Entity extraction technology-based liability and responsibility list construction method and system

Abstract

The invention provides a liability list construction method and system based on an entity extraction technology, and belongs to the technical field of administrative law enforcement judicial auxiliary supervision. The method comprises the steps that legal provision information is collected, the collected legal provision information is retrieved, and legal provisions with relevant regulations of rights and responsibilities are analyzed; key elements including law enforcement situations, law enforcement subjects and right responsibilities are extracted according to the legal provision content; marking law enforcement situations, law enforcement subjects and right and responsibility entity information on legal provisions, and constructing a right and responsibility list by adopting a trained entity extraction model; and obtaining and analyzing the latest regulation change in real time based on the law and regulation dynamic library and the policy document library, and automatically updating each content of the liability list. Accuracy and timeliness of right and responsibility list construction are effectively improved, and human errors and omission are effectively reduced. The responsibility list is automatically identified and updated, so that the adaptability of the responsibility list in a rapidly changing legal environment is ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of administrative law enforcement and judicial auxiliary supervision, and in particular relates to a method and system for constructing a rights and responsibilities list based on entity extraction technology. 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] As a legal norm and administrative management tool, the power and responsibility list not only helps clarify the functions and responsibilities of various administrative entities but also effectively prevents abuse of power and shirking of responsibility during law enforcement. Traditional methods of constructing power and responsibility lists often rely on manual interpretation and summarization, lacking efficient and automated tools to process and analyze large volumes of legal provisions, policy documents, and relevant judicial precedents. This approach is not only time-consuming and labor-intensive, but also prone to misunderstandings and omissions, resulting in incomplete or inaccurate power and responsibility lists, which in turn undermines the legitimacy and fairness of law enforcement.

[0004] As laws and regulations continue to evolve, new legal provisions, policies, measures, and judicial interpretations are introduced, requiring the system to dynamically adapt and update existing rights and responsibilities lists. Ensuring that the system accurately and promptly reflects changes in the legal environment is a key challenge in building a rights and responsibilities list system. Summary of the Invention

[0005] To overcome the shortcomings of the aforementioned existing technologies, the present invention provides a method and system for constructing a list of rights and responsibilities based on entity extraction technology. This method addresses issues inherent in traditional list construction, such as manual interpretation bias and inefficiency. By combining technologies such as big data analysis, natural language processing, and machine learning, the list of rights and responsibilities can be dynamically constructed and promptly updated with the latest laws and regulations, thereby improving the fairness, transparency, and efficiency of law enforcement.

[0006] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0007] The first aspect of the present invention provides a method for constructing a rights and responsibilities list based on entity extraction technology;

[0008] A method for constructing a rights and responsibilities list based on entity extraction technology, including:

[0009] Collect legal provisions information, search the collected legal provisions information and analyze the legal provisions with relevant provisions on rights and responsibilities;

[0010] Extract key elements based on the content of legal provisions, including law enforcement circumstances, law enforcement entities, and rights and responsibilities;

[0011] Mark the law enforcement situation, law enforcement subject and rights and responsibilities entity information on the legal provisions, and use the trained entity extraction model to build a list of rights and responsibilities;

[0012] Based on the dynamic database of laws and regulations and the policy document library, the latest regulatory changes are obtained and analyzed in real time, and the contents of the list of rights and responsibilities are automatically updated.

[0013] As a further technical solution, the process of retrieving the collected legal text information is as follows:

[0014] Using a large language model, a description of legal provisions containing provisions on rights and responsibilities is generated given a legal database. The description of the legal provisions provides a description of the enforcement situation and generates legal provisions based on the understanding of the case.

[0015] The generated legal provisions are used as queries and vector cosine similarity is used to retrieve relevant legal provisions from the legal provision database.

[0016] As a further technical solution, the process of obtaining and analyzing the latest regulatory changes in real time based on the dynamic database of laws and regulations and the policy document database, and automatically updating the contents of the list of rights and responsibilities is as follows:

[0017] When new legal provisions are introduced or existing regulations are revised, a change detection algorithm is used to automatically identify the differences between the new or revised provisions and the original legal provisions. Combined with semantic similarity analysis, it can accurately identify the relationship between the revised content and the original provisions.

[0018] By comparing newly added or revised legal provisions, incremental learning combined with knowledge graphs and logical reasoning is used to identify the changed administrative entities or responsibilities, and update the original list of rights and responsibilities based on the changed content.

[0019] The second aspect of the present invention provides a system for constructing a list of rights and responsibilities based on entity extraction technology.

[0020] A system for building a list of rights and responsibilities based on entity extraction technology, including:

[0021] A legal document parsing module, which is used to automatically parse current legal provisions, administrative regulations, policy documents, and judicial precedents;

[0022] An entity extraction module, which is used to extract rights and responsibilities clauses, law enforcement items, and functions and responsibilities of administrative entities from the parsed legal documents;

[0023] A rights and responsibilities list generation module, which is used to represent the extracted rights and responsibilities in the form of structured data and automatically generate a rights and responsibilities list based on the logical relationship of the legal provisions;

[0024] A legal change detection module, which is used to monitor changes in laws and regulations in real time and automatically identify newly added or revised legal provisions;

[0025] An incremental update module is used to automatically update the list of rights and responsibilities based on detected legal changes, combining knowledge graphs and logical reasoning.

[0026] As a further technical solution, the entity extraction module uses text analysis and information extraction technology to extract rights and responsibilities clauses, law enforcement items and the functions and responsibilities of administrative entities.

[0027] As a further technical solution, the rights and responsibilities list generation module uses knowledge graph technology to represent the rights and responsibilities relationship in the form of structured data, and automatically generates the rights and responsibilities list based on the logical relationship of the legal provisions.

[0028] As a further technical solution, the legal change detection module automatically compares the rights and responsibilities clauses and implementation standards before and after the update through semantic similarity analysis technology to identify new or revised legal provisions.

[0029] As a further technical solution, the incremental update module automatically updates the original list of rights and responsibilities through incremental learning technology, combined with knowledge graph and logical reasoning.

[0030] The third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in a method for constructing a list of rights and responsibilities based on entity extraction technology as described in the first aspect of the present invention.

[0031] The fourth aspect of the present invention provides an electronic device, including a memory, a processor, and a program stored in the memory and runnable on the processor. When the processor executes the program, it implements the steps in the method for constructing a list of rights and responsibilities based on entity extraction technology as described in the first aspect of the present invention.

[0032] One or more of the above technical solutions have the following beneficial effects:

[0033] (1) Through automated text parsing technology, the present invention can quickly extract rights and responsibilities information from current legal provisions and administrative regulations, automatically generate a list of rights and responsibilities, and update it in real time. This greatly improves the accuracy and timeliness of the construction of the list of rights and responsibilities, ensures that the rights and responsibilities of all parties are clear and unambiguous during the law enforcement process, and effectively reduces human errors and omissions;

[0034] (2) This invention uses machine learning and incremental learning technologies to achieve automatic updating of the list of rights and responsibilities. The system can monitor and analyze changes in laws, regulations, policy documents, and judicial interpretations in real time, automatically identify new or revised legal provisions, and automatically reflect these changes in the list of rights and responsibilities. This improves the accuracy and consistency of law enforcement, effectively reduces manual intervention, and ensures the adaptability of the list of rights and responsibilities in a rapidly changing legal environment.

[0035] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings, 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.

[0037] Figure 1 This is a flow chart of the method of the first embodiment.

[0038] Figure 2 This is a schematic diagram of the entity extraction model framework in the first embodiment.

[0039] Figure 3 This is a system structure diagram of the second embodiment. DETAILED DESCRIPTION

[0040] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0041] It should be noted that the terms used herein are for describing particular embodiments only and are not intended to limit the exemplary embodiments according to the present invention.

[0042] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0043] Example 1

[0044] This embodiment discloses a method for constructing a rights and responsibilities list based on entity extraction technology;

[0045] like Figure 1 As shown in the figure, a method for constructing a list of rights and responsibilities based on entity extraction technology includes:

[0046] Step S1: Collect legal provisions information, search the collected legal provisions information and analyze the legal provisions with relevant provisions on rights and responsibilities;

[0047] In step S11, a large language model is used to generate a description of a legal provision containing provisions on rights and responsibilities, given a legal database. The description of the legal provision specifies the law enforcement situation and generates legal provision a based on the understanding of the case. The process of generating legal provisions using the large language model can be expressed as the following formula:

[0048] a=LLM(x,p);

[0049] Among them, LLM represents the large language model, x represents laws and regulations, and p represents the corresponding prompt words generated by the large model. The prompt words "Please list the relevant laws and regulations that stipulate rights and responsibilities based on laws and regulations" guide the large language model to generate relevant laws and regulations.

[0050] Since legal provisions are strictly accurate, and the large language model may be inconsistent with the original legal provisions during the generation process, each generated legal provision is used as a query to retrieve relevant legal provisions from the legal provision database. The retrieval method uses the vector cosine similarity sorting method, and the vector representation of the predicted legal provision is:

[0051] v a =encoder(a);

[0052] The vector representation of the legal provisions in the legal provisions database is:

[0053]

[0054] The formula for calculating the cosine similarity between two vectors is:

[0055]

[0056] The encoder adopts a bidirectional encoder structure. Represents the kth legal article in the legal article library, and selects the top 5 legal articles ranked by cosine similarity as relevant legal articles.

[0057] Step S12, since the workload of manually analyzing legal provisions is large and the work efficiency is low, the model assists humans in completing the analysis of legal provisions by adopting a collaborative approach of manual and large language models. The model uses the prompt word: 'Please determine whether the legal provisions currently retrieved stipulate specific rights and responsibilities, and include law enforcement circumstances, law enforcement subjects, rights and responsibilities, etc. Please answer "yes" or "no" and explain the reason.' to analyze whether the retrieved legal provisions stipulate specific rights and responsibilities. If the answer is "yes", it means that the large language model has determined that the legal provisions currently retrieved stipulate complete rights and responsibilities information, and if the answer is "no", it means that the legal provisions currently retrieved do not stipulate corresponding rights and responsibilities information. Then return to step S11 to continue searching.

[0058] Step S2: extract key elements based on the content of the legal provisions, including law enforcement circumstances, law enforcement entities, and rights and responsibilities;

[0059] The key elements of the rights and responsibilities list include the enforcement situation, the enforcement subject, and the rights and responsibilities. Different elements describe the rights and responsibilities corresponding to different enforcement subjects in different enforcement situations:

[0060] Combine Figure 2 In this embodiment, the BERT-CRF entity extraction model is used, the WordPi ece word segmenter is used to process Chinese text, and special tokens [CLS] and [SEP] are added. These paragraphs are then passed independently through a BERT-based shared model (such as BERT-base-ch i nese), and a 12-layer Transformer encoder (hidden layer dimension 768, number of attention heads 12) is used to extract a list of context-independent paragraph representations. Next, the chunking model adds the token of each paragraph to the learnable paragraph position embedding (obtained through a trainable lookup table) and feeds the result into a 4-layer small Transformer encoder (hidden layer dimension 256, number of attention heads 8), using relative position encoding to make it aware of the surrounding paragraphs. The final law enforcement basis embedding is calculated by pooling the context-aware paragraph representation (using mean-pooling), that is:

[0061]

[0062] Finally, the embedding of each token is passed through a multi-classification head (containing 3 independent linear classifiers, each with an output dimension of K c , c∈{situation, subject, rights and responsibilities}) predicts the category of each character, uses the CRF layer to model the label transfer probability, and uses the weighted cross entropy loss function (the category weight is inversely proportional to the frequency), which includes law enforcement situations, law enforcement subjects, rights and responsibilities, etc.

[0063] (1) Extraction of law enforcement situation elements: Extraction of law enforcement situation elements refers to identifying and extracting specific situations and conditions related to law enforcement behavior from laws and regulations, and clarifying the key elements in the law enforcement process. The prompt "Please extract the prescribed law enforcement situations according to laws and regulations, and provide an analysis of the law enforcement situations. Law enforcement situations refer to the specific situations in which law enforcement behavior occurs and the relevant legal application conditions" can be used to guide the large language model to generate categories of law enforcement situation elements.

[0064] (2) Extraction of law enforcement subject elements: Extraction of law enforcement subject elements refers to identifying and extracting relevant law enforcement units in law enforcement behaviors from laws and regulations. Law enforcement subject elements include specific law enforcement personnel or law enforcement agencies, such as administrative law enforcement agencies such as the Market Supervision Administration. The prompt "Please extract the law enforcement subject according to laws and regulations and provide an analysis of the law enforcement subject. Law enforcement subject: refers to the person or organization that exercises law enforcement power in accordance with the law" is used to guide the large language model to generate the categories of law enforcement subject elements.

[0065] (3) Rights and responsibilities extraction: Rights and responsibilities extraction refers to identifying and extracting elements related to specific rights and responsibilities from legal texts. The prompt "Please extract the rights and responsibilities corresponding to the law enforcement entity according to laws and regulations, and provide an analysis of the rights and responsibilities. Rights and responsibilities refer to the rights that the entity should enjoy and the responsibilities that it should bear as stipulated in the legal text" is used to guide the large language model to generate categories of rights and responsibilities.

[0066] Step S3: Mark the law enforcement situation, law enforcement subject, and rights and responsibilities entity information on the legal provisions, and use the trained entity extraction model to build a rights and responsibilities list;

[0067] During the entity labeling phase, based on the previously defined definitions of three entity categories: "enforcement entity," "enforcement situation," and "rights and responsibilities," the accuracy of the extracted entities is determined by inputting the legal provisions and extracted entities into a large language model used for entity extraction. For example, "Market Supervision and Administration Bureau" is an enforcement entity, "unauthorized operation" is an enforcement situation, and "fine" is a right and responsibility. The final labeling results are output in standard JSON format.

[0068] Step S4, based on the dynamic database of laws and regulations and the policy document database, real-time acquisition and analysis of the latest changes in regulations, and automatic update of the contents of the list of rights and responsibilities.

[0069] By connecting to the legal database, policy document database, and judicial interpretation database in real time, we regularly monitor newly released legal documents and revised existing regulations. When new legal provisions or revisions to existing regulations are discovered, we use a change detection algorithm to automatically identify the differences between the new or revised provisions and the original legal provisions. Combined with semantic similarity analysis, we can accurately identify the relationship between the revised content and the original provisions.

[0070] By comparing newly added or revised legal provisions, incremental learning combined with knowledge graphs and logical reasoning is used to identify the changed administrative subjects or responsibilities, and update the original list of rights and responsibilities based on the changed content. Specifically: by connecting with the legal database, policy document database and judicial interpretation database in real time, regularly monitor newly released legal documents and revised existing regulations. When new legal provisions or revisions to existing regulations are found, a change detection algorithm (based on a hybrid model of text difference and semantic encoding) is used to automatically identify the differences between the new or revised provisions and the original legal provisions: First, the longest common subsequence (LCS) algorithm is used to locate the text difference segment, and the difference degree is calculated as:

[0071]

[0072] Where D is the difference; A and B represent the new and old texts respectively; semantic analysis is triggered when D > threshold θ (default 0.3). Combined with semantic similarity analysis, the BERT-wwm pre-trained model is used to generate sentence vectors and calculate cosine similarity. Accurately identify the relationship between the revised content and the original clauses; by comparing the newly added or revised legal provisions, identify the changed administrative subjects or responsibilities, and update the original list of rights and responsibilities based on the changed content. The update strategy is Where ΔR is the incremental triple of <subject, relation, object> extracted by pattern matching, KG ′ is the updated knowledge graph of the list of rights and responsibilities; KG is the knowledge graph of the original list of rights and responsibilities; Represents the exclusive OR operation.

[0073] Example 2

[0074] This embodiment discloses a system for constructing a list of rights and responsibilities based on entity extraction technology;

[0075] like Figure 2 As shown in the figure, a system for building a list of rights and responsibilities based on entity extraction technology includes:

[0076] A legal document parsing module, which is used to automatically parse current legal provisions, administrative regulations, policy documents, and judicial precedents;

[0077] An entity extraction module, which is used to extract rights and responsibilities clauses, law enforcement items, and functions and responsibilities of administrative entities from the parsed legal documents;

[0078] A rights and responsibilities list generation module, which is used to represent the extracted rights and responsibilities in the form of structured data and automatically generate a rights and responsibilities list based on the logical relationship of the legal provisions;

[0079] A legal change detection module, which is used to monitor changes in laws and regulations in real time and automatically identify newly added or revised legal provisions;

[0080] An incremental update module is used to automatically update the list of rights and responsibilities based on detected legal changes, combining knowledge graphs and logical reasoning.

[0081] The entity extraction module utilizes text analysis and information extraction techniques to extract rights and responsibilities clauses, enforcement items, and the functions and responsibilities of administrative entities. The rights and responsibilities list generation module uses knowledge graph technology to represent rights and responsibilities relationships as structured data and automatically generates a rights and responsibilities list based on the logical relationships within legal text. The legal change detection module uses semantic similarity analysis to automatically compare rights and responsibilities clauses and enforcement standards before and after updates, identifying new or revised legal provisions. The incremental update module uses incremental learning technology, combined with knowledge graphs and logical reasoning, to automatically update the existing rights and responsibilities list.

[0082] Example 3

[0083] The purpose of this embodiment is to provide a computer-readable storage medium.

[0084] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in a method for constructing a list of rights and responsibilities based on entity extraction technology as described in Example 1.

[0085] Example 4

[0086] The purpose of this embodiment is to provide an electronic device.

[0087] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, the processor implements the steps in a method for constructing a list of rights and responsibilities based on entity extraction technology as described in Example 1.

[0088] The steps involved in the apparatuses of Examples 2, 3, and 4 above correspond to those of Method Example 1. For detailed implementations, please refer to the relevant description of Example 1. The term "computer-readable storage medium" should be understood to mean a single medium or multiple media containing one or more instruction sets; it should also be understood to include any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and causing the processor to perform any method of the present invention.

[0089] Those skilled in the art will appreciate that the modules or steps of the present invention described above can be implemented using a general-purpose computer device. Alternatively, they can be implemented using program code executable by a computing device, which can then be stored in a storage device and executed by the computing device. Alternatively, they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0090] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A method for constructing a list of rights and responsibilities based on entity extraction technology, characterized in that: include: Collect legal provisions information, search the collected legal provisions information and analyze the legal provisions with relevant provisions on rights and responsibilities; Extract key elements based on the content of legal provisions, including law enforcement circumstances, law enforcement entities, and rights and responsibilities; Mark the law enforcement situation, law enforcement subject and rights and responsibilities entity information on the legal provisions, and use the trained entity extraction model to build a list of rights and responsibilities; Based on the dynamic database of laws and regulations and the policy document library, the latest regulatory changes are obtained and analyzed in real time, and the contents of the list of rights and responsibilities are automatically updated.

2. A method for constructing a list of rights and responsibilities based on entity extraction technology as claimed in claim 1, characterized in that: The process of retrieving the collected legal text information is as follows: Using a large language model, a description of legal provisions containing provisions on rights and responsibilities is generated given a legal database. The description of the legal provisions provides a description of the enforcement situation and generates legal provisions based on the understanding of the case. The generated legal provisions are used as queries and vector cosine similarity is used to retrieve relevant legal provisions from the legal provision database.

3. The method for constructing a list of rights and responsibilities based on entity extraction technology according to claim 1, characterized in that: The process of obtaining and analyzing the latest regulatory changes in real time based on the dynamic database of laws and regulations and the policy document database, and automatically updating the contents of the list of rights and responsibilities is as follows: When new legal provisions are introduced or existing regulations are revised, a change detection algorithm is used to automatically identify the differences between the new or revised provisions and the original legal provisions. Combined with semantic similarity analysis, it can accurately identify the relationship between the revised content and the original provisions. By comparing newly added or revised legal provisions, incremental learning combined with knowledge graphs and logical reasoning is used to identify the changed administrative entities or responsibilities, and update the original list of rights and responsibilities based on the changed content.

4. A system for building a list of rights and responsibilities based on entity extraction technology, characterized by: include: A legal document parsing module, which is used to automatically parse current legal provisions, administrative regulations, policy documents, and judicial precedents; An entity extraction module, which is used to extract rights and responsibilities clauses, law enforcement items, and functions and responsibilities of administrative entities from the parsed legal documents; A rights and responsibilities list generation module, which is used to represent the extracted rights and responsibilities in the form of structured data and automatically generate a rights and responsibilities list based on the logical relationship of the legal provisions; A legal change detection module, which is used to monitor changes in laws and regulations in real time and automatically identify newly added or revised legal provisions; An incremental update module is used to automatically update the list of rights and responsibilities based on detected legal changes, combining knowledge graphs and logical reasoning.

5. A system for constructing a list of rights and responsibilities based on entity extraction technology as described in claim 4, characterized in that: The entity extraction module uses text analysis and information extraction technology to extract rights and responsibilities clauses, law enforcement items and the functions and responsibilities of administrative entities.

6. A system for constructing a list of rights and responsibilities based on entity extraction technology as claimed in claim 4, characterized in that: The rights and responsibilities list generation module uses knowledge graph technology to represent the rights and responsibilities relationship in the form of structured data, and automatically generates the rights and responsibilities list based on the logical relationship of the legal provisions.

7. A system for constructing a list of rights and responsibilities based on entity extraction technology as claimed in claim 4, characterized in that: The legal change detection module uses semantic similarity analysis technology to automatically compare the rights and responsibilities clauses and implementation standards before and after the update to identify new or revised legal provisions.

8. A system for constructing a list of rights and responsibilities based on entity extraction technology as claimed in claim 4, characterized in that: The incremental update module automatically updates the original list of rights and responsibilities through incremental learning technology, combined with knowledge graphs and logical reasoning.

9. A computer-readable storage medium having a program stored thereon, characterized in that: When the program is executed by a processor, the steps in the method for constructing a list of rights and responsibilities based on entity extraction technology as described in any one of claims 1 to 3 are implemented.

10. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, wherein: When the processor executes the program, it implements the steps in the method for constructing a list of rights and responsibilities based on entity extraction technology as described in any one of claims 1-3.

Images