Judgment method and device for case jurisdiction court and related equipment

Abstract

The embodiment of the invention discloses a judgment method and device for a case jurisdiction court and related equipment. The method comprises the following steps: acquiring court information, court jurisdiction rules, preset constraint rules and court portrait information to construct a jurisdiction knowledge base; obtaining case information of a current case, and performing analysis and structured processing on the case information to generate a case fact atlas; based on the case fact atlas, detecting whether the current case has an effective jurisdiction agreement; determining a compliance candidate court set according to the effective jurisdiction protocol and the jurisdiction knowledge base; and performing matching calculation on the basis of the court portrait information of each compliance candidate court in the compliance candidate court set and the case fact atlas to obtain the matching degree of each compliance candidate court, and performing descending sorting on the compliance candidate courts in the compliance candidate court set according to the matching degree to generate a court recommendation result. The method improves the efficiency of judging the case jurisdiction court.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of artificial intelligence technology, and in particular to a method, apparatus and related equipment for determining the court with jurisdiction over a case. Background Technology

[0002] With the construction of smart court systems, the intelligentization of case filing and triage has become key to improving judicial efficiency. However, in actual legal service scenarios, users (the public or lawyers) often describe cases in unstructured natural language, such as "I have business in both City A and City B, but the other party's company is in City A, and the debt occurred in City B."

[0003] Existing jurisdiction determination technologies suffer from the following key pain points: 1. The “separation” between rules and semantics: Traditional expert systems rely on rigid “keyword matching” (such as IF-THEN rules), which is difficult to handle ambiguous semantics such as “place of habitual residence” and “place of actual performance”, resulting in poor generalization ability; while classification models based on pure deep learning (such as BERT / LLM) output probabilistic results, which cannot guarantee that the results must comply with the rigid constraints of legal provisions such as the Civil Procedure Law (such as the exclusivity of exclusive jurisdiction), leading to “inexplicable” or “illegal filing” in legal logic.

[0004] 2. Missing elements lead to reasoning interruption: User input is often incomplete (e.g., the target amount is not mentioned). Existing systems lack a mechanism to dynamically detect missing elements and guide their completion, resulting in the inability to make accurate judgments.

[0005] 3. Lack of interpretable reasoning path: The "black box" model cannot output a logical deduction chain similar to a judge's judgment (such as "because the amount in dispute exceeds 50 million and it is a foreign-related case, it is under the jurisdiction of the intermediate court"), which makes it difficult to meet the requirements of judicial scenarios for decision-making transparency. Summary of the Invention

[0006] The embodiments of the present invention provide a method, apparatus and related equipment for determining the court with jurisdiction over a case, aiming to solve the technical problem that it is difficult to prevent the leakage of sensitive data while ensuring the availability of semantic retrieval in traditional technologies.

[0007] In a first aspect, embodiments of the present invention provide a method for determining the court with jurisdiction over a case, comprising: Obtain court information, court jurisdiction rules, pre-set constraint rules, and court profile information, and construct a jurisdiction knowledge base based on the aforementioned court information, court jurisdiction rules, pre-set constraint rules, and court profile information; Obtain the case information of the current case, and parse and structure the case information to generate a case fact map; Based on the aforementioned case fact map, it is determined whether a valid jurisdiction agreement exists for the current case. The jurisdiction agreement is agreed upon by both parties and includes at least the name of the court with jurisdiction or the geographical scope of the jurisdiction. If a valid jurisdiction agreement exists, the corresponding court with jurisdiction shall be determined in accordance with the jurisdiction agreement. If the jurisdiction agreement has expired or no jurisdiction agreement is detected, then jurisdiction rules are retrieved and constraint rules are determined in a pre-set jurisdiction knowledge base based on the case fact graph to obtain a set of constraint rules; a set of candidate courts for the current case is determined based on the relational regions in the case fact graph; candidate courts that do not conform to the rules in the candidate court set and the courts under the agreement jurisdiction are filtered out according to the set of constraint rules to obtain a set of compliant candidate courts; the court profile information of each compliant candidate court in the set of compliant candidate courts is matched with the case fact graph to obtain the matching degree of each compliant candidate court, and the compliant candidate courts in the set of compliant candidate courts are sorted in descending order according to the matching degree to generate court recommendation results.

[0008] Secondly, embodiments of the present invention provide a device for determining the jurisdiction of a court in a case, comprising: The acquisition module is used to acquire court information, court jurisdiction rules, preset constraint rules, and court profile information, and to construct a jurisdiction knowledge base based on the court information, court jurisdiction rules, preset constraint rules, and court profile information; The parsing module is used to obtain the case information of the current case, and to parse and structure the case information to generate a case fact map; The detection module is used to detect whether there is a valid jurisdiction agreement for the current case based on the case fact map. The jurisdiction agreement is agreed upon by both parties and the content of the jurisdiction agreement includes at least the name of the court with jurisdiction or the scope of the jurisdiction. The first determining module is used to determine the corresponding court of jurisdiction based on the jurisdiction agreement if a valid jurisdiction agreement exists. The matching module is used to, if the jurisdiction agreement has expired or no jurisdiction agreement is detected, perform jurisdiction rule retrieval and constraint rule determination in a pre-set jurisdiction knowledge base based on the case fact graph to obtain a set of constraint rules; determine a set of candidate courts for the current case based on the relational regions in the case fact graph; filter out candidate courts that do not conform to the rules in the candidate court set and the courts under the agreement jurisdiction according to the set of constraint rules to obtain a set of compliant candidate courts; perform matching calculations on the court profile information of each compliant candidate court in the set of compliant candidate courts and the case fact graph to obtain the matching degree of each compliant candidate court; and sort the compliant candidate courts in the set of compliant candidate courts in descending order according to the matching degree to generate court recommendation results.

[0009] Thirdly, embodiments of the present invention provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for determining the court of jurisdiction described in the first aspect above.

[0010] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the method for determining the jurisdiction of a court described in the first aspect.

[0011] This invention provides a method, apparatus, and related equipment for determining the court of jurisdiction in a case. The method involves: constructing a jurisdiction knowledge base by acquiring court information, court jurisdiction rules, preset constraint rules, and court profile information; acquiring case information for the current case and parsing and structuring the case information to generate a case fact graph; detecting whether a valid jurisdiction agreement exists for the current case based on the case fact graph; determining the corresponding court of jurisdiction based on the agreement if a valid jurisdiction agreement exists; and removing candidate courts that do not conform to the rules from the candidate court set and the court of jurisdiction agreement based on the constraint rule set, based on the case fact graph; performing a matching calculation between the court profile information of each compliant candidate court in the compliant candidate court set and the case fact graph to obtain the matching degree of each compliant candidate court, and sorting the compliant candidate courts in the compliant candidate court set in descending order according to the matching degree to generate a court recommendation result. This method reduces the workload of manual retrieval and repeated comparisons by extracting case elements and matching jurisdiction rules, and recommends relevant courts for the current case according to the matched jurisdiction rules, thereby improving the efficiency of determining the court with jurisdiction over the case. Attached Figure Description

[0012] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 A flowchart illustrating an embodiment of the method for determining the court of jurisdiction in a case, provided by an embodiment of the present invention; Figure 2 A schematic block diagram of a case jurisdiction determination device provided in an embodiment of the present invention. Detailed Implementation

[0014] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0016] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0017] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0018] Please see Figure 1 This is a flowchart illustrating a method for determining the court of jurisdiction in a case, as provided in an embodiment of the present invention. The method includes steps S110 to S132.

[0019] Step S110: Obtain the case information of the current case, and parse and structure the case information to generate a case fact map; In this embodiment, the system receives case information for the current case. Input formats include: case description text in natural language, case element data in structured form, and attached documents such as contracts and agreements, or their text summaries. The system supports any combination of the above input formats. When multiple input formats exist simultaneously, structured data takes precedence, while natural language descriptions and attachment content serve as supplementary information sources. Case information includes at least: 1. Identifying the case type and cause of action: Analyzing the case information to identify the dispute type and cause of action to which the case belongs. When the user directly provides the cause of action, the validity of the cause of action code is verified and adopted directly. When the user does not provide the cause of action, the following method is used for identification: extracting keywords and entities from the case description, matching them with the cause of action knowledge base, and outputting the cause of action code with the highest matching degree and its confidence level. If the confidence level is lower than the threshold, a list of candidate causes of action is output for the user to confirm. 2. Extract information about the parties involved, including relevant information for the plaintiff and defendant: Subject type (natural person, legal person, unincorporated organization); subject name; domicile or registered address (accurate to the provincial, municipal, or county level); whether it is a foreign-related subject (based on nationality and whether the registered address is outside of China); whether it is a special subject (including military units, classified units, diplomatic missions, etc.). Generate party nodes from the extracted results and establish relationship edges between party nodes and case nodes, with relationship types including "plaintiff" and "defendant". 3. Extract information about the subject matter: Extract the type and monetary value of the disputed subject matter. Subject matter types include: monetary claims, movable property, immovable property, intellectual property, equity, debt, and other property rights. The method for extracting the subject matter amount includes: direct user input, extraction from the litigation claim, and deduction from the contract amount. When multiple amounts exist, the litigation claim amount is used as the subject matter amount. When the subject matter amount cannot be precisely determined, extract the order of magnitude range of the amount and indicate the source and confidence level of the amount. 4. Extract relevant geographical regions for the current case: Depending on the case type, relevant geographical regions include: (1) the defendant's domicile, extracted from the party information; (2) the place of contract signing, extracted from the contract text or case description; (3) the place of contract performance, extracted according to the contract agreement, or presumed according to legal provisions if there is no agreement; (4) the place where the tortious act was committed, extracted from the case description; (5) the place where the tortious result occurred, extracted from the case description; (6) the location of the real estate, extracted from the subject matter information; (7) the company's domicile, extracted from the party information; and (8) the location of the estate, extracted from the case description. Create a geographical node for each relevant geographical region, record the administrative division code, and establish a relationship edge between the geographical node and the case node, with the relationship type being the corresponding geographical region type.

[0020] Furthermore, identify whether there are any special jurisdictional elements in the current case: (1) Real estate property rights elements, the judgment standard is whether the case dispute involves the confirmation of ownership, division, or neighboring relations of real estate; (2) Port operation elements, the judgment standard is whether the case involves disputes over port loading and unloading, warehousing, etc.; (3) Inheritance elements, the judgment standard is whether the case is an inheritance dispute and the estate is real estate or the location of the main estate is clearly defined; (4) Corporate litigation elements, the judgment standard is whether the case involves the establishment of a company, confirmation of shareholder qualifications, dissolution, division, or merger of a company; (5) Intellectual property elements, the judgment standard is whether the case involves the establishment of a company, confirmation of shareholder qualifications, dissolution, division, or merger of a company; The criteria for judgment are whether the case involves intellectual property disputes such as patents, trademarks, copyrights, and trade secrets; (6) Maritime and commercial elements, the criteria for judgment are whether the case involves maritime disputes such as maritime transportation, ship collisions, and marine pollution; (7) Negotiable instruments elements, the criteria for judgment are whether the case involves negotiable instruments disputes; (8) Financial elements, the criteria for judgment are whether the case involves financial disputes such as securities, futures, trusts, and insurance and falls within the jurisdiction of the financial court; (9) Internet elements, the criteria for judgment are whether the case involves Internet disputes such as online shopping, online service contracts, and online infringement and falls within the jurisdiction of the Internet court.

[0021] Based on special jurisdictional elements, it is determined whether a case falls under the jurisdiction of a special court. If so, the court with jurisdiction over the case is determined according to the relevant regulations for the special jurisdictional elements. Specifically, cases involving technology-related intellectual property disputes such as patents, new plant varieties, integrated circuit layout designs, trade secrets, and computer software, or monopoly disputes, are under the jurisdiction of the intellectual property court or an intermediate court with jurisdiction over intellectual property cases. Cases involving maritime tort disputes, maritime contract disputes, marine development and utilization disputes, and ship disputes are under the jurisdiction of the maritime court. Cases involving financial civil and commercial disputes where the parties' domicile or place of contract performance is within the jurisdiction of the financial court are under the jurisdiction of the financial court. Cases involving online shopping contract disputes, online service contract disputes, online tort disputes, and internet finance disputes that fall within the jurisdiction of the internet court are under the jurisdiction of the internet court.

[0022] The extracted and identified case information is integrated into a case fact graph. The graph structure includes: (1) a case root node, associated with the case number, cause of action code, case type label, amount in dispute and its confidence level; (2) a set of party nodes, each associated with subject information and attribute labels; (3) a set of relational region nodes, each associated with the regional administrative division code and relational region type; (4) a set of special element labels, recording the identified special jurisdictional elements; (5) a set of uncertainty labels, recording all pending elements and low confidence information; and (6) graph edges represent the relationships between nodes, supporting subsequent rule matching and path backtracking.

[0023] In addition, when case elements cannot be extracted or the confidence level of the extraction results is insufficient, the following procedures shall be followed: For any missing key elements, targeted follow-up questions are generated to explain to the user the impact of that element on jurisdictional determination and request its supplementation. Key elements include: the defendant's domicile, the amount in dispute, the case type, and the legally defined geographical area corresponding to the case type.

[0024] For information that users cannot immediately supplement, mark the element as "pending" and record the type of uncertainty (missing, ambiguous, contradictory) and the current confidence score.

[0025] For contradictory information (such as discrepancies between case descriptions and contract terms), record the contradictions and indicate that user confirmation is required. Pass this uncertainty information to subsequent steps as input for risk assessment, rather than terminating the process.

[0026] Step S120: Based on the case fact map, detect whether there is a valid jurisdiction agreement for the current case; Step S131: If a valid jurisdiction agreement exists, determine the corresponding court of jurisdiction according to the jurisdiction agreement; In this embodiment, priority is given to detecting whether a valid jurisdiction agreement exists in the current case, reflecting the principle of party autonomy in civil litigation. A jurisdiction agreement is agreed upon by both parties and its content must at least include the name of the court with jurisdiction or the geographical scope of jurisdiction. The existence of jurisdiction agreement clauses is detected in the case information, including: dispute resolution clauses in contract texts, independent jurisdiction agreement documents, jurisdiction agreements in electronic texts, and jurisdiction clauses in supplementary agreements. Detection methods include: keyword matching detection, searching for keywords such as "jurisdiction," "court," "litigation," and "dispute resolution," and their combinations; semantic understanding detection, identifying clauses expressing the meaning of the jurisdiction agreement, including non-standard expressions. If a jurisdiction agreement clause is detected, the agreed content is extracted, including the agreed court name or scope, the agreed geographical scope, and the agreed case type scope. The validity of the detected jurisdiction agreement is verified, and the verification requirements include: Formal requirements verification determines whether the agreement is in written form or other tangible form that can represent the content, including contracts, letters, telegrams, telexes, faxes, electronic data interchange, emails, etc.

[0027] The clarity requirement is verified to determine whether the court stipulated in the agreement is clear or definite, excluding situations where the agreement is unclear, multiple courts are stipulated, and the selection rules cannot be determined.

[0028] The actual connection requirement is verified to determine whether the agreed court location has an actual connection with the dispute. The actual connection location includes the defendant's domicile, the place of performance of the contract, the place of signing the contract, the plaintiff's domicile, and the location of the subject matter.

[0029] Mandatory requirement verification is imposed to determine whether the agreement violates the provisions on hierarchical jurisdiction and exclusive jurisdiction. Hierarchical jurisdiction restriction: The court level selected in the agreement cannot be lower than the level determined by statutory hierarchical jurisdiction. Exclusive jurisdiction restriction: Cases under exclusive jurisdiction cannot have their jurisdiction changed through the agreement.

[0030] The agreement is verified to confirm its authenticity, determining whether there are any invalid standard clauses, fraud, coercion, or other circumstances that would render the agreement invalid. If the user claims that the agreement has defects in validity, these are marked.

[0031] In one embodiment, the status of the governance protocol is output based on the verification result: Status 1 is "valid". The agreement has passed all the verification requirements, and the agreed court scope is output as a high-priority recommendation result.

[0032] Status 2 is "invalid". The agreement failed a certain verification requirement. The type of invalidity reason is recorded (non-compliant form, unclear agreement, no actual connection, violation of mandatory provisions, defect of agreement). The agreement does not stipulate that it is included in the jurisdiction determination, but it is recorded as a risk factor.

[0033] Status 3 is "Uncertain". The result of a certain verification requirement cannot be determined (such as disputes over actual contact or doubts about the validity of standard terms). The reason for the uncertainty is recorded, and the agreement is included as a possible valid candidate in the subsequent process. At the same time, the risk of objection is marked.

[0034] Step S132: If the jurisdiction agreement has expired or no jurisdiction agreement is detected, then based on the case fact graph, a jurisdiction rule retrieval and constraint rule determination are performed in a preset jurisdiction knowledge base to obtain a set of constraint rules; based on the relational regions in the case fact graph, a set of candidate courts for the current case is determined; candidate courts that do not conform to the rules in the set of candidate courts are filtered out according to the set of constraint rules to obtain a set of compliant candidate courts; based on the court profile information of each compliant candidate court in the set of compliant candidate courts and the case fact graph, a matching calculation is performed to obtain the matching degree of each compliant candidate court; and the compliant candidate courts in the set of compliant candidate courts are sorted in descending order according to the matching degree to generate a court recommendation result.

[0035] In this embodiment, court information, court jurisdiction rules, pre-defined constraint rules, and court profile information are obtained in advance, and a jurisdiction knowledge base is constructed based on these information. Specifically, this includes: Step 1: Construct a basic court information database. Each court is treated as a node, associated with the following information: court name, unique court code, court level (basic people's court, intermediate people's court, higher people's court, supreme people's court), jurisdictional area, administrative division code, special court attribute identifiers (ordinary court, intellectual property court, maritime court, financial court, internet court, etc.), and a set of identifiers for types of cases that can be accepted. An index structure is established for rapid retrieval: the first index maps "administrative division code to court node set," the second index maps "court level to court node set," the third index maps "special court type to court node set," and the fourth index maps "case type to court node set with jurisdiction."

[0036] Step Two: Compile a set of normative documents related to jurisdiction. Collect normative texts related to jurisdiction determination, including: the Civil Procedure Law of the People's Republic of China and its judicial interpretations, the Supreme People's Court's provisions on hierarchical jurisdiction, the provisions of various Higher People's Courts on standards for hierarchical jurisdiction, the provisions on the jurisdictional scope of special courts, the provisions on jurisdiction in foreign-related civil litigation, and the provisions on jurisdiction in special types of cases. Mark each normative text with the following attributes: normative name, issuing authority, issuance date, effective date, expiration date (if any), applicable geographical scope, and applicable case types.

[0037] Step 3: Construct the jurisdictional norm ontology and structured rule base.

[0038] The specification text is parsed into structured rule entries, each rule entry containing the following fields: The rule number is used to uniquely identify the rule.

[0039] Rule types, with possible values ​​including: exclusive jurisdiction rule, hierarchical jurisdiction rule, agreed jurisdiction rule, general territorial jurisdiction rule, special territorial jurisdiction rule, special court jurisdiction rule, and foreign-related jurisdiction rule.

[0040] Triggering conditions describe the preconditions under which the rule applies. They are expressed in a structured manner and include case type conditions, subject conditions, subject matter conditions, geographical conditions, monetary conditions, etc. The logical relationships between the conditions are "AND" or "OR".

[0041] Jurisdiction conclusion describes the scope of courts with jurisdiction determined by the rule, including three types: court level limitation, geographical limitation, and specific court limitation.

[0042] Legal basis: Record the legal provision number and summary of the rule.

[0043] Scope of application: Record the geographical and time range in which this rule applies.

[0044] A jurisdiction normative ontology graph is constructed, with nodes including: factual element nodes (case type, party attributes, subject matter attributes, geographical relationship, etc.), rule requirement nodes, and jurisdiction conclusion nodes (court level, geographical scope, specific court). Graph edges represent relationships between nodes, including "satisfaction" relationships (factual elements satisfy rule requirements), "derivation" relationships (rule requirements derive jurisdiction conclusions), "conflict" relationships (conflicts between different rule conclusions), and "priority" relationships (priority relationships between rules).

[0045] Step 4: Define the constraint types and conflict resolution mechanisms Each rule is assigned a constraint type, which is divided into two categories: hard constraints and soft constraints. Hard constraints are mandatory rules that must be followed, and courts must exclude rules that are violated. Soft constraints are optional rules, and parties may choose one when multiple soft constraints exist.

[0046] Set rule priorities, from highest to lowest: First priority: Exclusive jurisdiction rules, which are hard constraints and exclude the parties' choice of jurisdiction or other statutory jurisdiction.

[0047] Second priority: the rule of jurisdiction by level, which is a hard constraint, and once the court level is determined, it cannot be changed by agreement between the parties.

[0048] The third priority is the jurisdictional agreement rule, which is a hard constraint. Under the premise of not violating exclusive jurisdiction and hierarchical jurisdiction, a valid jurisdictional agreement excludes statutory territorial jurisdiction.

[0049] Fourth priority: Special jurisdiction rules, which are soft constraints, are chosen by the plaintiff when they conflict with general territorial jurisdiction.

[0050] Fifth priority: General territorial jurisdiction rules, which are soft constraints and serve as a catch-all rule.

[0051] Establish a conflict resolution matrix, where rows and columns represent different rule types, and matrix elements represent conflict resolution results. Resolution results include: higher priority rule excludes lower priority rule; two competing rules are chosen by the parties; and two rules are applied in parallel, and their intersection is taken.

[0052] The specific rules for adjudication are as follows: When exclusive jurisdiction conflicts with any other rule, exclusive jurisdiction applies. When hierarchical jurisdiction conflicts with agreed jurisdiction, any agreement in the agreed jurisdiction that violates hierarchical jurisdiction is invalid. Hierarchical jurisdiction and territorial jurisdiction do not conflict; the intersection is taken after determining the court level and territorial scope separately. When agreed jurisdiction is valid, statutory territorial jurisdiction is excluded. When special territorial jurisdiction and general territorial jurisdiction conflict, the plaintiff may choose one.

[0053] Step 5: Construction of the Court Profile Model S5a: Collection of Judgment Document Data Raw data was collected from publicly available court judgments, including data from the China Judgments Online website, publicly available judgments from local courts, and legal database data obtained with authorization.

[0054] The collected court judgments should cover the main types of cases, span at least the past three years, and the total amount of data should meet the statistical significance requirement.

[0055] S5b: Structured Analysis of Judgment Documents The collected court documents were processed using natural language processing and structured parsing to extract the following fields: Basic case information fields: case number, cause of action, case type code, date of filing, date of closing, and trial procedure (ordinary procedure for first instance, simplified procedure for first instance, procedure for second instance, and procedure for retrial).

[0056] Court information fields: name of the court hearing the case, unique court code, court level, and administrative division to which the court belongs.

[0057] The parties information fields include: number of plaintiffs, plaintiff type (natural person, legal person, unincorporated organization), number of defendants, defendant type, whether the case involves foreign parties, and whether special entities are involved.

[0058] Subject matter information fields: amount claimed, amount supported by judgment, and range code for the subject matter amount.

[0059] Case closure information fields: Case closure method (judgment, mediation, withdrawal of lawsuit, dismissal of lawsuit, transfer of jurisdiction, other), type of judgment result (plaintiff wins, defendant wins, partial support, dismissal of lawsuit), whether an appeal was filed, and the second instance result of the appealed case (upholding, reversal, remand for retrial, mediation, withdrawal of lawsuit).

[0060] Trial period field: Trial days (case closing date minus case filing date), deducting trial time limit days (if any), actual trial days.

[0061] S5c: Data Cleaning and Standardization The parsed data is cleaned to remove records with missing key fields, including case number, cause of action, court of trial, date of filing, date of closing, and method of closing.

[0062] Standardize the mapping of the cause of action field to uniformly map different expressions of cause of action to the standard cause of action code system.

[0063] Standardize the mapping of the court name field and associate the court name with the unique court code in the S101 court basic information database.

[0064] The amount field is normalized to unify the currency unit and divide the amount into ranges. The ranges are divided as follows: less than 10,000 yuan, 10,000 yuan to 100,000 yuan, 100,000 yuan to 500,000 yuan, 500,000 yuan to 1 million yuan, 1 million yuan to 5 million yuan, 5 million yuan to 10 million yuan, 10 million yuan to 50 million yuan, 50 million yuan to 100 million yuan, and more than 100 million yuan.

[0065] Outlier handling is performed on the trial period field, and records with negative trial periods or those exceeding a reasonable range (such as more than three years) are removed.

[0066] S5d: Definition and Indicator Calculation of Court Profiling Dimensions The system defines the dimensions of the court profile, with each dimension containing several specific indicators.

[0067] The first dimension is the trial efficiency dimension, which includes the following indicators: The average trial period indicator is calculated as the arithmetic mean of the number of days it takes to process all cases in the court.

[0068] The average trial period index for each cause of action is calculated by first grouping cases by cause of action and then taking the arithmetic mean of the number of days each group of cases takes to complete, thus creating a mapping table from "court plus cause of action" to "average trial period".

[0069] The simplified procedure application rate indicator is calculated by dividing the number of cases applying the simplified procedure by the total number of first-instance cases.

[0070] The case completion rate within the statutory time limit is calculated by dividing the number of cases completed within the statutory time limit by the total number of cases.

[0071] The second dimension is the case experience dimension, which includes the following indicators: The case volume index is calculated as the total number of cases of a certain cause of action heard by the court.

[0072] The case concentration index is calculated by dividing the number of cases for a specific case type by the total number of cases in that court.

[0073] The indicator for handling complex cases is calculated by dividing the number of cases with a value of over one million yuan by the total number of cases.

[0074] The experience indicator for foreign-related cases is calculated as the number of foreign-related cases and their proportion of the total number of cases.

[0075] The third dimension is the referee quality dimension, which includes the following indicators: The mediation settlement rate is calculated by dividing the number of cases settled through mediation by the total number of cases settled.

[0076] The withdrawal rate is calculated by dividing the number of withdrawn cases by the total number of cases.

[0077] The appeal rate is calculated by dividing the number of appealed cases by the number of cases adjudicated in the first instance.

[0078] The second instance upholding rate is calculated by dividing the number of cases in which the original judgment was upheld by the total number of appealed cases.

[0079] The fourth dimension is the convenience of litigation, which includes the following indicators: The online case filing support index is categorized as either "support" or "dissatisfaction," with data sourced from publicly available court information or actual test results.

[0080] The electronic service of process support index is rated as either "support" or "dissatisfaction," and the data source is publicly available information from the court's official website.

[0081] The online court hearing support index is categorized as either "support" or "dissatisfaction," with data sourced from publicly available court information or actual test results.

[0082] The cross-jurisdictional case filing support index is rated as either "support" or "dissatisfaction," and the data source is publicly available information from the courts.

[0083] S5e: Court Image Vector Generation The above indicators are aggregated according to the court dimension to generate a court profile vector. The structure of the court profile vector is: a unique court code, an overall profile vector, and a case-specific profile matrix.

[0084] The overall profile vector contains normalized values ​​of indicators for each dimension. The normalization method uses min-max normalization to map each indicator value to the range of zero to one.

[0085] The behavioral case type codes in the case type profile matrix are listed as indicators for each dimension, and the matrix elements are the normalized values ​​of the indicators for that court under that case type.

[0086] For courts or case combinations where the amount of data is insufficient to form stable statistics, the data confidence level is marked as low, and the data is downweighted in subsequent use.

[0087] S5f: Image Model Update Mechanism Define the update strategy for the portrait model: The full update strategy involves recalculating all metrics and profile vectors using all historical data every quarter.

[0088] The incremental update strategy involves collecting newly added court judgment data weekly and updating the profile vector using a sliding window method with a sliding window length of three years.

[0089] The version management mechanism saves the portrait model version number and update timestamp after each update, and supports historical version rollback.

[0090] In this embodiment, when no jurisdiction agreement is detected, or the verification result of the jurisdiction agreement is invalid, jurisdiction rules are retrieved and constraint rules are determined based on the case fact graph in a pre-set jurisdiction knowledge base to obtain a set of hard constraints, providing a basis for candidate court selection. Specifically, using the case fact graph as input, potentially applicable constraint rules are retrieved from the rule knowledge base. Retrieval anchors include: case type tags, cause of action codes, party attribute tags, subject matter type, special jurisdiction element tags, and territorial relationship territorial type. The retrieval method involves traversing the rule entries in the rule base, matching the triggering conditions of the rule with the case fact graph, and if all or part of the triggering conditions are met (when there are pending elements), the rule is included in the candidate rule set. For each candidate rule, the matching status is recorded: complete match (all conditions are met), partial match (pending conditions exist), and matching degree score.

[0091] Further, determining whether the current case falls under the jurisdiction of a specific court is known as the exclusive jurisdiction rule. Exclusive jurisdiction rules are identified from the candidate rules. The triggering conditions for exclusive jurisdiction rules include: lawsuits arising from real estate disputes (jurisdiction of the court where the real estate is located); lawsuits arising from disputes occurring during port operations (jurisdiction of the court where the port is located); lawsuits arising from inheritance disputes (jurisdiction of the court where the deceased resided at the time of death or where the main estate is located); and lawsuits arising from disputes related to company establishment, confirmation of shareholder status, profit distribution, and dissolution (jurisdiction of the court where the company is located). If an exclusive jurisdiction rule is identified and the triggering conditions are met, the jurisdictional conclusion of that rule is treated as the highest priority hard constraint, excluding jurisdiction agreed upon by the parties and other statutory territorial jurisdictions.

[0092] Further, determine which level of court jurisdiction the current case falls under. First, determine if there are any case types explicitly stipulated by law or judicial interpretation to be under the jurisdiction of a specific level of court. In intellectual property cases, patents, new plant varieties, integrated circuit layout designs, trade secrets, computer software, and monopoly disputes are under the jurisdiction of intermediate people's courts or intellectual property courts. Maritime and commercial cases are under the jurisdiction of maritime courts, which are specialized intermediate courts. If no such special provisions exist, the level is determined based on the amount in dispute and the jurisdictional standards of the location of the court handling the case. The method for querying jurisdictional standards is as follows: using the possible jurisdictional region of the case (such as the province where the defendant resides) and the amount in dispute as input, query the jurisdictional standards published by the higher people's court of that province, and output the court level that should have jurisdiction. When the amount in dispute is uncertain, output the possible set of level domains and the corresponding amount conditions for each level. For example: amounts in dispute below 500,000 yuan are under the jurisdiction of basic-level courts; amounts in dispute between 500,000 yuan and 100 million yuan are under the jurisdiction of intermediate courts.

[0093] Integrating the above identification results, a set of constraint rules is generated. The set of constraint rules includes: Exclusive jurisdiction constraints record the geographical scope defined by the exclusive jurisdiction rules.

[0094] Level jurisdiction constraints record the court level or level domain determined by the level jurisdiction.

[0095] Special court constraints record constraints that should be under the jurisdiction of a special court and the type of special court.

[0096] Jurisdiction by agreement (if any) records the scope of courts stipulated in a valid agreement, but is less effective than exclusive jurisdiction and hierarchical jurisdiction.

[0097] Perform a consistency check on the set of constraint rules: if there is a conflict between the agreed jurisdiction and the exclusive jurisdiction or the hierarchical jurisdiction, the conflicting part of the agreed jurisdiction constraint is invalid, and the non-conflicting part is retained; if there is a logical contradiction among the constraint rules, report the constraint conflict and output possible interpretation paths.

[0098] Based on the geographical relationships within the case elements, a set of candidate courts is generated. Specifically, the applicable jurisdictional rules are selected according to the case type. For contract disputes: the first geographical relationship is the defendant's domicile, and the second geographical relationship is the place of performance of the contract. The rules for determining the place of performance are: if the contract stipulates the place of performance, the stipulation shall prevail; if the contract does not stipulate or the stipulation is unclear, the statutory presumption rule applies according to the contract type. For disputes involving monetary payments, the place of performance is the location of the party receiving the money; for disputes involving the delivery of real estate, the place of performance is the location of the real estate; for other disputes, the place of performance is the location of the party performing the obligation. For tort disputes: the first geographical relationship is the defendant's domicile, and the second geographical relationship is the place of the tort. The place of the tort includes the place where the tortious act was committed and the place where the tortious result occurred. For corporate litigation: the jurisdiction is the company's domicile (exclusive jurisdiction). For inheritance disputes: the jurisdiction is the deceased's domicile at the time of death or the location of the principal estate (exclusive jurisdiction). For labor disputes: the jurisdiction is the employer's location or the place of performance of the labor contract. The administrative division codes of relevant regions are mapped to candidate regions. The mapping method is as follows: taking the administrative division codes of the relevant regions as input, the administrative division index of the rule knowledge base is queried, and the set of candidate courts within the scope of the administrative division is output.

[0099] The candidate court set is filtered based on a set of constraints: if an exclusive jurisdiction constraint exists, the candidate court set retains only courts within the exclusive jurisdiction's geographical area, eliminating courts from other areas; courts that do not meet the requirements for hierarchical jurisdiction are also eliminated; if a case should be under the jurisdiction of a special court, special courts of the corresponding type are eliminated; and courts that do not have jurisdiction over the case type are eliminated, resulting in a compliant candidate court set. This embodiment ensures that the results absolutely comply with the rigid provisions of the law (avoiding basic legal errors) and utilizes deep learning to solve the optimal court recommendation problem under complex semantics.

[0100] Furthermore, profile vectors of each compliant candidate court are extracted from the rule knowledge base. For each candidate court, a matching score is calculated between the current case feature vector and the court profile vector. The matching score is calculated using a dimension-weighted summation method, where the matching score equals the sum of the products of each dimension's score and its corresponding weight. For example, the dimensions of trial efficiency, case experience, judgment quality, and litigation convenience are defined, with default weight configurations of 0.3 for trial efficiency, 0.25 for case experience, 0.25 for judgment quality, and 0.2 for litigation convenience. Users can also customize the weights according to their actual needs. The matching score between each dimension and the current case is calculated as follows: Trial efficiency score: The average trial period of the court under the case cause is queried. The shorter the trial period, the higher the score. The calculation formula is 100 minus (actual period divided by the benchmark period and then multiplied by 100). The lower limit of the score is zero.

[0101] Case experience dimension score: Search for the number of cases handled by the court under the case type. The higher the number of cases handled, the higher the score. The calculation formula is (the logarithm of the number of cases handled by the court divided by the logarithm of the national maximum) and then multiplied by one hundred.

[0102] The score for the quality of judgment is based on the court's rate of upholding judgments in the second instance. The higher the rate of upholding judgments, the higher the score. The calculation formula is the rate of upholding judgments in the second instance multiplied by one hundred.

[0103] Litigation convenience score: Online case filing support adds 20 points, electronic service of process adds 20 points, online court hearing support adds 30 points, cross-jurisdictional case filing support adds 30 points, and the sum is the score for this dimension.

[0104] In addition, a dimension can be set to assess the favorability of each compliant candidate court relative to the user's current status, allowing the user to intuitively understand which compliant candidate court is more advantageous for subsequent rights protection matters. The favorability dimension score can be based on a comprehensive consideration of factors such as status-based success rate, claim support rate, and procedural bias, and the weight of each factor can be adjusted according to user needs.

[0105] In one embodiment, the risk of jurisdictional objection is assessed for each candidate court, and the objection risk factors include: Risk of territorial evidence: If the strength of territorial evidence corresponding to the court is low, there is a risk that the opposing party may deny the territorial relationship, thus increasing the risk score.

[0106] The risk of the agreement's validity: If the court enters the case for jurisdiction based on the agreement and the validity of the agreement is uncertain, there is a risk that the other party may claim that the agreement is invalid, thus increasing the risk score.

[0107] The risk of competition and substitution exists if multiple related regions correspond to different courts. In such cases, the opposing party may claim jurisdiction by other courts. The risk score is determined based on the strength of evidence in the region of substitution.

[0108] Uncertainty risk: If there are undetermined elements in the case fact map, the risk score will be increased according to the importance of the undetermined elements.

[0109] The formula for the objection risk score is: the risk score equals the sum of the scores of each risk factor multiplied by their corresponding weights. A higher risk score indicates a greater likelihood of a jurisdictional objection being raised.

[0110] In one embodiment, the matching score and the objection risk score are fused to calculate a comprehensive score. The fusion formula is: the comprehensive score equals the matching score multiplied by (one minus the risk adjustment coefficient multiplied by the risk score divided by one hundred). The risk adjustment coefficient defaults to 0.3 and can be configured by the user. Candidate courts are sorted in descending order according to the comprehensive score, and the sorting results are output. Furthermore, detailed scores for each court can be retained, including scores for each dimension, matching score, scores for each risk factor, risk score, and comprehensive score, for subsequent interpretation and generation.

[0111] In one embodiment, the court's recommendation result is output, and the result includes: Recommended courts: The court with the highest overall score. Output the court's name, level, administrative division, and contact information (if any).

[0112] Alternative court list: The courts ranked second to fifth in overall score (if any), output the court name, overall score, and the difference between the first choice and the second choice.

[0113] Explanation of the total number of candidate courts and the scope of compliant candidates.

[0114] In one embodiment, a complete explanation of the reasoning path is generated for the preferred recommended court report, and the explanation is organized in the following order: The first part is a summary of case elements, including the cause of action, basic information of the parties, amount in dispute, and key geographical areas; The second part is an explanation of the applicable jurisdiction rules: if jurisdiction by agreement applies, explain the content of the agreement terms and the results of the validity verification; if exclusive jurisdiction applies, explain the exclusive jurisdiction rules and triggering elements; if statutory territorial jurisdiction applies, explain the applicable territorial relationship and legal basis. Part Three is an explanation of the determination of jurisdiction by level, which explains the correspondence between the amount in dispute and the standards for jurisdiction by level; Part Four explains the process of generating candidate courts, illustrating the mapping and selection process from relational regions to candidate courts. Part Five explains the ranking criteria, detailing the scores for each dimension and the overall ranking results.

[0115] The output report should include legal citations, formatted as law name plus clause number.

[0116] In one embodiment, risk warnings and supplementary suggestions are output based on the risk assessment results. For example, if the risk of evidence regarding the territorial relationship is high, the warning states, "The evidence you claim regarding a certain territorial relationship (such as the place of performance of the contract) is weak, and the other party may raise jurisdictional objections. It is recommended to supplement relevant evidence materials." If the risk of the agreement's validity is high, the warning states, "The validity of the jurisdictional agreement is uncertain, possibly due to a specific reason. It is recommended to consult a professional lawyer to assess the validity of the agreement." If the risk of competition and substitution is high, the warning states, "This case involves multiple territorial jurisdictions. The other party may claim jurisdiction by a certain court, and the court you chose may face jurisdictional objections." If there are pending elements, the warning states, "Due to incomplete information on a certain element, the jurisdictional determination result may be affected. It is recommended to supplement specific information." Supplementary suggestions are provided according to different risk types, such as evidence supplementation suggestions, explaining the types of evidence that can be supplemented and their impact on the jurisdictional determination; agreement improvement suggestions, explaining where the agreement clauses can be improved (applicable to situations where the contract has not been signed or can be modified); and alternative solution suggestions, if the risk of the preferred court is high, explaining the alternative courts with lower risk and their advantages and disadvantages.

[0117] This method constructs a jurisdiction knowledge base by acquiring court information, court jurisdiction rules, pre-set constraint rules, and court profile information; it acquires case information for the current case, parses and structures the case information to generate a case fact graph; based on the case fact graph, it detects whether there is a valid jurisdiction agreement for the current case; if a valid jurisdiction agreement exists, it determines the corresponding court with jurisdiction based on the agreement; if the jurisdiction agreement has expired or no jurisdiction agreement is detected, it searches for jurisdiction rules and determines constraint rules in the jurisdiction knowledge base based on the case fact graph to obtain a set of constraint rules; based on the set of candidate courts for the current case in the case fact graph, it filters out candidate courts that do not conform to the rules from the candidate court set and the court with jurisdiction under the agreement, obtaining a set of compliant candidate courts; it matches the court profile information of each compliant candidate court in the compliant candidate court set with the case fact graph to obtain the matching degree of each compliant candidate court, and sorts the compliant candidate courts in the compliant candidate court set in descending order according to the matching degree to generate court recommendation results. This method reduces the workload of manual retrieval and repeated comparisons by extracting case elements and matching jurisdiction rules, and recommends relevant courts for the current case according to the matched jurisdiction rules, thereby improving the efficiency of determining the court with jurisdiction over the case.

[0118] This invention also provides a device for determining the court of jurisdiction, which is used to execute any embodiment of the aforementioned method for determining the court of jurisdiction. Specifically, please refer to... Figure 2 , Figure 2This is a schematic block diagram of a case jurisdiction determination device 100 provided in an embodiment of the present invention. The case jurisdiction determination device 100 can be configured in a server.

[0119] like Figure 2 As shown, the case jurisdiction court judgment device 100 includes an acquisition module 110, an analysis module 120, a detection module 130, a first determination module 140, and a matching module 150.

[0120] The acquisition module 110 is used to acquire court information, court jurisdiction rules, preset constraint rules and court profile information, and to construct a jurisdiction knowledge base based on the court information, court jurisdiction rules, preset constraint rules and court profile information; The parsing module 120 is used to obtain the case information of the current case, and to parse and structure the case information to generate a case fact map; The detection module 130 is used to detect whether there is a valid jurisdiction agreement for the current case based on the case fact map. The jurisdiction agreement is agreed upon by both parties and the content of the jurisdiction agreement includes at least the name of the court with jurisdiction or the scope of the jurisdiction. The first determining module 140 is used to determine the corresponding court of jurisdiction based on the jurisdiction agreement if a valid jurisdiction agreement exists. The matching module 150 is used to, if the jurisdiction agreement has expired or no jurisdiction agreement is detected, perform jurisdiction rule retrieval and constraint rule determination in a preset jurisdiction knowledge base based on the case fact graph to obtain a set of constraint rules; determine a set of candidate courts for the current case based on the relational regions in the case fact graph; filter out candidate courts that do not conform to the rules in the candidate court set and the courts under the agreement jurisdiction according to the set of constraint rules to obtain a set of compliant candidate courts, wherein the constraint set includes exclusive jurisdiction constraints, hierarchical jurisdiction constraints, special court constraints, and agreement jurisdiction constraints; perform matching calculations based on the court profile information of each compliant candidate court in the set of compliant candidate courts and the case fact graph to obtain the matching degree of each compliant candidate court; and sort the compliant candidate courts in the set of compliant candidate courts in descending order according to the matching degree to generate court recommendation results.

[0121] In one embodiment, the device 100 for determining the jurisdiction of a court includes: The identification module is used to identify whether there are special jurisdictional elements in the case information. The special jurisdictional elements include real estate property rights elements, port operation elements, inheritance elements, corporate litigation elements, intellectual property elements, maritime and commercial elements, negotiable instruments elements, financial elements, and Internet elements. The second determination module is used to determine the court with jurisdiction over the current case according to the regulations corresponding to the special jurisdictional elements if special jurisdictional elements exist.

[0122] In one embodiment, the matching module 150 includes: The calculation unit is used to calculate the matching score between the feature vector of the current case and the court profile vector of each compliant candidate court in multiple dimensions, and obtain the matching score in multiple dimensions. The weighted summation unit is used to sum the matching scores of multiple dimensions according to preset dimension weights to obtain the matching degree between each compliant candidate court and the current case.

[0123] This invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for determining the court of jurisdiction as described above.

[0124] In another embodiment of the invention, a computer-readable storage medium is provided. This computer-readable storage medium may be a non-volatile computer-readable storage medium. The computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to perform the method for determining jurisdiction of a court as described above.

[0125] Those skilled in the art will readily understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention.

[0126] In the embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Units with the same function may be grouped into one unit. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, devices, or units, or it may be an electrical, mechanical, or other form of connection.

[0127] 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 the embodiments of the present invention, depending on actual needs.

[0128] 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.

[0129] 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 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 various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), magnetic disks, or optical disks.

[0130] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for determining the court with jurisdiction over a case, characterized in that, include: Obtain the case information of the current case, and parse and structure the case information to generate a case fact map; Based on the aforementioned case fact map, it is determined whether a valid jurisdiction agreement exists for the current case. The jurisdiction agreement is agreed upon by both parties and includes at least the name of the court with jurisdiction or the geographical scope of the jurisdiction. If a valid jurisdiction agreement exists, the corresponding court with jurisdiction shall be determined in accordance with the jurisdiction agreement. If the jurisdiction agreement has expired or no jurisdiction agreement is detected, then the jurisdiction rules are retrieved and the constraint rules are determined in the preset jurisdiction knowledge base based on the case fact map to obtain a set of constraint rules. Based on the relational regions in the case fact graph, a set of candidate courts for the current case is determined. Based on the set of constraint rules, candidate courts that do not conform to the rules in the set of candidate courts and the courts with agreed jurisdiction are screened out to obtain a set of compliant candidate courts. Based on the court profile information and case fact map of each compliant candidate court in the set of compliant candidate courts, a matching calculation is performed to obtain the matching degree of each compliant candidate court. The compliant candidate courts in the set of compliant candidate courts are then sorted in descending order according to the matching degree to generate court recommendation results.

2. The method for determining the court of jurisdiction as described in claim 1, characterized in that, include: Obtain court information, court jurisdiction rules, pre-set constraint rules, and court profile information; The jurisdiction knowledge base is constructed based on the court information, court jurisdiction rules, pre-set constraint rules, and court profile information.

3. The method for determining the court of jurisdiction as described in claim 1, characterized in that, After parsing and structuring the case information to generate a case fact map, the process includes: Identify whether there are special jurisdictional elements in the case information. These special jurisdictional elements include real estate property rights elements, port operation elements, inheritance elements, corporate litigation elements, intellectual property elements, maritime and commercial elements, negotiable instruments elements, financial elements, and internet elements. If special jurisdictional elements exist, the court with jurisdiction over the current case shall be determined in accordance with the relevant laws and regulations governing such special jurisdictional elements.

4. The method for determining the court with jurisdiction over a case as described in claim 1, characterized in that, The process of parsing and structuring the case information to generate a case fact map includes: From the case information, extract the case type and cause of action, party information, subject matter information, and geographical location of the current case. The party information includes relevant information about the plaintiff and defendant. The subject matter information includes the type and value information of the subject matter. The geographical location includes the defendant's domicile, the place of contract signing, the place of contract performance, the place where the tort was committed, the place where the tort occurred, the location of the real estate, the company's domicile, and the location of the estate.

5. The method for determining the court of jurisdiction as described in claim 1, characterized in that, The matching calculation based on the court profile information of each compliant candidate court in the set of compliant candidate courts and the case fact map is used to obtain the matching degree of each compliant candidate court, including: Calculate the matching scores of the feature vector of the current case and the court profile vector of each compliant candidate court in multiple dimensions to obtain the multi-dimensional matching score; The matching scores of multiple dimensions are weighted and summed according to the preset dimension weights to obtain the matching degree between each compliant candidate court and the current case.

6. A device for determining the court with jurisdiction over a case, characterized in that, include: The acquisition module is used to acquire court information, court jurisdiction rules, preset constraint rules, and court profile information, and to construct a jurisdiction knowledge base based on the court information, court jurisdiction rules, preset constraint rules, and court profile information; The parsing module is used to obtain the case information of the current case, and to parse and structure the case information to generate a case fact map; The detection module is used to detect whether there is a valid jurisdiction agreement for the current case based on the case fact map. The jurisdiction agreement is agreed upon by both parties and the content of the jurisdiction agreement includes at least the name of the court with jurisdiction or the scope of the jurisdiction. The first determining module is used to determine the corresponding court of jurisdiction based on the jurisdiction agreement if a valid jurisdiction agreement exists. The matching module is used to perform jurisdiction rule retrieval and constraint rule determination in a preset jurisdiction knowledge base based on the case fact graph if the jurisdiction agreement has expired or no jurisdiction agreement is detected, so as to obtain a set of constraint rules. Based on the relational regions in the case fact graph, a set of candidate courts for the current case is determined. Based on the set of constraint rules, candidate courts that do not conform to the rules in the set of candidate courts and the courts with agreed jurisdiction are screened out to obtain a set of compliant candidate courts. Based on the court profile information and case fact map of each compliant candidate court in the set of compliant candidate courts, a matching calculation is performed to obtain the matching degree of each compliant candidate court. The compliant candidate courts in the set of compliant candidate courts are then sorted in descending order according to the matching degree to generate court recommendation results.

7. The device for determining the court of jurisdiction as described in claim 6, characterized in that, include: The identification module is used to identify whether there are special jurisdictional elements in the case information. The special jurisdictional elements include real estate property rights elements, port operation elements, inheritance elements, corporate litigation elements, intellectual property elements, maritime and commercial elements, negotiable instruments elements, financial elements, and Internet elements. The second determination module is used to determine the court with jurisdiction over the current case according to the regulations corresponding to the special jurisdictional elements if special jurisdictional elements exist.

8. The device for determining the court of jurisdiction as described in claim 6, characterized in that, The matching module includes: The calculation unit is used to calculate the matching score between the feature vector of the current case and the court profile vector of each compliant candidate court in multiple dimensions, and obtain the matching score in multiple dimensions. The weighted summation unit is used to sum the matching scores of multiple dimensions according to preset dimension weights to obtain the matching degree between each compliant candidate court and the current case.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements a method for determining the jurisdiction of a court as claimed in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to perform the method for determining the jurisdiction of a court as described in any one of claims 1 to 6.

Images