Credit data collection rule identification method and system based on document numbers
Through the method and system for identifying credit data collection rules based on document numbers, natural language processing and machine learning technology are used to automatically identify and collect credit data, solving the problem of inefficiency of traditional methods and achieving efficient and accurate credit data collection and identification have increased the support for credit assessment.
Patent Information
- Application Number
- CN202411078350.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-07
- Publication Date
- 2024-12-03
AI Technical Summary
Traditional credit data collection methods are inefficient and error-prone, and cannot effectively manage and utilize diversified and inconsistent credit data.
Using the credit data collection rules identification method and system based on document number, the document number can be automatically extracted and identified through natural language processing, text mining, machine learning and other technologies, and the collection rules can be established and updated to realize the automation and standardized credit data collection. Set and recognize.
It improves the accuracy and reliability of credit data, reduces the manual operation error rate, and provides more accurate data support for credit evaluation and management.
Smart Images

Figure CN119067215A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of document number parsing and identification, rule matching and identification, and automated process management, and in particular to a method and system for identifying credit data collection rules based on document numbers. Background Art
[0002] As the credit system becomes increasingly perfect, the management and use of credit data is particularly important. Whether it is an individual or an enterprise, credit data has become an important basis for measuring their credit status. However, due to the diversity of data sources and non-uniform formats, traditional credit data collection methods are often inefficient and prone to errors. Therefore, a more intelligent and efficient method is needed to collect and identify credit data. Secondly, with the rapid development of big data and artificial intelligence technologies, new technical means are provided for the identification of credit data collection rules. By utilizing big data processing and analysis capabilities, it is possible to quickly collect and organize massive amounts of credit data. At the same time, with the help of artificial intelligence's machine learning, deep learning and other technologies, the collection rules in the document number can be automatically identified and analyzed, thereby achieving accurate identification and classification of credit data. Summary of the invention
[0003] In response to the problems in the related art, the present invention proposes a method and system for identifying credit data collection rules based on document numbers to overcome the above-mentioned technical problems existing in the existing related art.
[0004] To this end, the specific technical solution adopted by the present invention is as follows: A method and system for identifying credit data collection rules based on document numbers, comprising the following steps: Step S1: extract document number; Step S2: credit information identification; Step S3: Establishing aggregation rules; Step S4: rule base construction and matching; Step S5: Data collection and rule update.
[0005] As a further solution of the present invention, the step of obtaining the document number in step S1 specifically includes: Use the Named Entity Recognition (NER) algorithm in Natural Language Processing (NLP) technology to identify potential document numbers appearing in documents; Apply regular expressions (Regex) to pattern match the submission documents and extract the document numbers that conform to the specified format; The extracted documents are verified and deduplicated to ensure the uniqueness and accuracy of the document number.
[0006] As a further solution of the present invention, the step of credit information identification in step S2 specifically includes: Use text mining technology, such as TF-IDF.Word2Vec, to extract keywords and perform topic modeling on the submissions to identify credit-related text fragments; Combined with the domain knowledge base, a rule engine is built to identify key information in documents, such as amount, date, default record, etc. Apply machine learning algorithms (such as support vector machine SVM, random forest RF, etc.) to classify documents to distinguish data of different credit levels.
[0007] As a further solution of the present invention, the step of establishing the aggregation rule in step S3 specifically includes: Analyze historical data and business needs to determine the basic framework and classification standards of the collection rules; Use machine learning algorithms such as decision trees and logistic regression to learn from existing data and automatically generate or optimize aggregation rules; Combine manual review and expert knowledge to revise and improve automatically generated rules.
[0008] As a further solution of the present invention, the rule base construction and matching steps in step S4 specifically include: Establish a rule base containing all known submission book numbers and their corresponding collection rules, and structure the rule base to facilitate retrieval and query; On the basis of document number extraction and credit information identification, the rule engine is used for rule matching, which can be achieved through fuzzy matching, regular expression matching, etc. During the matching process, a similarity-based matching algorithm (such as cosine similarity, Jaccard similarity, etc.) can be used to calculate the similarity between the document number or credit information and the rules in the rule library to find the most matching collection rule.
[0009] As a further solution of the present invention, the steps of data collection and rule updating in step S5 specifically include: Once the submission guide successfully matches the collection rules, the system can trigger the data extraction and integration process to collect the relevant credit data to the designated location; For newly appeared document numbers or document numbers that cannot be matched, the system can automatically record and mark them as pending for further manual review and processing; Regularly update and maintain the collection rule base, including adding new rules, deleting or modifying old rules, to ensure the adaptability and accuracy of the system. This can be achieved by regularly running machine learning algorithms to optimize rules, or adjusting rules based on manual review results.
[0010] Beneficial Effects of the Invention By realizing automated and standardized credit data collection and processing, different types of document numbers can be identified. Collection rules are established based on the type of credit data, source, format, etc. corresponding to the document number. At the same time, new document numbers and collection rules are automatically identified, and the data collection rule library is automatically updated and maintained. The credit data collection rule identification method and system based on the document number can effectively improve the accuracy and reliability of credit data through automated and intelligent collection and identification processes, reduce the error rate of manual operations, and provide more accurate data support for credit assessment and management. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 A flowchart of a method for identifying credit data collection rules based on document numbers according to the present invention; DETAILED DESCRIPTION
[0012] According to an embodiment of the present invention, a method and system for identifying credit data aggregation rules based on document numbers are provided.
[0013] A method and system for identifying credit data collection rules based on document numbers according to an embodiment of the present invention includes the following steps: Step S1: extract document number; Step S2: credit information identification; Step S3: Establishing aggregation rules; Step S4: rule base construction and matching; Step S5: Data collection and rule update.
[0014] In an optional embodiment, the step of obtaining the document number in step S1 specifically includes: Use the Named Entity Recognition (NER) algorithm in Natural Language Processing (NLP) technology to identify potential document numbers appearing in documents; Apply regular expressions (Regex) to pattern match the submission documents and extract the document numbers that conform to the specified format; The extracted documents are verified and deduplicated to ensure the uniqueness and accuracy of the document number.
[0015] In an optional embodiment, the step of credit information identification in step S2 specifically includes: Use text mining technology, such as TF-IDF.Word2Vec, to extract keywords and perform topic modeling on the submissions to identify credit-related text fragments; Combined with the domain knowledge base, a rule engine is built to identify key information in documents, such as amount, date, default record, etc. Apply machine learning algorithms (such as support vector machine SVM, random forest RF, etc.) to classify documents to distinguish data of different credit levels.
[0016] In an optional embodiment, the step of establishing the aggregation rule in step S3 specifically includes: Analyze historical data and business needs to determine the basic framework and classification standards of the collection rules; Use machine learning algorithms such as decision trees and logistic regression to learn from existing data and automatically generate or optimize aggregation rules; Combine manual review and expert knowledge to revise and improve automatically generated rules.
[0017] In an optional embodiment, the rule base construction and matching steps in step S4 specifically include: Establish a rule base containing all known submission book numbers and their corresponding collection rules, and structure the rule base to facilitate retrieval and query; On the basis of document number extraction and credit information identification, the rule engine is used for rule matching, which can be achieved through fuzzy matching, regular expression matching, etc. During the matching process, a similarity-based matching algorithm (such as cosine similarity, Jaccard similarity, etc.) can be used to calculate the similarity between the document number or credit information and the rules in the rule library to find the most matching collection rule.
[0018] In an optional embodiment, the steps of data collection and rule updating in step S5 specifically include: Once the submission guide successfully matches the collection rules, the system can trigger the data extraction and integration process to collect the relevant credit data to the designated location; For newly appeared document numbers or document numbers that cannot be matched, the system can automatically record and mark them as pending for further manual review and processing; Regularly update and maintain the collection rule base, including adding new rules, deleting or modifying old rules, to ensure the adaptability and accuracy of the system. This can be achieved by regularly running machine learning algorithms to optimize rules, or adjusting rules based on manual review results.
[0019] The present invention can be applied to: The present invention realizes automated and standardized credit data collection and processing, and identifies different types of document numbers. Collection rules are established according to the credit data type, source, format, etc. corresponding to the document number. At the same time, new document numbers and collection rules are automatically identified, and the data collection rule library is automatically updated and maintained. The credit data collection rule identification method and system based on the document number can effectively improve the accuracy and reliability of credit data through automated and intelligent collection and identification processes, reduce the error rate of manual operations, and provide more accurate data support for credit evaluation and management.
[0020] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for identifying credit data collection rules based on document numbers, characterized in that: The following steps are involved: Step S1: extract document number; Step S2: credit information identification; Step S3: Establishing aggregation rules; Step S4: rule base construction and matching; Step S5: Data collection and rule update.
2. A method for identifying credit data collection rules based on document numbers according to claim 1, characterized in that: The steps of obtaining the document number in step S1 specifically include: Use the Named Entity Recognition (NER) algorithm in Natural Language Processing (NLP) technology to identify potential document numbers appearing in documents; Apply regular expressions (Regex) to pattern match the submission documents and extract the document numbers that conform to the specified format; The extracted documents are verified and deduplicated to ensure the uniqueness and accuracy of the document number.
3. A method for identifying credit data collection rules based on document number according to claim 1, characterized in that: The steps of credit information identification in step S2 specifically include: Use text mining technology, such as TF-IDF.Word2Vec, to extract keywords and perform topic modeling on the submissions to identify credit-related text fragments; Combined with the domain knowledge base, a rule engine is built to identify key information in documents, such as amount, date, default record, etc. Apply machine learning algorithms (such as support vector machine SVM, random forest RF, etc.) to classify documents to distinguish data of different credit levels.
4. A method for identifying credit data collection rules based on document numbers according to claim 1, characterized in that: The steps of establishing the aggregation rules in step S3 specifically include: Analyze historical data and business needs to determine the basic framework and classification standards of the collection rules; Use machine learning algorithms such as decision trees and logistic regression to learn from existing data and automatically generate or optimize aggregation rules; Combine manual review and expert knowledge to revise and improve automatically generated rules.
5. The method for identifying credit data collection rules based on document number according to claim 1, characterized in that: The steps of rule base construction and matching in step S4 specifically include: Establish a rule base containing all known submission book numbers and their corresponding collection rules, and structure the rule base to facilitate retrieval and query; On the basis of document number extraction and credit information identification, the rule engine is used for rule matching, which can be achieved through fuzzy matching, regular expression matching, etc. During the matching process, a similarity-based matching algorithm (such as cosine similarity, Jaccard similarity, etc.) can be used to calculate the similarity between the document number or credit information and the rules in the rule library to find the most matching collection rule.
6. A method for identifying credit data collection rules based on document numbers according to claim 1, characterized in that: The steps of data collection and rule updating in step S5 specifically include: Once the submission guide successfully matches the collection rules, the system can trigger the data extraction and integration process to collect the relevant credit data to the designated location; For newly appeared document numbers or document numbers that cannot be matched, the system can automatically record and mark them as pending for further manual review and processing; Regularly update and maintain the collection rule base, including adding new rules, deleting or modifying old rules, to ensure the adaptability and accuracy of the system. This can be achieved by regularly running machine learning algorithms to optimize rules, or adjusting rules based on manual review results.
