehensive scores to determine a number of candidate advertising delivery plans;

[0098] A multi-objective optimization algorithm is used to screen and optimize several candidate advertising delivery plans to determine the target advertising delivery plan.

[0099] In this embodiment, a multi-branch tree data structure is used to represent the advertising delivery decision tree model. Each node of the tree represents a decision attribute or delivery plan, and the edges between the nodes represent decision conditions, so as to obtain a decision tree model structure that can meet the needs of various delivery scenarios. A depth-first search algorithm is used, starting from the root node of the decision tree, and recursively traversing each node to obtain an advertising delivery plan represented by the subtree with the current node as the root. During the traversal process, the direction is determined according to the decision attributes and conditions of the node until the leaf node is reached, and a complete advertising delivery execution plan is obtained. Through the depth-first search, all paths of the decision tree are traversed to obtain multiple advertising delivery plans covering various delivery scenarios. Each plan contains a series of decision attributes and their corresponding values, representing the optimal delivery execution strategy under specific audiences and delivery requirements, which can guide the actual advertising delivery process.

[0100] According to the target audience, delivery channel, delivery time, creative content and other attributes of the advertising plan, a multi-dimensional evaluation model is used to obtain the score of each plan on each evaluation indicator. The scores of each evaluation indicator are weighted and summed by the weighted average method to obtain the comprehensive score of each advertising plan. Among them, the weight of each indicator can be set and adjusted according to the needs of the advertiser and historical data. All advertising plans are sorted according to the comprehensive score, and several preferred plans with the highest scores are obtained as candidates.

[0101] By analyzing the business attributes of each advertising delivery plan, the objective function in the multi-objective optimization algorithm is used to determine the initial fitness value of each plan. According to the business attributes of advertising delivery, such as the budget, target audience, and delivery time, the initial fitness value is adjusted by using constraint conditions to obtain the adjusted fitness value. Through the adjusted fitness value, a selection operation is used to obtain an advertising delivery plan with a higher fitness value, and a set of advertising delivery plans after preliminary screening is obtained. According to the set of advertising delivery plans after preliminary screening, a crossover operation is used to obtain a new advertising delivery plan, and a set of advertising delivery plans after crossing is obtained. Through the set of advertising delivery plans after crossing, a mutation operation is used to obtain a mutated advertising delivery plan, and a set of mutated advertising delivery plans is obtained. According to the set of mutated advertising delivery plans, fitness evaluation is used to obtain the fitness value of each plan, and an evaluated set of advertising delivery plans is obtained. Through the evaluated set of advertising delivery plans, non-dominated sorting is used to obtain a non-dominated solution set, and an advertising delivery plan in the non-dominated solution set is obtained. According to the advertising delivery plans in the non-dominated solution set, congestion calculation is used to obtain the congestion value of each plan, and a set of advertising delivery plans after congestion sorting is obtained. The set of advertising delivery plans sorted by congestion is used to obtain the final advertising delivery plan by adopting the elite retention strategy, and the target advertising delivery plan is determined.

[0102] In some embodiments, in the above step S105, the display analysis of the delivered advertisements that do not conform to the multi-branch tree rule specifically includes:

[0103] Acquire several delivered advertisements that do not conform to the multi-branch tree rule, group each delivered advertisement using a K-means clustering algorithm, perform feature statistics on the delivered advertisements in each group, calculate support and confidence between different combinations of advertisement features, and obtain a first analysis result;

[0104] Analyzing the association relationship between a number of placed advertisements that do not conform to the multi-tree rule and a number of placed advertisements that conform to the multi-tree rule by using an association rule mining algorithm to obtain a second analysis result;

[0105] The multi-branch tree rule is optimized and improved according to the first analysis result and the second analysis result.

[0106] In this embodiment, an initial data set of these advertisements is obtained by obtaining a number of advertisements that do not conform to the multi-branch tree rule. The initial data set is grouped by using the K-means clustering algorithm to determine the group to which each advertisement belongs. Based on the advertisements in each group, feature data of these advertisements are obtained to obtain feature statistics of each group. The support and confidence between different combinations of advertisement features are determined by analyzing the feature statistics of each group. Based on the calculated support and confidence, a first analysis result is obtained to obtain an association rule for the combination of advertisement features.

[0107] According to the attribute information of the advertisements, the multi-tree rule is used to classify the advertisements, and a set A of advertisements that conform to the multi-tree rule and a set B of advertisements that do not conform to the multi-tree rule are obtained. By scanning the advertisements in the advertisement sets A and B, various attribute combinations of the advertisements are obtained, and a set of advertisement attribute combinations is obtained. According to the combination set of advertisement attributes, the Apriori algorithm is used to generate frequent item sets, and the frequent combination pattern of advertisement attributes is obtained. By calculating the support and confidence of the frequent item sets, the association rules that conform to the minimum support and minimum confidence thresholds are obtained, and the association relationship between the advertisement attributes is obtained. According to the association rules of the advertisement attributes, the correlation between the advertisements that do not conform to the multi-tree rule and the advertisements that conform to the multi-tree rule is determined, and the second analysis result is obtained to obtain the association pattern between the advertisements.

[0108] Through the first analysis result, determine which advertising feature combinations have higher support and confidence in advertising delivery, and obtain the specific features of these combinations. Based on these specific features, obtain advertising feature combinations that do not conform to the multi-tree rule, and determine the performance of these combinations in advertising delivery. By further analyzing the performance of these advertising feature combinations, determine their distribution in different groups, and obtain the feature distribution results of each group. According to the feature distribution results of each group, obtain advertising feature combinations of different groups, and determine the support and confidence of these combinations in each group. By comparing the support and confidence of each group, determine which groups of advertising feature combinations have higher correlation, and obtain the specific features of these groups. Based on these specific features, obtain the performance of advertising feature combinations that do not conform to the multi-tree rule in different groups, and determine their effects in advertising delivery. By analyzing these effects, determine which advertising feature combinations have higher delivery effects in different groups, and obtain the specific performance of these combinations. Based on these specific performances, obtain the delivery effects of advertising feature combinations that do not conform to the multi-tree rule in different groups, and determine their actual performance in advertising delivery. By summarizing these actual performances, we can determine which combinations of advertising features have higher delivery effects in different groups.

[0109] Based on the second analysis result, the association mode is used to analyze the advertisements that do not conform to the multi-tree rules, obtain the advertisements related to them that conform to the multi-tree rules, and obtain the optimized advertising recommendation set. By analyzing the attribute distribution and delivery effect of the advertising recommendation set, the optimized delivery strategy is obtained, and the method of improving the accuracy and conversion rate of advertising delivery is determined.

[0110] Reference Figure 2 An embodiment of the present invention provides a processing system 2 for an automated rule configuration process, and the system 2 specifically includes:

[0111] The first processing module 201 is used to obtain first delivery rule data, perform semantic analysis and structure extraction on the first delivery rule data, and form second delivery rule data;

[0112] The second processing module 202 is used to establish a standard indicator library, and to construct an indicator system for the second delivery rule data according to the standard indicator library to generate a multi-branch tree rule;

[0113] The third processing module 203 is used to automatically check the multi-tree rule according to the rule checking tool to determine whether the multi-tree rule is reasonable, and if it is unreasonable, optimize the multi-tree rule;

[0114] The fourth processing module 204 is used to obtain historical advertisement delivery data, and perform advertisement delivery scenario simulation test on the multi-branch tree rule according to the historical advertisement delivery data;

[0115] The fifth processing module 205 is used to determine the advertising scheme for each advertisement according to the multi-branch tree rule, and to perform display analysis on the advertisements that do not comply with the multi-branch tree rule.

[0116] It is understandable that if Figure 1 The contents of the processing method embodiment of the automated rule configuration process shown in the figure are applicable to the processing system embodiment of the automated rule configuration process. The functions specifically implemented by the processing system embodiment of the automated rule configuration process are similar to those of the embodiment of the automated rule configuration process. Figure 1 The processing method embodiment of the automated rule configuration process shown in FIG. 1 is the same as that of FIG. 1 , and the beneficial effects achieved are the same as those of FIG. Figure 1 The beneficial effects achieved by the processing method embodiment of the automated rule configuration process shown are also the same.

[0117] It should be noted that the information interaction, execution process and other contents between the above-mentioned systems are based on the same concept as the embodiment of the method of the present invention. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0118] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0119] Reference Figure 2 The embodiment of the present invention further provides a computer device 3, comprising: a memory 302 and a processor 301 and a computer program 303 stored in the memory 302. When the computer program 303 is executed on the processor 301, a processing method for an automated rule configuration process as described in any one of the above methods is implemented.

[0120] The computer device 3 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The computer device 3 may include, but is not limited to, a processor 301 and a memory 302. Those skilled in the art will appreciate that Figure 2 It is only an example of computer device 3 and does not constitute a limitation on computer device 3. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components, for example, it may also include input and output devices, network access devices, etc.

[0121] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0122] In some embodiments, the memory 302 may be an internal storage unit of the computer device 3, such as a hard disk or memory of the computer device 3. In other embodiments, the memory 302 may also be an external storage device of the computer device 3, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the computer device 3. Further, the memory 302 may also include both an internal storage unit of the computer device 3 and an external storage device. The memory 302 is used to store an operating system, an application program, a boot loader (Boot Loader), data, and other programs, such as the program code of the computer program, etc. The memory 302 may also be used to temporarily store data that has been output or is to be output.

[0123] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the processing method of the automation rule configuration process as described in any one of the above methods is implemented.

[0124] In this embodiment, if the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the camera / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0125] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0126] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0127] In the embodiments disclosed in the present application, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0128] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

Claims

1. A method for processing an automated rule configuration process, characterized in that: The method specifically comprises: Acquire first delivery rule data, perform semantic analysis and structured extraction on the first delivery rule data, and form second delivery rule data; Establishing a standard indicator library, and constructing an indicator system for the second delivery rule data according to the standard indicator library to generate a multi-branch tree rule; Automatically verify the multi-branch tree rule according to the rule verification tool to determine whether the multi-branch tree rule is reasonable, and if it is unreasonable, optimize the multi-branch tree rule; Acquire historical advertisement delivery data, and perform advertisement delivery scenario simulation test on the multi-branch tree rule according to the historical advertisement delivery data; An advertisement delivery scheme for each advertisement is determined according to the multi-branch tree rule, and display analysis is performed on advertisements that do not comply with the multi-branch tree rule.

2. The method according to claim 1, characterized in that The performing semantic analysis and structural extraction on the first delivery rule data to form second delivery rule data specifically includes: Performing semantic analysis on the first delivery rule data using natural language processing technology to determine key elements in the first delivery rule data and obtain third delivery rule data; Extracting the core semantic information of the third delivery rule data, constructing a knowledge graph through the core semantic information, and vectorizing the knowledge graph using a graph embedding algorithm to obtain a low-dimensional dense vector representation; According to the low-dimensional dense vector representation, a clustering algorithm is used to perform semantic similarity calculation and cluster analysis on the third delivery rule data to generate a plurality of rule clusters; An association rule mining algorithm is used to identify association relationships and dependency relationships between multiple rule clusters, and second delivery rule data is formed according to the association relationships and dependency relationships.

3. The method according to claim 2, characterized in that The step of constructing an indicator system for the second delivery rule data according to the standard indicator library to generate a multi-branch tree rule specifically includes: Using the hierarchical analysis method to classify and organize the second delivery rule data, and obtain dimension information and attribute information of the second delivery rule data; According to the dimension information and attribute information, a multi-level indicator system is constructed by using a decision tree algorithm, and indicator weights are assigned to indicators at each level of the multi-level indicator system by using a fuzzy comprehensive evaluation method; According to the standard indicator library, the naming of indicators at each level in the multi-level indicator system is standardized and revised to form a multi-branch tree rule.

4. The method according to claim 1, characterized in that The method of automatically checking the polytree rule according to the rule checking tool to determine whether the polytree rule is reasonable, and optimizing the polytree rule if it is unreasonable, specifically includes: Using a depth-first search algorithm to traverse the multi-branch tree rule, obtaining attribute information and child node relationships of each node of the multi-branch tree rule; Based on the attribute information of each node and the relationship between its sub-nodes, determine whether the attribute values ​​of each node meet the predefined constraints, and whether the relationship between nodes meets the preset business logic requirements; If the attribute value of each node does not meet the predefined constraint conditions or the relationship between the nodes does not meet the preset business logic requirements, a heuristic search algorithm is used to locally adjust the multi-branch tree rules; The adjusted multi-branch tree rules are evaluated by using the hierarchical analysis method, and quantitative scoring is performed from multiple dimensions such as structural complexity, business coverage, and redundancy to obtain quantitative scoring results. Based on the quantitative scoring results, it is determined whether the multi-branch tree rules should be re-optimized.

5. The method according to claim 1, characterized in that The performing of an advertisement delivery scenario simulation test on the multi-branch tree rule according to the historical advertisement delivery data specifically includes: Performing an advertising delivery scenario simulation on the multi-branch tree rule according to the historical advertising delivery data, and simultaneously modeling the advertising delivery scenario simulation process using a Markov model to obtain an advertising delivery scenario simulation result; According to the simulation results of the advertising delivery scenario, the AdaBoost algorithm is used to train the integrated decision tree classifier to obtain the delivery strategy optimization model.

6. The method according to claim 1, characterized in that The step of determining the advertisement placement scheme for each advertisement placement according to the multi-branch tree rule specifically includes: Constructing an advertising placement decision tree model according to the multi-branch tree rule, and using a depth-first search algorithm to traverse the tree structure of the advertising placement decision tree model to obtain multiple advertising placement plans; For each advertising plan, a comprehensive score of each advertising plan is calculated by setting multiple evaluation indicators; Sorting all advertising delivery plans according to the comprehensive scores to determine a number of candidate advertising delivery plans; A multi-objective optimization algorithm is used to screen and optimize several candidate advertising delivery plans to determine the target advertising delivery plan.

7. The method according to claim 6, characterized in that The display analysis of the advertisements not meeting the multi-tree rule specifically includes: Acquire several delivered advertisements that do not conform to the multi-branch tree rule, group each delivered advertisement using a K-means clustering algorithm, perform feature statistics on the delivered advertisements in each group, calculate support and confidence between different combinations of advertisement features, and obtain a first analysis result; Analyzing the association relationship between a number of placed advertisements that do not conform to the multi-tree rule and a number of placed advertisements that conform to the multi-tree rule by using an association rule mining algorithm to obtain a second analysis result; The multi-branch tree rule is optimized and improved according to the first analysis result and the second analysis result.

8. A processing system for automated rule configuration process, characterized in that: The system specifically comprises: A first processing module is used to obtain first delivery rule data, perform semantic analysis and structured extraction on the first delivery rule data, and form second delivery rule data; The second processing module is used to establish a standard indicator library, and to construct an indicator system for the second delivery rule data according to the standard indicator library to generate a multi-branch tree rule; A third processing module is used to automatically check the multi-tree rule according to a rule checking tool to determine whether the multi-tree rule is reasonable, and if it is unreasonable, optimize the multi-tree rule; A fourth processing module is used to obtain historical advertising data, and perform advertising delivery scenario simulation test on the multi-branch tree rule according to the historical advertising data; The fifth processing module is used to determine the advertising scheme for each advertisement according to the multi-branch tree rule, and to perform display analysis on the advertisements that do not comply with the multi-branch tree rule.

9. A computer device, characterized in that: include: A memory, a processor and a computer program stored in the memory, when the computer program is executed on the processor, implements the processing method of the automation rule configuration process as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the processing method of the automation rule configuration process according to any one of claims 1 to 7 is implemented.

Images