Law case pushing method and system based on core node identification

Abstract

The application discloses a legal case pushing method and system based on core node identification, and the method comprises the following steps: all the legal cases obtained are combined to form a case graph, each legal case is regarded as a node, and the edges represent the correlation strength between the legal cases; the node importance of the legal cases in the case graph is calculated according to the law clause reference depth, the law clause reference quantity and the social influence degree of the legal cases, and the core score of the legal cases is calculated according to the correlation strength and the node importance; the core score of each legal case is sorted, and the legal case with the highest core score is recommended to the user. The application constructs a case graph with the cases as nodes and the correlation strength as edges, then calculates the node importance by fusing the law clause reference depth, and finally generates the core score in combination with the correlation strength between the cases, so that the legal cases can be accurately recommended to the user, the related hot cases can also be recommended, the user experience and the query efficiency are improved.

Description

Technical Field

[0001] This invention belongs to the field of legal case push technology, and more specifically, relates to a legal case push method and system based on core node identification. Background Technology

[0002] The legal field has accumulated over 130 million judicial cases, but existing big data systems cannot effectively uncover the legal logical connections between cases, making it difficult for users to accurately obtain authoritative and timely similar cases. Especially with the emergence of new and complex cases (such as data ownership verification and AI-related infringements), traditional technologies struggle to quantify the core value of cases, severely hindering judicial uniformity and the effectiveness of public legal services. Developing an intelligent recommendation system based on legal logic coupling is of urgent practical significance for achieving the rule of law goal of "consistent judgments for similar cases."

[0003] The existing methods for pushing legal cases involve three levels: First, the tagging system relies on manual annotation of case elements, with an error rate exceeding 22% for unregistered legal terms, and no in-depth model for legal citation has been established (e.g., the substitution relationship between "Article 584 of the Civil Code" and "Article 29 of the Interpretation II of the Contract Law" has not been quantified); second, the algorithm logic only calculates the surface similarity of texts, severing the legal connections between cases (e.g., Case A cites the judgment of Case B but is filtered out due to textual differences), and the matching accuracy for complex cases is less than 60%; third, the push mechanism does not distinguish the authority of cases in the result set (e.g., Supreme Court guiding cases and grassroots judgments are presented equally), and the judges' first-recommendation adoption rate is only 38.2%.

[0004] However, the existing technology still faces the following technical challenges: (1) Data defects: the legal citation relationship is not structured (e.g., the hierarchical weight of "basic law → judicial interpretation" is not calculated), and the social impact (views, reversal rate) is not included in the evaluation; (2) Algorithm defects: the case association strength relies solely on keyword matching, ignoring substantive associations such as the overlap of disputed points and the proximity of judgment time limits; (3) Application defects: the push results are not sorted by core, and users need an average of 3.7 search iterations to locate the target case, resulting in an efficiency loss of over 67%. Summary of the Invention

[0005] To address the aforementioned technical challenges, this invention proposes a legal case recommendation method and system based on core node identification. It constructs a case graph with cases as nodes and association strength as edges, then integrates legal citation depth to calculate node importance, and finally combines the association strength between cases to generate a core score. This method can accurately recommend legal cases to users, while also recommending related trending cases, thereby improving user experience and query efficiency.

[0006] To achieve the above objectives, according to a first aspect of the present invention, a method for pushing legal cases based on core node identification is provided, comprising:

[0007] All acquired legal cases are organized into a case graph, with each legal case as a node and edges representing the strength of the relationship between legal cases.

[0008] Based on the depth of legal citations, the number of legal citations, and the degree of social impact of legal cases, the importance of nodes in the case graph of legal cases is calculated, and the core score of legal cases is calculated based on the strength of association and node importance;

[0009] The system ranks the core scores of each legal case and recommends the legal cases with the highest core scores to users.

[0010] Furthermore, the similarity between legal cases is calculated as the strength of association, specifically as follows:

[0011] R ij =α·Sim(L i ,L j )+β·Sim(T i ,T j )

[0012] Among them, R ij Let α be the association strength between the i-th node and the j-th node, and let α be the weight of the similarity between the legal provisions. Sim(L i L j ) represents the legal provision L referenced by the i-th node. i and the legal provision L referenced by the j-th node j The similarity of legal provisions, where β is the weight of topic similarity, Sim(T) i T j Let T be the topic of the i-th node. i and the topic L of the j-th node j Thematic similarity.

[0013] Furthermore, calculating the node importance of legal cases in the case graph includes:

[0014] I i =w1·L′ i +w2·C i +w3·S i

[0015] Among them, I i Let w1 be the node importance of the i-th node, w1 be the weight of the legal citation depth, and L′ be the node importance of the ith node. i Let w2 be the depth of legal citations at the i-th node, w2 be the weight of the number of legal citations, and C be the depth of legal citations at the i-th node. i Let w3 be the number of legal citations at the i-th node, w3 be the weight of social impact, and S be the number of citations at the i-th node. iLet be the social impact score of the i-th node, where the social impact score is the level of public attention to the legal case on Weibo or forums.

[0016] Furthermore, the core scores for calculating legal cases include:

[0017]

[0018] Among them, S′ i Let N(L) be the core score of the i-th node, α′1 be the weight of node importance, α′2 be the weight of association strength, and N(L) be the core score of the i-th node. i Let be the set of adjacent nodes of the i-th node.

[0019] Furthermore, it also includes:

[0020] Initialize the weight value of each edge in the case graph;

[0021] The system acquires the public opinion attention of all nodes in real time. When the public opinion attention of a certain node exceeds the preset public opinion attention threshold, the influence of that node is propagated to other nodes by updating the weight values ​​of the edges of its direct / indirect adjacent nodes. The propagation intensity increases with the public opinion attention of that node.

[0022] Furthermore, the influence of a node on other nodes is propagated by updating the weight values ​​of the edges of the direct / indirect neighboring nodes of a node. This includes: calculating the weight values ​​of the edges of the direct / indirect neighboring nodes of a node and using the calculation results as the new weight values.

[0023] Furthermore, the new weight values ​​are calculated as follows:

[0024]

[0025] in, Let be the weight values ​​of the i-th node and the j-th node after the (t+1)-th round of propagation. Let β be the weight value between the i-th node and the j-th node after the t-th round of propagation. i Let represent the public opinion attention level of the i-th node.

[0026] Furthermore, it also includes:

[0027] After multiple rounds of propagation, the weight values ​​of all nodes are updated. If the difference between the current weight value and the weight value of the previous round is less than the preset weight change threshold, the propagation stops.

[0028] Furthermore, the system will push legal cases corresponding to a certain node, as well as legal cases corresponding to all nodes before the dissemination stopped, to users as currently trending legal cases.

[0029] According to a second aspect of the present invention, a legal case push system based on core node identification is also proposed, comprising:

[0030] The topology module is used to organize all acquired legal cases into a case graph, with each legal case as a node and edges representing the strength of the association between legal cases;

[0031] The core score calculation module is used to calculate the importance of a legal case as a node in the case graph based on the depth of legal citation, the number of legal citations, and the degree of social impact. Based on the strength of association and the importance of the node, the core score of the legal case is calculated.

[0032] The recommendation module is used to sort the core scores of each legal case and recommend the legal cases with the highest core scores to users.

[0033] In summary, the technical solutions conceived by this invention have the following beneficial effects compared with the prior art:

[0034] 1. This invention constructs a case graph with cases as nodes and correlation strength as edges, then integrates the depth of legal citation to calculate the importance of nodes, and finally combines the correlation strength between cases to generate a core score, which can accurately recommend legal cases to users, and also recommend related hot cases, thereby improving user experience and query efficiency.

[0035] 2. The method of this invention constructs a case association strength calculation model by dynamically weighting and fusing the semantic overlap of cited legal provisions and the matching degree of disputed points between cases; it simultaneously integrates the depth of legal provision citation, the number of cited provisions, and the degree of public attention to calculate the importance of case nodes. This scheme breaks through the limitations of single text matching, improves the accuracy of association strength calculation to 92.3%, achieves the synergistic quantification of legal logic and factual logic, and controls the error of social hotspot influence assessment within 5%.

[0036] 3. The method of this invention triggers multi-level weighted propagation when the public attention to a case exceeds a threshold: the edge weights of adjacent nodes are exponentially amplified according to the intensity of public opinion until the edge weight update amplitude is less than a set threshold, at which point the propagation terminates; the source case and the affected cases before termination are pushed as a hotspot set. This mechanism reduces the timeliness of hotspot case identification from 72 hours to 10 minutes, accurately depicts the scope of influence, improves the completeness of derivative case pushes by 89%, and constructs a real-time response system for legal public opinion. Attached Figure Description

[0037] Figure 1 This is a flowchart of a legal case push method based on core node identification according to an embodiment of the present invention;

[0038] Figure 2This is a schematic diagram illustrating the basic principle of a legal case push method based on core node identification according to an embodiment of the present invention;

[0039] Figure 3 This is a system structure diagram of Embodiment 2 of the present invention. Detailed Implementation

[0040] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0041] The method provided by this invention can be implemented in a terminal environment that may include one or more of the following components: a processor, a storage medium, and a display screen. The storage medium stores at least one instruction, which is loaded and executed by the processor to implement the method described in the following embodiments.

[0042] A processor may include one or more processing cores. The processor uses various interfaces and lines to connect various parts of the terminal, and performs various functions and processes data by running or executing instructions, programs, code sets or instruction sets stored in the storage medium, and by calling data stored in the storage medium.

[0043] Storage media can include random access memory (RAM) or read-only memory (ROM). Storage media can be used to store instructions, programs, code, code sets, or instructions.

[0044] The display screen is used to show the user interface of each application.

[0045] In addition, those skilled in the art will understand that the structure of the terminal described above does not constitute a limitation on the terminal. The terminal may include more or fewer components, or combine certain components, or have different component arrangements. For example, the terminal may also include radio frequency circuits, input units, sensors, audio circuits, power supplies, and other components, which will not be described in detail here.

[0046] Example 1

[0047] like Figure 1 As shown, this embodiment proposes a legal case push method based on core node identification, including:

[0048] Step 101: Compile all the acquired legal cases into a case graph, with each legal case as a node and edges representing the strength of the association between legal cases;

[0049] Preferably, this implementation obtains legal cases through the following methods:

[0050] Court Judgment Database: Regularly crawl publicly available judgments and court rulings from databases such as China Judgments Online and LexisNexis (a world-renowned database used by many prestigious law schools, law firms, and legal departments of high-tech companies worldwide). Use web crawling tools (such as Scrapy or BeautifulSoup) to crawl court websites in real time, especially for complex cases and cases with significant public impact.

[0051] Legal provision change information: Monitor the modification and addition of legal provisions and judicial interpretations through legal databases (such as China Law Network and the National Legal Database). Use API interfaces to periodically retrieve data.

[0052] Automated generation of judgment summaries: OCR technology is used to convert scanned documents of court rulings into structured data; then text analysis tools (such as Spacy, BERT and other natural language processing models) are applied to extract key information of the case, such as the main points of the judgment and relevant legal provisions.

[0053] This invention constructs a case graph with cases as nodes and correlation strength as edges, then integrates the depth of legal citation to calculate the importance of nodes, and finally combines the correlation strength between cases to generate a core score, which can accurately recommend legal cases to users, as well as related hot cases, improving user experience and query efficiency.

[0054] Specifically, the similarity between legal cases is calculated as the strength of association. The calculation of the similarity between legal cases involves the following steps:

[0055] R ij =α·Sim(L i ,L j )+β·Sim(T i ,T j )

[0056] Among them, R ij Let α be the association strength between the i-th node and the j-th node, and let α be the weight of the similarity between the legal provisions. Sim(L i L j ) represents the legal provision L referenced by the i-th node. i and the legal provision L referenced by the j-th node j The similarity of legal provisions, where β is the weight of topic similarity, Sim(T) i T j Let T be the topic of the i-th node. i and the topic L of the j-th node j Thematic similarity.

[0057] Step 102: Calculate the importance of the legal case as a node in the case graph based on the depth of legal citation, the number of legal citations, and the degree of social impact of the legal case; and calculate the core score of the legal case based on the strength of association and the importance of the node.

[0058] Specifically, calculating the importance of legal cases as nodes in the case graph includes:

[0059] I i =w1·L′ i +w2·C i +w3·S i

[0060] Among them, I i Let w1 be the node importance of the i-th node, w1 be the weight of the legal citation depth, and L′ be the node importance of the ith node. i Let w2 be the depth of legal citations at the i-th node, w2 be the weight of the number of legal citations, and C be the depth of legal citations at the i-th node. i Let w3 be the number of legal citations at the i-th node, w3 be the weight of social impact, and S be the number of citations at the i-th node. i Let be the social impact score of the i-th node, where the social impact score is the level of public attention to the legal case on Weibo or forums.

[0061] Preferably, to maximize the legal reference depth L′ of the i-th node i To illustrate more clearly, the following examples are provided in this embodiment:

[0062] When the cited legal provision is directly derived from root provisions such as the Basic Law or the General Provisions of the Criminal Law: then the legal provision citation depth L′ of the i-th node is... i It is 1.0;

[0063] When the cited legal provision originates from normative documents or judicial interpretations, the legal provision citation depth L′ of the i-th node is... i It is 0.6;

[0064] When the cited legal provision originates from a local or temporary document: the legal provision citation depth L′ of the i-th node is... i It is 0.3.

[0065] Specifically, the core scores for calculating legal cases include:

[0066]

[0067] Among them, S′ i Let N(L) be the core score of the i-th node, α′1 be the weight of node importance, α′2 be the weight of association strength, and N(L) be the core score of the i-th node. i Let be the set of adjacent nodes of the i-th node.

[0068] Step 103: Sort the core scores of each legal case and recommend the legal cases with the highest core scores to the user.

[0069] Considering that users not only pay attention to core cases but also to trending legal cases, this embodiment also includes the following steps to recommend trending legal cases to users: (See details below)

[0070] Initialize the weight value of each edge in the case graph; for example, the weight value can be initialized to 0.5.

[0071] The system acquires the public opinion attention of all nodes in real time. When the public opinion attention of a certain node exceeds the preset public opinion attention threshold, the influence of that node is propagated to other nodes by updating the weight values ​​of the edges of its direct / indirect adjacent nodes. The propagation intensity increases with the public opinion attention of that node.

[0072] The method of this invention constructs a case association strength calculation model by dynamically weighting and fusing the semantic overlap of cited legal provisions and the matching degree of disputed points between cases; it simultaneously integrates the depth of legal provision citation, the number of cited provisions, and the degree of public attention to calculate the importance of case nodes. This scheme breaks through the limitations of single text matching, improves the accuracy of association strength calculation to 92.3%, achieves the coordinated quantification of legal logic and factual logic, and controls the error of social hotspot influence assessment within 5%.

[0073] Specifically, the influence of a node on other nodes is propagated by updating the weight values ​​of the edges between the direct / indirect neighboring nodes of a node. This includes: calculating the weight values ​​of the edges between the direct / indirect neighboring nodes of a node and using the calculation results as the new weight values.

[0074] Specifically, the calculation of the new weight values ​​is as follows:

[0075]

[0076] in, Let be the weight values ​​of the i-th node and the j-th node after the (t+1)-th round of propagation. Let β be the weight value between the i-th node and the j-th node after the t-th round of propagation. i Let represent the public opinion attention level of the i-th node.

[0077] Specifically, after multiple rounds of propagation, the weight values ​​of all nodes are updated. If the difference between the current weight value and the weight value of the previous round is less than the preset weight change threshold, the propagation stops.

[0078] Specifically, the legal cases corresponding to a certain node, as well as all legal cases corresponding to all nodes before the dissemination stopped, are pushed to users as the current hot legal cases.

[0079] Preferably, when a user searches for legal cases on a related topic (e.g., labor disputes), legal cases corresponding to the core nodes of all labor dispute-related legal cases will be pushed to the user. At the same time, hot legal cases in the labor dispute-related legal cases will also be pushed to the user.

[0080] Example 2

[0081] like Figure 2 As shown, this embodiment proposes a legal case push system based on core node identification, including:

[0082] The topology module is used to organize all acquired legal cases into a case graph, with each legal case as a node and edges representing the strength of the association between legal cases;

[0083] Specifically, the similarity between legal cases is calculated as the strength of association. The calculation of the similarity between legal cases involves the following steps:

[0084] R ij =α·Sim(L i ,L j )+β·Sim(T i ,T j )

[0085] Among them, R ij Let α be the association strength between the i-th node and the j-th node, and let α be the weight of the similarity between the legal provisions. Sim(L i L j ) represents the legal provision L referenced by the i-th node. i and the legal provision L referenced by the j-th node j The similarity of legal provisions, where β is the weight of topic similarity, Sim(T) i T j Let T be the topic of the i-th node. i and the topic L of the j-th node j Thematic similarity.

[0086] The core score calculation module is used to calculate the importance of a legal case as a node in the case graph based on the depth of legal citation, the number of legal citations, and the degree of social impact. Based on the strength of association and the importance of the node, the core score of the legal case is calculated.

[0087] Specifically, calculating the importance of legal cases as nodes in the case graph includes:

[0088] I i =w1·L′ i +w2·C i +w3·S i

[0089] Among them, I i Let w1 be the node importance of the i-th node, w1 be the weight of the legal citation depth, and L′ be the node importance of the ith node. i Let w2 be the depth of legal citations at the i-th node, w2 be the weight of the number of legal citations, and C be the depth of legal citations at the i-th node. i Let w3 be the number of legal citations at the i-th node, w3 be the weight of social impact, and S be the number of citations at the i-th node. i Let be the social impact score of the i-th node, where the social impact score is the level of public attention to the legal case on Weibo or forums.

[0090] Specifically, the core scores for calculating legal cases include:

[0091]

[0092] Among them, S′ i Let N(L) be the core score of the i-th node, α′1 be the weight of node importance, α′2 be the weight of association strength, and N(L) be the core score of the i-th node. i Let be the set of adjacent nodes of the i-th node.

[0093] The recommendation module is used to sort the core scores of each legal case and recommend the legal cases with the highest core scores to users.

[0094] Considering that users not only pay attention to core cases but also to trending legal cases, this embodiment also includes the following steps to recommend trending legal cases to users: (See details below)

[0095] Initialize the weight value of each edge in the case graph;

[0096] The system acquires the public opinion attention of all nodes in real time. When the public opinion attention of a certain node exceeds the preset public opinion attention threshold, the influence of that node is propagated to other nodes by updating the weight values ​​of the edges of its direct / indirect adjacent nodes. The propagation intensity increases with the public opinion attention of that node.

[0097] Specifically, the influence of a node on other nodes is propagated by updating the weight values ​​of the edges between the direct / indirect neighboring nodes of a node. This includes: calculating the weight values ​​of the edges between the direct / indirect neighboring nodes of a node and using the calculation results as the new weight values.

[0098] Specifically, the calculation of the new weight values ​​is as follows:

[0099]

[0100] in, Let be the weight values ​​of the i-th node and the j-th node after the (t+1)-th round of propagation. Let β be the weight value between the i-th node and the j-th node after the t-th round of propagation. i Let represent the public opinion attention level of the i-th node.

[0101] Specifically, after multiple rounds of propagation, the weight values ​​of all nodes are updated. If the difference between the current weight value and the weight value of the previous round is less than the preset weight change threshold, the propagation stops.

[0102] Specifically, the legal cases corresponding to a certain node, as well as all legal cases corresponding to all nodes before the dissemination stopped, are pushed to users as the current hot legal cases.

[0103] Example 3

[0104] This invention also proposes a storage medium storing multiple instructions for implementing the aforementioned legal case push method based on core node identification.

[0105] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0106] like Figure 2 As shown, optionally, in this embodiment, the storage medium is configured to store program code for performing the following steps: Step 101, forming a case graph of all acquired legal cases, with each legal case as a node and edges representing the strength of association between legal cases;

[0107] Specifically, the similarity between legal cases is calculated as the strength of association. The calculation of the similarity between legal cases involves the following steps:

[0108] R ij =α·Sim(L i ,L j )+β·Sim(T i ,T j )

[0109] Among them, R ij Let α be the association strength between the i-th node and the j-th node, and let α be the weight of the similarity between the legal provisions. Sim(L i ,L j ) represents the legal provision L referenced by the i-th node. i and the legal provision L referenced by the j-th node j The similarity of legal provisions, where β is the weight of topic similarity, Sim(T) i T j Let T be the topic of the i-th node. i and the topic L of the j-th node jThematic similarity.

[0110] Step 102: Calculate the importance of the legal case as a node in the case graph based on the depth of legal citation, the number of legal citations, and the degree of social impact of the legal case; and calculate the core score of the legal case based on the strength of association and the importance of the node.

[0111] Specifically, calculating the importance of legal cases as nodes in the case graph includes:

[0112] I i =w1·L′ i +w2·C i +w3·S i

[0113] Among them, I i Let w1 be the node importance of the i-th node, w1 be the weight of the legal citation depth, and L′ be the node importance of the ith node. i Let w2 be the depth of legal citations at the i-th node, w2 be the weight of the number of legal citations, and C be the depth of legal citations at the i-th node. i Let w3 be the number of legal citations at the i-th node, w3 be the weight of social impact, and S be the number of citations at the i-th node. i Let be the social impact score of the i-th node, where the social impact score is the level of public attention to the legal case on Weibo or forums.

[0114] Specifically, the core scores for calculating legal cases include:

[0115]

[0116] Among them, S′ i Let N(L) be the core score of the i-th node, α′1 be the weight of node importance, α′2 be the weight of association strength, and N(L) be the core score of the i-th node. i Let be the set of adjacent nodes of the i-th node.

[0117] Step 103: Sort the core scores of each legal case and recommend the legal cases with the highest core scores to the user.

[0118] Considering that users not only pay attention to core cases but also to trending legal cases, this embodiment also includes the following steps to recommend trending legal cases to users: (See details below)

[0119] Initialize the weight value of each edge in the case graph;

[0120] The system acquires the public opinion attention of all nodes in real time. When the public opinion attention of a certain node exceeds the preset public opinion attention threshold, the influence of that node is propagated to other nodes by updating the weight values ​​of the edges of its direct / indirect adjacent nodes. The propagation intensity increases with the public opinion attention of that node.

[0121] Specifically, the influence of a node on other nodes is propagated by updating the weight values ​​of the edges between the direct / indirect neighboring nodes of a node. This includes: calculating the weight values ​​of the edges between the direct / indirect neighboring nodes of a node and using the calculation results as the new weight values.

[0122] Specifically, the calculation of the new weight values ​​is as follows:

[0123]

[0124] in, Let be the weight values ​​of the i-th node and the j-th node after the (t+1)-th round of propagation. Let β be the weight value between the i-th node and the j-th node after the t-th round of propagation. i Let represent the public opinion attention level of the i-th node.

[0125] Specifically, after multiple rounds of propagation, the weight values ​​of all nodes are updated. If the difference between the current weight value and the weight value of the previous round is less than the preset weight change threshold, the propagation stops.

[0126] Specifically, the legal cases corresponding to a certain node, as well as all legal cases corresponding to all nodes before the dissemination stopped, are pushed to users as the current hot legal cases.

[0127] Example 4

[0128] This invention also proposes an electronic device, including a processor and a storage medium connected to the processor. The storage medium stores multiple instructions, which can be loaded and executed by the processor to enable the processor to execute the aforementioned method for pushing legal cases based on core node identification.

[0129] Specifically, the electronic device in this embodiment can be a computer terminal, which may include one or more processors and a storage medium.

[0130] The storage medium can be used to store software programs and modules, such as the legal case push method based on core node identification in this embodiment of the invention. The corresponding program instructions / modules are executed by the processor through running the software programs and modules stored in the storage medium, thereby performing various functional applications and data processing, thus realizing the aforementioned legal case push method based on core node identification. The storage medium may include high-speed random access storage media, and may also include non-volatile storage media, such as one or more magnetic storage systems, flash memory, or other non-volatile solid-state storage media. In some instances, the storage medium may further include storage media remotely configured relative to the processor, which can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0131] The processor can call the information and application stored in the storage medium through the transmission system to perform the following steps: Step 101, organize all the legal cases that have been acquired into a case graph, with each legal case as a node and the edges representing the strength of the association between legal cases;

[0132] Specifically, the similarity between legal cases is calculated as the strength of association. The calculation of the similarity between legal cases involves the following steps:

[0133] R ij =α·Sim(L i ,L j )+β·Sim(T i ,T j )

[0134] Among them, R ij Let α be the association strength between the i-th node and the j-th node, and let α be the weight of the similarity between the legal provisions. Sim(L i L j ) represents the legal provision L referenced by the i-th node. i and the legal provision L referenced by the j-th node j The similarity of legal provisions, where β is the weight of topic similarity, Sim(T) i T j Let T be the topic of the i-th node. i and the topic L of the j-th node j Thematic similarity.

[0135] Step 102: Calculate the importance of the legal case as a node in the case graph based on the depth of legal citation, the number of legal citations, and the degree of social impact of the legal case; and calculate the core score of the legal case based on the strength of association and the importance of the node.

[0136] Specifically, calculating the importance of legal cases as nodes in the case graph includes:

[0137] I i =w1·L′ i +w2·C i +w3·S i

[0138] Among them, I i Let w1 be the node importance of the i-th node, w1 be the weight of the legal citation depth, and L′ be the node importance of the ith node. i Let w2 be the depth of legal citations at the i-th node, w2 be the weight of the number of legal citations, and C be the depth of legal citations at the i-th node. i Let w3 be the number of legal citations at the i-th node, w3 be the weight of social impact, and S be the number of citations at the i-th node. i Let be the social impact score of the i-th node, where the social impact score is the level of public attention to the legal case on Weibo or forums.

[0139] Specifically, the core scores for calculating legal cases include:

[0140]

[0141] Among them, S′ i Let N(L) be the core score of the i-th node, α′1 be the weight of node importance, α′2 be the weight of association strength, and N(L) be the core score of the i-th node. i Let be the set of adjacent nodes of the i-th node.

[0142] Step 103: Sort the core scores of each legal case and recommend the legal cases with the highest core scores to the user.

[0143] Considering that users not only pay attention to core cases but also to trending legal cases, this embodiment also includes the following steps to recommend trending legal cases to users: (See details below)

[0144] Initialize the weight value of each edge in the case graph;

[0145] The system acquires the public opinion attention of all nodes in real time. When the public opinion attention of a certain node exceeds the preset public opinion attention threshold, the influence of that node is propagated to other nodes by updating the weight values ​​of the edges of its direct / indirect adjacent nodes. The propagation intensity increases with the public opinion attention of that node.

[0146] Specifically, the influence of a node on other nodes is propagated by updating the weight values ​​of the edges between the direct / indirect neighboring nodes of a node. This includes: calculating the weight values ​​of the edges between the direct / indirect neighboring nodes of a node and using the calculation results as the new weight values.

[0147] Specifically, the calculation of the new weight values ​​is as follows:

[0148]

[0149] in, Let be the weight values ​​of the i-th node and the j-th node after the (t+1)-th round of propagation. Let β be the weight value between the i-th node and the j-th node after the t-th round of propagation. i Let represent the public opinion attention level of the i-th node.

[0150] Specifically, after multiple rounds of propagation, the weight values ​​of all nodes are updated. If the difference between the current weight value and the weight value of the previous round is less than the preset weight change threshold, the propagation stops.

[0151] Specifically, the legal cases corresponding to a certain node, as well as all legal cases corresponding to all nodes before the dissemination stopped, are pushed to users as the current hot legal cases.

[0152] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0153] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0154] In the several embodiments provided by this invention, it should be understood that the disclosed technical content can be implemented in other ways. The system embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection between units or modules, and may be electrical or other forms.

[0155] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0156] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0157] If the integrated unit is implemented as 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 technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, optical disks, and other media capable of storing program code.

[0158] The method of this invention triggers multi-level weighted propagation when the public attention to a case exceeds a threshold: the edge weights of adjacent nodes are exponentially amplified according to the intensity of public opinion until the edge weight update amplitude is less than a set threshold, at which point the propagation terminates; the source case and the affected cases before termination are pushed as a hotspot set. This mechanism reduces the timeliness of hotspot case identification from 72 hours to 10 minutes, accurately depicts the scope of influence, improves the completeness of derivative case pushes by 89%, and constructs a real-time response system for legal public opinion.

[0159] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for pushing legal cases based on core node identification, characterized in that, include: All acquired legal cases are organized into a case graph, with each legal case as a node and edges representing the strength of the relationship between legal cases. Initialize the weight value of each edge in the case graph; The system acquires the public opinion attention of all nodes in real time. When the public opinion attention of a certain node exceeds the preset public opinion attention threshold, the influence of the node is propagated to other nodes by updating the weight values ​​of the edges of the direct / indirect adjacent nodes of the node. The propagation intensity increases with the public opinion attention of the node. Based on the depth of legal citation, the number of legal citations, and the degree of social impact of legal cases, the importance of legal cases in the case graph is calculated. Based on the strength of association and the importance of nodes, the core score of legal cases is calculated. The closer the cited legal citation is to the root legal citation, the greater the depth of legal citation. The system ranks the core scores of each legal case and recommends the legal cases with the highest core scores to users.

2. The legal case push method based on core node identification as described in claim 1, characterized in that, The similarity between legal cases is calculated as the strength of association. Specifically, the calculation of the similarity between legal cases involves: in, For the first The node and the first The strength of the association between nodes The weight for the similarity of legal provisions. For the first The legal provisions referenced by each node and the The legal provisions referenced by each node Similarity of legal provisions Weights for topic similarity For the first The theme of each node and the The theme of each node Thematic similarity.

3. The legal case push method based on core node identification as described in claim 1, characterized in that, Calculating the importance of legal cases as nodes in the case graph includes: in, For the first The importance of each node. Weighting for the depth of legal citation. For the first The depth of legal citations at each node. Weighting based on the number of legal citations. For the first The number of legal citations for each node. As a weight for social impact, For the first The social impact of each node refers to the level of public attention paid to legal cases on Weibo or forums.

4. The legal case push method based on core node identification as described in claim 1, characterized in that, The core scoring criteria for legal cases include: in, For the first The core score of each node, Weights representing the importance of nodes. The weights for the strength of the association. For the first The set of adjacent nodes of a node.

5. The legal case push method based on core node identification as described in claim 4, characterized in that, The influence of a node on other nodes is propagated by updating the weight values ​​of the edges between the direct / indirect neighboring nodes of a node. This includes: calculating the weight values ​​of the edges between the direct / indirect neighboring nodes of a node and using the calculation results as the new weight values.

6. The legal case push method based on core node identification as described in claim 5, characterized in that, The new weight values ​​are calculated as follows: in, For the first The node and the first The node at the th The weight values ​​after round propagation. For the first The node and the first The node at the th The weight values ​​after round propagation. For the first Public opinion attention at each node.

7. The legal case push method based on core node identification as described in claim 6, characterized in that, Also includes: After multiple rounds of propagation, the weight values ​​of all nodes are updated. If the difference between the current weight value and the weight value of the previous round is less than the preset weight change threshold, the propagation stops.

8. The legal case push method based on core node identification as described in claim 7, characterized in that, The system pushes legal cases corresponding to a specific node, as well as legal cases corresponding to all nodes before the dissemination stopped, to users as currently trending legal cases.

9. A legal case push system based on core node identification, characterized in that, include: The topology module is used to organize all acquired legal cases into a case graph, with each legal case as a node and edges representing the strength of the association between legal cases; Initialize the weight value of each edge in the case graph; The system acquires the public opinion attention of all nodes in real time. When the public opinion attention of a certain node exceeds the preset public opinion attention threshold, the influence of the node is propagated to other nodes by updating the weight values ​​of the edges of the direct / indirect adjacent nodes of the node. The propagation intensity increases with the public opinion attention of the node. The core score calculation module is used to calculate the importance of a legal case in the case graph based on the depth of legal citation, the number of legal citations, and the degree of social impact. It also calculates the core score of the legal case based on the strength of association and the importance of the node. The closer the cited legal citation is to the root legal citation, the greater the depth of legal citation. The recommendation module is used to sort the core scores of each legal case and recommend the legal cases with the highest core scores to users.

Images