A self-service billing file physical examination method, system, medium and device

Abstract

The present invention belongs to the field of electricity fee accounting, and provides a self-service billing file physical examination method, system, medium and equipment. Its technical solution is to establish a file verification rule library and a policy library by sorting out the marketing file data rules, the anomalies affecting the billing files, and other issues. Business personnel use the file physical examination tool regularly or on demand, and conduct a one-click self-service file physical examination and verification, output an abnormality detail list, calculate the file physical examination score, and generate a physical examination report. At the same time, based on the policy library, it can provide business personnel with standardized processing procedures and operation steps to guide when repairing file defects, thereby improving the one-time success rate of file repair, and at the same time improving the file quality, the billing quality, and the accounting processing efficiency. It solves the problem that the original algorithm is no longer sufficient to support the rapid and accurate detection of power data anomalies.

Description

Technical Field

[0001] The invention belongs to the field of electricity fee accounting, and in particular relates to a self-service billing file physical examination method, system, medium and equipment. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Traditional billing file inspection methods are mainly based on rules and statistical models, which determine whether the data is abnormal by setting thresholds or using statistical algorithms. Although this method is relatively simple and direct, it is difficult to capture complex abnormal patterns. In recent years, with the rapid development of natural language processing, knowledge graphs, and association analysis technologies, more and more studies have begun to apply them to the field of power data anomaly detection. However, as the volume of power system data continues to increase and the database becomes larger and larger, the original algorithm is no longer sufficient to support the rapid and accurate detection of power data anomalies. Summary of the invention

[0004] In order to solve at least one technical problem existing in the above-mentioned background technology, the present invention provides a self-service billing file physical examination method, system, medium and equipment, which establishes a file verification rule library and a policy library by sorting out the marketing file data rules, the abnormalities of the billing files, and other issues. Business personnel use the file physical examination tool regularly or on demand, and conduct a one-click self-service file physical examination and verification, output a detailed list of abnormalities, calculate the file physical examination score, and generate a physical examination report. At the same time, based on the policy library, it can provide business personnel with standardized processing procedures and operation steps to guide when repairing file defects, thereby improving the one-time success rate of file repair, and at the same time, it can also improve the file quality, billing quality, and improve the efficiency of accounting processing.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] A first aspect of the present invention provides a self-service billing file physical examination method, comprising the following steps:

[0007] Obtain business data during electricity billing;

[0008] Convert the business data obtained in the electricity fee calculation process into text data;

[0009] Combine text data and billing archive knowledge graph to build an automated verification rule base and policy base; the specific construction process includes:

[0010] Based on text data and billing file knowledge graph, billing file entities and the relationships between entities are extracted to build a graph database;

[0011] Based on the graph database, the association rule analysis method is used to obtain the initial association rules;

[0012] Based on the initial association rules, the embedded representation of nodes and edges is learned on the knowledge graph, so that the metrological archive entities with similar relationships are constantly close in space, and an automated verification rule base is generated; in the interaction between the verification rule generation and the knowledge graph, a reinforcement learning framework is introduced to learn the processing strategies for abnormal situation matching corresponding to the verification rules in different scenarios, and generate a verification strategy library;

[0013] Based on the automated verification rule library, the billing files to be verified are checked item by item, the file physical examination results are generated, and the corresponding processing strategies are matched for the abnormal detection results based on the verification strategy library.

[0014] Furthermore, based on the text data and the billing file knowledge graph, the billing file entities and the relationships between the entities are extracted to construct a graph data structure, including:

[0015] Extract billing file entities from text data, associate or create new entities based on existing entities in the billing file knowledge graph; based on the extracted billing file entities, identify and extract the relationships between each entity, and construct a graph data structure based on the extracted billing file entities and the relationships between each entity.

[0016] Furthermore, the initial association rules are obtained by using an association rule analysis method based on a graph database, including:

[0017] The billing archive data is formed into a collection of itemsets as a transaction database;

[0018] Based on the transaction database, find out the combination of items that frequently appear in the billing archive data and obtain the frequent item sets;

[0019] Based on frequent item sets, the association strength between these item sets is analyzed through association rules, and the rules with association strength greater than the set threshold are used as initial association rules.

[0020] Furthermore, after learning the embedded representations of nodes and edges on the knowledge graph, the embedded representations learned on the knowledge graph are used as features. Combined with the inductive logic programming method, the embeddings of nodes and edges are used as heuristic information by searching the hypothesis space, interacting with background knowledge and positive and negative examples to automatically generate logical rules.

[0021] Furthermore, after obtaining the automated verification rule base, verification rule maintenance is performed, including adding rule application descriptions and rule corresponding attachment descriptions, compiling the approved verification rules, obtaining the compiled rules, and generating entity physical examination rules based on the compiled rules.

[0022] Furthermore, the business data in the electricity fee calculation process includes business data that affects abnormal fee calculation problems and key business parameters that affect the accuracy of electricity price execution.

[0023] Furthermore, users who do not meet the verification rules of the automated verification rule base are added to the whitelist. During the validity period of the whitelist, user objects are excluded during rule verification and no abnormal data is generated. The whitelist supports setting an effective period and will automatically become invalid after the expiration. When verifying the invalid object user, it will be executed as a normal user not added to the whitelist, which may generate an exception.

[0024] A second aspect of the present invention provides a self-service billing file physical examination system, comprising:

[0025] A business data acquisition module, which is used to acquire business data in the electricity fee calculation process;

[0026] A business data preprocessing module, which is used to convert the business data obtained in the electricity fee calculation process into text data;

[0027] The verification rule generation module is used to build an automated verification rule base and policy base by combining text data and billing archive knowledge graph. The specific construction process includes:

[0028] Based on text data and billing file knowledge graph, billing file entities and the relationships between entities are extracted to build a graph database;

[0029] Based on the graph database, the association rule analysis method is used to obtain the initial association rules;

[0030] Based on the initial association rules, the embedded representation of nodes and edges is learned on the knowledge graph, so that the metrological archive entities with similar relationships are constantly close in space, and an automated verification rule base is generated;

[0031] In the interaction between verification rule generation and knowledge graph, a reinforcement learning framework is introduced to learn the processing strategies for abnormal situations corresponding to verification rules in different scenarios and generate a verification strategy library;

[0032] The file physical examination module is used to verify the file physical examination execution tasks item by item based on the automated verification rule library, generate the file physical examination results, and match the corresponding processing strategies for the abnormal detection results based on the verification strategy library.

[0033] A third aspect of the present invention provides a computer-readable storage medium.

[0034] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in the self-service billing file physical examination method as described above.

[0035] A fourth aspect of the present invention provides a computer device.

[0036] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the self-service billing file physical examination method as described above are implemented.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] The present invention establishes an archive verification rule library and a policy library by sorting out the marketing archive data rules, issues affecting the abnormalities of the billing archives, etc. Business personnel use the archive physical examination tool regularly or on demand, conduct a one-click self-service archive physical examination and verification, output an abnormality detail list, calculate the archive physical examination score, and generate a physical examination report. At the same time, based on the policy library, it can provide business personnel with standardized processing procedures and operation steps to guide when repairing archive defects, thereby improving the success rate of archive repair in one go, and at the same time, it can also improve the archive quality, billing quality, and improve the efficiency of accounting processing.

[0039] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0041] Figure 1 The present invention provides a flowchart of a self-service billing file physical examination method. DETAILED DESCRIPTION

[0042] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0043] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0044] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0045] In view of the defect that the original algorithm is no longer sufficient to support the rapid and accurate detection of power data anomalies, the present invention establishes an archive verification rule library and a policy library by sorting out the marketing archive data rules, the anomalies that affect the billing archives, and other issues. Business personnel use the archive physical examination tool regularly or on demand, and conduct a one-click self-service archive physical examination and verification, output an abnormality detail list, calculate the archive physical examination score, and generate a physical examination report. At the same time, based on the policy library, it can provide business personnel with standardized processing procedures and operation steps to guide when repairing archive defects, thereby improving the one-time success rate of archive repair, while also improving the archive quality and billing quality, and improving the efficiency of accounting processing.

[0046] Based on the NLP principle, the present invention can extract key information such as entities, relationships and events from unstructured or semi-structured texts as the basis for automatic rule generation. With the knowledge graph, the information extracted by NLP can be organized in the form of nodes (entities) and edges (relationships), forming a knowledge structure that is easy for machines to understand and reason, and further provides a clear and standardized framework for the automatic synthesis of rules. Through entity linking and ontology mapping, the consistency and accuracy of the subsequent rule base can be ensured. Combined with machine learning and deep learning technology, focusing on the GNNS graph neural network, it continuously generates more abstract and generalized rules from the embedded representation of the knowledge graph to ensure that the model can learn the implicit features of nodes and edges, and then automatically construct rules that can cover a wide range of metering archive physical examination situations, enabling the improvement of the quality and efficiency of power data anomaly detection.

[0047] Embodiment 1

[0048] This embodiment provides a self-service billing file physical examination method, including the following steps:

[0049] Step 1: Obtain business data in the electricity fee calculation process;

[0050] In this embodiment, the business data in the electricity fee calculation process includes common and frequent abnormal charging problems such as the quality of collected meter codes, billing parameters, change vouchers, and key business parameters that affect the accuracy of electricity price execution.

[0051] Step 2: Convert the business data obtained in the electricity fee calculation process into text data;

[0052] After obtaining the business data in the electricity billing process, the initial stage of structured data and semi-structured data collection and unstructured text to data conversion operations are completed through sorting and summarizing operations.

[0053] Step 3: Combine text data and billing archive knowledge graph to build an automated verification rule base and policy base, which includes the following steps:

[0054] Step 301: extract entities from the text using named entity recognition (NER) technology, and associate them with existing entities in the billing profile knowledge graph through entity linking or create new entities based on them;

[0055] Step 302: Based on the extracted relational entities, identify and extract the relationships between the entities, organize the constructed entities and relationships into corresponding graph data structures, and store them in the corresponding graph database or triple storage system as the basis for subsequent rule association analysis.

[0056] Step 303: Based on the graph data structure, an association rule analysis method is used to obtain initial association rules;

[0057] In this embodiment, knowledge graph and pattern learning are innovatively integrated. Based on the strong scalability principle of knowledge graph, frequent subgraph mining algorithm is used to sort out and analyze the frequently appearing patterns in the billing archive knowledge graph, and gain insights into potential rules or structural laws.

[0058] The goal of using association rule analysis algorithms is to discover the connections between different items in the transaction database. The rules formed by these connections can assist in exploring the characteristics of certain behaviors. It mainly includes transaction database construction, frequent item set (frequent pattern in knowledge graph) mining, association rule analysis, etc.

[0059] The transaction database is designed to extract and organize relevant billing archive data into a collection of itemsets;

[0060] Frequent itemsets aim to find out the combinations of items that appear frequently in billing archive data;

[0061] Based on frequent item sets, association rules further analyze whether there is a strong correlation between these item sets.

[0062] (1) Transactional database

[0063] definition A collection of global items, where Is the unique identifier of the item. Transactional database is a collection of transactions, each transaction All correspond A set consisting of some or all of the items is an itemset, and any non-empty itemset does not contain duplicate items.

[0064] (2) Association rules

[0065] Association rules represent the relationship between items, such as: , which become the antecedent and subsequent of the rule respectively. Support measures the frequency of an item set (whether a single item or a combination of multiple items) appearing in all transactions. It is expressed as shown in the following formula: Things included in Percentage of:

[0066] ,

[0067] in For each transaction, is an item set, A collection of transactions.

[0068] If a transaction already contains the antecedent (X) of the rule, then the probability that the transaction also contains the consequent (Y) can be expressed as It is represented as follows:

[0069] ,

[0070] in For each transaction, is an item set, A collection of transactions.

[0071] If a transaction already contains the antecedent (X) of the rule, then the probability that the transaction also contains the consequent (Y) can be expressed as It is represented as follows:

[0072] ,

[0073] The numerator is the probability of events X and Y occurring simultaneously, and the denominator is the probability of event X occurring.

[0074] The above minimum support and minimum confidence are the minimum support threshold and the minimum confidence threshold respectively. The association rule that satisfies both of them becomes a strong rule.

[0075] (3) Frequent itemsets

[0076] Given a global itemset and a transaction database, for a non-empty itemset of the former, if its support is greater than or equal to the minimum support threshold, it is called a frequent itemset; if it contains k items, it is called a frequent k-itemset.

[0077] Step 304: Based on the initial association rules, combined with machine learning and deep learning technologies, further learning is performed to generate an automated verification rule base and a corresponding verification strategy base, which specifically includes the following steps:

[0078] Step 3041: Based on the above association rule analysis and combing results, the existing rules and patterns are integrated. First, the graph neural network model GNNS in machine learning is used to learn the embedded representation of nodes and edges on the knowledge graph. GNNS adopts the information propagation mechanism to integrate the features of neighboring nodes and generate an embedding vector for each node that can capture its structural features and semantic context. This process not only considers the attributes of the node itself, but also fully integrates the complex relationships between nodes, laying the foundation for subsequent rule generation.

[0079] For example, by training the GNN model, metrology archival entities can be embedded into a vector space, so that metrology archival entities with similar relationships are constantly close in space.

[0080] Step 3042: Next, in combination with the Inductive Logic Programming (ILP) method, the embedded representation obtained from the GNNs is used as a feature, interacting with background knowledge and positive and negative examples to automatically generate logical rules.

[0081] ILP searches the possible hypothesis space to identify logical rules that can explain the observed data and generalize to unseen instances. In this process, the embedding of nodes and edges can serve as heuristic information to guide the ILP algorithm to explore the rule space more efficiently and generate new rules that are consistent with data reality, business reality, and have the ability to generalize and evolve.

[0082] For example, based on the embedding similarity of metering file entities, ILP can learn rules such as "low-voltage purchasing lacks time-sharing meters" and "low-voltage purchasing only has peak and valley electricity" both require "metering file standardization verification".

[0083] Step 3043: Introduce a reinforcement learning framework so that the rule generation model interacts with the underlying knowledge graph through continuous trial and error and feedback loops. Reinforcement learning prompts the automated verification rule base to master how to most effectively utilize and expand the rule set and rule base of the knowledge graph in different scenarios, thereby more accurately matching the corresponding strategies, and continuously iterating and integrating based on the strategy generation results to finally generate the corresponding strategy library.

[0084] Some of the rules expanded by the automation rule base are shown in Table 1 below:

[0085] Table 1 Partial verification rules

[0086]

[0087] Table 1 Partial Verification Rules

[0088]

[0089] Step 4: Maintain the verification rules, including:

[0090] Support provincial and local units to provide optimization suggestions through online processes, and achieve closed-loop management of the entire process including verification rule application, approval, script writing, and rule publishing.

[0091] Rule application: Business personnel at all levels of units add rule application descriptions, corresponding rule attachments, and other content to initiate the rule application process.

[0092] Rule application approval: The rule verification management personnel receive rule applications from units at all levels and conduct approval confirmation. Those that fail to pass will be returned to the application stage, and those that pass will be sent to the rule compilation stage.

[0093] Rule content preparation: According to the rule application description and attachments, perform operations such as adding new rules, modifying rules (upgrading the original rule version), and deactivating rules to complete the rule application confirmation goal.

[0094] Release of rule content: After completing the compiled rules and confirming that they are correct, execute the release operation to generate actual physical examination rules, or update the corresponding time status of the original rules.

[0095] Step 5: Based on the automated verification rule library, the file physical examination tasks to be verified are verified item by item, the file physical examination results are generated, and the corresponding processing strategies are matched for the abnormal detection results based on the verification strategy library.

[0096] On-demand physical examination. Business personnel can retrieve and execute applicable rules in real time with one click, conduct a comprehensive physical examination of the billing files of the unit and subordinate units, calculate the physical examination scores and generate physical examination reports, and carry out file defect elimination management with reference to the disposal strategy. Based on the automated verification rule library, the physical examination execution tasks of the files to be verified are checked item by item, and the file physical examination results are generated. Based on the verification strategy library, the corresponding processing strategy is matched for the abnormal detection results.

[0097] Regular physical examinations. Business personnel can define the execution time and frequency of verification rules based on their own work conditions and business characteristics. The system automatically performs verification operations at regular intervals, generates verification reports, and reminds business personnel to handle them.

[0098] In this embodiment, for suspected abnormalities found out, the verification rules and users that are confirmed to be normal after verification and analysis can be added to the whitelist, and will be automatically removed during the next physical examination. At the same time, the situation where the physical examination is incomplete due to accumulation of whitelists can be avoided, and a dynamic cancellation mechanism for the whitelist can be established, that is, an effective period is set for the whitelist, which will no longer be effective after the expiration.

[0099] In this embodiment, the number of physical examinations, the number of abnormal problems found in each physical examination, the number of problems repaired, etc. are recorded according to the login account; when the rule library is upgraded or the login account has not performed a physical examination operation for a long time, the system will actively pop up a reminder to guide the user to actively check.

[0100] Embodiment 2

[0101] This embodiment provides a self-service billing file physical examination system, including:

[0102] A business data acquisition module, which is used to acquire business data in the electricity fee calculation process;

[0103] A business data preprocessing module, which is used to convert the business data obtained in the electricity fee calculation process into text data;

[0104] The verification rule generation module is used to build an automated verification rule base and policy base by combining text data and billing archive knowledge graph. The specific construction process includes:

[0105] Based on text data and billing file knowledge graph, billing file entities and the relationships between entities are extracted to build a graph database;

[0106] Based on the graph database, the association rule analysis method is used to obtain the initial association rules;

[0107] Based on the initial association rules, the embedded representation of nodes and edges is learned on the knowledge graph, so that the metrological archive entities with similar relationships are constantly close in space, and an automated verification rule base is generated;

[0108] In the interaction between verification rule generation and knowledge graph, a reinforcement learning framework is introduced to learn the processing strategies for abnormal situations corresponding to verification rules in different scenarios and generate a verification strategy library;

[0109] The file physical examination module is used to verify the file physical examination execution tasks item by item based on the automated verification rule library, generate the file physical examination results, and match the corresponding processing strategies for the abnormal detection results based on the verification strategy library.

[0110] Embodiment 3

[0111] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps in the self-service billing file physical examination method as described above are implemented.

[0112] Embodiment 4

[0113] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps in the self-service billing file physical examination method described above are implemented.

[0114] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A self-service billing file physical examination method, It is characterized in that The steps include: Obtain business data during electricity bill calculation; Convert the business data obtained in the electricity fee calculation process into text data; Combine text data and billing archive knowledge graph to build an automated verification rule base and policy base; the specific construction process includes: Based on the text data and the billing file knowledge graph, the billing file entities and the relationships between the entities are extracted to construct a graph database, including: extracting the billing file entities from the text data, associating or creating new entities according to the existing entities in the billing file knowledge graph; based on the extracted billing file entities, identifying and extracting the relationships between the entities, and constructing a graph data structure based on the extracted billing file entities and the relationships between the entities; Based on the graph database, the association rule analysis method is used to obtain the initial association rules, including: The billing archive data is formed into a collection of itemsets as a transaction database; Based on the transaction database, find out the combination of items that frequently appear in the billing archive data and obtain the frequent item sets; Based on frequent item sets, the association strength between these item sets is analyzed through association rules, and the rules with association strength greater than the set threshold are used as initial association rules; Based on the initial association rules, the embedded representation of nodes and edges is learned on the knowledge graph, so that the metrological archive entities with similar relationships are constantly close in space, and an automated verification rule base is generated; in the interaction between the verification rule generation and the knowledge graph, a reinforcement learning framework is introduced to learn the processing strategies for abnormal situation matching corresponding to the verification rules in different scenarios, and generate a verification strategy library; Based on the automated verification rule library, the billing files to be verified are checked item by item, the file physical examination results are generated, and the corresponding processing strategies are matched for the abnormal detection results based on the verification strategy library.

2. A self-service billing file physical examination method as claimed in claim 1, It is characterized in that After learning the embedded representation of nodes and edges on the knowledge graph, the embedded representation learned on the knowledge graph is used as a feature. Combined with the inductive logic programming method, the embedding of nodes and edges is used as heuristic information by searching the hypothesis space, interacting with background knowledge and positive and negative examples to automatically generate logical rules.

3. A self-service billing file physical examination method as claimed in claim 1, It is characterized in that After obtaining the automated verification rule base, verification rule maintenance is performed, including adding rule application descriptions and rule corresponding attachment descriptions, compiling the approved verification rules, obtaining the compiled rules, and generating entity physical examination rules based on the compiled rules.

4. A self-service billing file physical examination method as claimed in claim 1, It is characterized in that The business data in the electricity fee calculation process includes business data that affects abnormal fee calculation problems and key business parameters that affect the accuracy of electricity price execution.

5. A self-service billing file physical examination method as claimed in claim 1, It is characterized in that For users who do not meet the verification rules of the automated verification rule base, they are added to the whitelist. During the validity period of the whitelist, user objects are excluded during rule verification and no abnormal data is generated. The whitelist supports setting an effective period and will automatically become invalid after the expiration. When the invalid object user is verified, it will be executed as a normal user not added to the whitelist, which may generate an exception.

6. A self-service billing file physical examination system, It is characterized in that include: A business data acquisition module, which is used to acquire business data in the electricity fee calculation process; A business data preprocessing module, which is used to convert the business data obtained in the electricity fee calculation process into text data; The verification rule generation module is used to build an automated verification rule base and policy base by combining text data and billing archive knowledge graph. The specific construction process includes: Based on the text data and the billing file knowledge graph, the billing file entities and the relationships between the entities are extracted to construct a graph database, including: extracting the billing file entities from the text data, associating or creating new entities according to the existing entities in the billing file knowledge graph; based on the extracted billing file entities, identifying and extracting the relationships between the entities, and constructing a graph data structure based on the extracted billing file entities and the relationships between the entities; Based on the graph database, the association rule analysis method is used to obtain the initial association rules, including: The billing archive data is formed into a collection of itemsets as a transaction database; Based on the transaction database, find out the combination of items that frequently appear in the billing archive data and obtain the frequent item sets; Based on frequent item sets, the association strength between these item sets is analyzed through association rules, and the rules with association strength greater than the set threshold are used as initial association rules; based on the initial association rules, the embedded representation of nodes and edges is learned on the knowledge graph, so that the metrological archive entities with similar relationships are constantly close in space, and an automated verification rule base is generated; In the interaction between verification rule generation and knowledge graph, a reinforcement learning framework is introduced to learn the processing strategies for abnormal situations corresponding to verification rules in different scenarios and generate a verification strategy library; The file physical examination module is used to verify the file physical examination execution tasks item by item based on the automated verification rule library, generate the file physical examination results, and match the corresponding processing strategies for the abnormal detection results based on the verification strategy library.

7. A self-service billing file physical examination system as claimed in claim 6, It is characterized in that After learning the embedded representation of nodes and edges on the knowledge graph, the embedded representation learned on the knowledge graph is used as a feature. Combined with the inductive logic programming method, the embedding of nodes and edges is used as heuristic information by searching the hypothesis space, interacting with background knowledge and positive and negative examples to automatically generate logical rules.

8. A self-service billing file physical examination system as claimed in claim 6, It is characterized in that After obtaining the automated verification rule base, verification rule maintenance is performed, including adding rule application descriptions and rule corresponding attachment descriptions, compiling the approved verification rules, obtaining the compiled rules, and generating entity physical examination rules based on the compiled rules.

9. A self-service billing file physical examination system as claimed in claim 6, It is characterized in that The business data in the electricity fee calculation process includes business data that affects abnormal fee calculation problems and key business parameters that affect the accuracy of electricity price execution.

10. A self-service billing file physical examination system as claimed in claim 6, It is characterized in that For users who do not meet the verification rules of the automated verification rule base, they are added to the whitelist. During the validity period of the whitelist, user objects are excluded during rule verification and no abnormal data is generated. The whitelist supports setting an effective period and will automatically become invalid after the expiration. When the invalid object user is verified, it will be executed as a normal user not added to the whitelist, which may generate an exception.

11. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the program is executed by a processor, the steps in a self-service billing file physical examination method as described in any one of claims 1-5 are implemented.

12. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the program, the steps in the self-service billing file physical examination method as described in any one of claims 1-5 are implemented.

Images