WO2021073409A1 - 数据处理方法、装置、设备及存储介质 - Google Patents

数据处理方法、装置、设备及存储介质 Download PDF

Info

Publication number
WO2021073409A1
WO2021073409A1 PCT/CN2020/118273 CN2020118273W WO2021073409A1 WO 2021073409 A1 WO2021073409 A1 WO 2021073409A1 CN 2020118273 W CN2020118273 W CN 2020118273W WO 2021073409 A1 WO2021073409 A1 WO 2021073409A1
Authority
WO
WIPO (PCT)
Prior art keywords
trial
case
dispute
focus
evidence
Prior art date
Application number
PCT/CN2020/118273
Other languages
English (en)
French (fr)
Inventor
刘嘉伟
于修铭
汪伟
陈晨
李可
Original Assignee
平安科技(深圳)有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 平安科技(深圳)有限公司 filed Critical 平安科技(深圳)有限公司
Publication of WO2021073409A1 publication Critical patent/WO2021073409A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/36Creation of semantic tools, e.g. ontology or thesauri
    • G06F16/374Thesaurus
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33Querying
    • G06F16/332Query formulation
    • G06F16/3329Natural language query formulation or dialogue systems
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/35Clustering; Classification
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10Services
    • G06Q50/18Legal services
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02PCLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30Computing systems specially adapted for manufacturing

Definitions

  • This application relates to the field of data processing technology, and in particular to a data processing method, device, device, and computer-readable storage medium.
  • the main purpose of this application is to provide a data processing method, device, equipment, and computer-readable storage medium, aiming to solve the technical problem of low efficiency in the existing manual trial business processing method.
  • the first aspect of this application provides a data processing method, which includes: obtaining the plaintiff and the Yankee's case files, and extracting key information from the case files through identification technology to identify the case files At least two dispute points of the dispute, wherein the case documents include at least a petition filed by the plaintiff and a defense file filed by the court; the type of the case is determined according to the dispute focus, and the laws and regulations corresponding to the type are inquired, Historical case information and the corresponding legal knowledge map framework; according to the laws and regulations and the historical case information, construct a logical relationship between at least two of the dispute focal points in the case file, and according to the logical relationship,
  • the dispute focus is filled into the legal knowledge map framework to form a trial map, wherein the trial map is a reasoning tree for the trial of a case generated based on the dispute focus; the dispute focus is performed according to the trial map Review to obtain the theoretical trial result.
  • the review includes judging whether the reasoning logic of the petition document and the defense document is reasonable, and whether the defense reasons in the defense
  • the second aspect of the present application provides a data processing device, including a memory, a processor, and computer-readable instructions stored on the memory and running on the processor, and the processor executes the computer-readable instructions.
  • the following steps are implemented during the instruction: obtain the case files of the plaintiff and the court, and extract key information from the case files through identification technology to identify at least two dispute points in the case files, wherein the case files are at least Including the petition files filed by the plaintiff and the defense files filed by the court; determine the type of the case according to the focus of the dispute, and query the laws and regulations corresponding to the type, historical case information, and the corresponding legal knowledge map framework; Laws and regulations and the historical case information, construct a logical relationship between at least two of the dispute focal points in the case file, and fill the dispute focal point into the legal knowledge graph framework according to the logical relationship, A trial atlas is formed, wherein the trial atlas is a reasoning tree for the trial of cases generated based on the dispute focus; the dispute focus is reviewed according to the trial atlas to
  • the third aspect of the present application provides a computer-readable storage medium having computer instructions stored in the computer-readable storage medium, and when the computer instructions are run on a computer, the computer executes the following steps: obtaining information about the plaintiff and the lawyer Case documents, by extracting key information from the case documents through identification technology to identify at least two controversies in the case documents, where the case documents include at least the petition documents filed by the plaintiff and the court files Defense document; determine the type of the case according to the focus of the dispute, and query the laws and regulations, historical case information corresponding to the type, and the corresponding legal knowledge graph framework; construct the legal knowledge graph framework according to the laws and regulations and the historical case information State the logical relationship between at least two of the dispute focal points in the case file, and fill the dispute focal point into the legal knowledge graph framework according to the logical relationship to form a trial graph, where the trial graph is The reasoning tree for the trial of the case generated based on the dispute focus; the dispute focus is reviewed according to the trial graph to obtain a theoretical trial result, and the review includes the reasoning logic for
  • the fourth aspect of the present application provides a data processing device, including: a collection module for obtaining case files of the plaintiff and the latter, and extracting key information from the case files through identification technology to identify the case files At least two dispute points of the dispute, where the case file includes at least the petition filed by the plaintiff and the defense file filed by the court; the query module is used to determine the type of the case according to the dispute focus, and query the type Corresponding laws and regulations, historical case information, and corresponding legal knowledge map framework; a map generation module for constructing a logical relationship between at least two of the dispute focal points in the case file according to the laws and regulations and historical case information , And according to the logical relationship, fill the dispute focus into the legal knowledge graph framework to form a trial graph, wherein the trial graph is a reasoning tree for the trial of a case generated based on the dispute focus; a trial module, It is used to review the focus of the dispute according to the trial graph to obtain a theoretical trial result, wherein the review includes judging whether the reasoning logic of the petition document and the
  • the data processing method provided in this application is mainly to assist court judgments by constructing a legal knowledge graph of court cases.
  • the knowledge graph is composed of various nodes and node relationships, and has the inherent advantage of representing various complex relationships. It is a relationship map combining knowledge methods and human cognition, and it is also the basis for realizing artificial intelligence. It is an important method for machines to understand natural language and construct knowledge networks. The use of this kind of atlas to achieve auxiliary judgments has greatly helped practitioners to quickly retrieve relevant legal content online, thereby improving the quality and efficiency of the court's trial work.
  • FIG. 1 is a schematic structural diagram of a business system involved in a solution according to an embodiment of the application
  • Figure 3 is a schematic diagram of the construction of the trial atlas provided by this application.
  • Fig. 5 is a schematic diagram of functional modules of a data processing device provided by this application.
  • This application provides a business system. As shown in Figure 1, the system is mainly based on the opinions and evidence presented by the plaintiff and the lawyer. The judge only needs to check the attributes of the evidence, and then proceed according to the knowledge graph and judicial engine of this application. At the same time, the output of the judgment plan will explain the judgment result in a visual form to enhance the judge's credibility of this method.
  • the method proposed in this application can also be combined with other applications, such as evidence guidance and category pushes, which can give judges more effective information.
  • the process of judges using this method is also a process of supplementing the knowledge map, which can optimize the underlying legal knowledge map itself. Using the method of this application can greatly shorten the time for judges to decide cases and increase efficiency.
  • the business system includes: a processor 101, such as a CPU, a communication bus 102, a user interface 103, a network interface 104, and a memory 105.
  • the communication bus 102 is used to implement connection and communication between these components.
  • the user interface 103 may include a display screen (Display) and an input unit such as a keyboard (Keyboard), and the network interface 104 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface).
  • the memory 105 may be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as a magnetic disk memory.
  • the memory 105 may also be a storage system independent of the aforementioned processor 101.
  • FIG. 1 does not constitute a limitation on the data processing device, and may include more or less components than those shown in the figure, or a combination of certain components, or different components. Component arrangement.
  • the memory 105 as a computer-readable storage medium may include an operating system, a network communication module, a user interface module, and data processing for automatic trial of appeal files and defense files of court cases. program.
  • the operating system is a program that manages and controls the software resource calls in the data processing device and the memory, and supports the operation of the data processing program and other software and/or programs.
  • the network interface 104 is mainly used to access the network; the user interface 103 is mainly used to detect whether data processing operations are required in the system or whether the data in the monitoring system is updated or abnormal.
  • the processor 101 can be used to call the data processing program stored in the memory 105 and execute the operations of the following embodiments of the data processing method.
  • the implementation of FIG. 1 can also be a networked system composed of a mobile terminal and a server, where the mobile terminal is used as a task data table generating device, and the server is used to synchronize the task data table on the mobile terminal.
  • the processor of the server can distinguish the incremental task data by reading the program code stored in the buffer or the storage unit that can realize the data processing method, and realize the synchronous operation.
  • the data processing method provided by the present application can also be used to implement synchronized database data processing between multiple servers S and terminals TE, or between several terminals TE and several servers S, where the synchronized database It can be broadly understood as a storage system that can store any data. If synchronization is to be performed between the terminal TE or the server S, the terminal TE is used as a generating device for the task data table to be synchronized, the server S is used as a synchronization server, and the server S is usually a network server or a PC. TE is usually a mobile phone, PC (personal computer), laptop or PDA device.
  • the data processing method specifically includes the following steps:
  • Step S210 Obtain the case files of the plaintiff and the court, and extract key information of the case files through identification technology to identify at least two dispute points in the case files;
  • the case file includes at least the petition filed by the plaintiff and the defense file filed by the court;
  • the plaintiff will file a complaint request through the legal affairs agency or through the corresponding website on the judicial Internet, and the judicial agency can obtain the plaintiff’s case in real time through the real-time monitoring of the legal affairs agency or the communication interface on the judicial Internet. file.
  • the case documents include the reasons for the petition and the evidence of the petition.
  • the rationality review of the grounds of the petition refers to whether the petition meets the requirements of laws and regulations and the verification of the authenticity of the evidence.
  • the verification of the authenticity generally involves reviewing whether the source of the evidence is legal.
  • the judiciary after reviewing the plaintiff’s petition, the judiciary will issue a court hearing request to the court based on the petition.
  • the court provides defense documents based on the court hearing notice, specifically the reasons for the defense and the defense evidence.
  • the judicial agency can obtain the court's defense document in real time by monitoring the legal affairs agency or the communication interface on the judicial Internet.
  • the identification method adopted is the same as the identification method used to identify the case file of the plaintiff.
  • this step after identifying the focal points of dispute in the case files of the plaintiff and the lawyer, it also includes comparing the focal points identified by the two, and selecting the same or corresponding dispute focal points to determine the final dispute of the case focus.
  • Step S220 Determine the type of the case according to the focus of the dispute, and query the laws and regulations, historical case information and the corresponding legal knowledge graph framework corresponding to the type;
  • the type of case may be adjusted according to the location of the dispute focus. Therefore, the legal knowledge map structure here is a preliminary identification map of the dispute focus, and the map that is really suitable for the case is still It needs to be re-determined according to the focus of the dispute, and the actual type of the case is determined based on the combination of the focus of the dispute, and some historical case information of the same type and corresponding laws and regulations are obtained from the judicial structure system according to the determined actual type.
  • Step S230 Construct a logical relationship between at least two of the dispute focal points in the case file according to the laws and regulations and historical case information, and fill the dispute focal point into the legal knowledge according to the logical relationship
  • a trial map is formed
  • the trial atlas is a reasoning tree for the trial of the case generated based on the dispute focus; for the construction of the trial atlas, it can be specifically obtained through model training and learning, and it can be implemented by a random forest algorithm.
  • a storage node for fact evidence is set under each small tree branch.
  • the trial map is obtained through training, the dispute focus, elements and evidence of the trial case can be stored correspondingly according to the trial map. Go to the corresponding node of the trial atlas, then directly call, and execute step S240.
  • Step S240 Trial the focus of the dispute according to the trial atlas, and obtain a theoretical trial result
  • the trial specifically includes judging whether the reasoning logic of the appeal document and the defense document is reasonable, and whether the defense reasons in the defense document are correct;
  • each node stores audit elements, audit rules, and factual evidence, and the audit Rules can be understood as the logical rules of the relationship between some laws and regulations and some nodes; and then the audit operation of the upper-level nodes, so that the step-by-step accounting is performed in turn.
  • Step S250 Output a trial opinion based on the theoretical trial result and use the trial opinion as a reference document for the court judgment.
  • the trial opinion output here may specifically be the result of a judgment, such as the success of the litigation or the failure of the litigation.
  • it may also be a judgment of a court trial, which contains the result of the litigation and the litigation.
  • the auxiliary trial was realized based on the form of a knowledge map.
  • the underlying knowledge map information was completely derived from the judgment document, legal rules and regulations and the judgment manual, and there is no doubt about the authenticity of the map.
  • the application scenarios of this method are completely derived from real court scenarios, that is, starting from the opinions and evidence presented by the plaintiff and the court, the judge only needs to check the attributes of the evidence, and then the judgment plan can be made based on the knowledge graph and judicial engine of this application.
  • the judgment result will be explained in a visual form to enhance the judge’s credibility with this method.
  • the method proposed in this application can also be combined with other applications, such as evidence guidance and category pushes, which can give judges more effective information.
  • the process of judges using this method is also a process of supplementing the knowledge map, which can optimize the underlying legal knowledge map itself. Using the method of this application can greatly shorten the time for judges to decide cases and increase efficiency.
  • link evidence includes contract documents, grounds of appeal, factual evidence of appeals, grounds of defense, evidence of defenses, and judicial review of the laws and regulations involved in the case Explanation;
  • the link evidence is analyzed to screen out the dispute focus among the focus keywords.
  • the plaintiff’s petition document and the court’s defense document are compared with each other, combined with semantic recognition to distinguish the corresponding dispute focus, and at the same time classify the type of focus, so as to extract If the focus of the dispute of the litigation case is revealed, in the subsequent trial, the trial can be conducted directly on these focus issues, which greatly accelerates the efficiency of the trial.
  • the relationship between each dispute focus is determined, and the graph is constructed according to the relationship.
  • the knowledge graph structure in the historical case information can be extracted;
  • a litigation often has multiple dispute points, and each dispute point has multiple small elements, which are divided into large elements and small elements.
  • the reasoning tree is constructed based on such elements, and the resulting reasoning tree There are often multiple layers; at the same time, there will be a corresponding relationship between each element, and large elements may need to be reviewed and estimated on the premise of small elements.
  • the reasoning tree is composed of four layers, from top to bottom, they are the focus of dispute, the big facts, the small facts, and the evidence.
  • Each reasoning tree contains a dispute focus, a fact element, multiple small fact elements and multiple evidences.
  • the establishment condition is based on the logical calculation of the child nodes, and the method of automatic production judgment standards will be used here.
  • the calculation logic of the child node is identified here.
  • the step of using the knowledge graph framework as a basis and modifying the framework in accordance with the trial methods of the laws and regulations to generate the reasoning tree includes:
  • the reasoning block diagram is established for the key points of the case according to the parent-child relationship after the classification is completed;
  • the corresponding link evidence on the node is attached to the corresponding node on the reasoning block diagram, and the logical calculation relationship between it and the parent node or the subordinate node is determined to generate the reasoning tree.
  • the reasoning method perform alignment reasoning analysis on the grounds of appeal in the linked evidence, the facts of the appeal, the grounds for defense, and the defense evidence, and determine whether the link evidence of the node meets the establishment conditions, and the establishment conditions include the establishment conditions. Please succeed or the defense is established.
  • the trial results also include the process of accounting and reasoning based on the trial map and the explanation of the reasons for the results.
  • the final trial proposal is output based on the results of the accounting.
  • the format of the trial proposal can be output according to the preset document construction format, or it can be generated according to the template of the judgment letter of historical case information.
  • the process of outputting the trial proposal also includes generating the reason for each node in the trial graph (that is, the reasoning tree).
  • the reason for generating the reason is based on the evidence provided by the plaintiff, the reason, and the court.
  • the defenses and evidences put forward are explained in a one-to-one correspondence, and recorded in the trial proposal, so as to facilitate the judges' reference and understanding, and at the same time improve the persuasiveness and credibility of the trial opinion.
  • the reason for appeal in the linked evidence is the parent node in the reasoning tree
  • the evidence based on the fact of the appeal is determined according to the reasoning method.
  • the reason for defense is determined according to the reasoning method.
  • the above method can be applied to two scenarios.
  • One is for court trial, that is, it is just in the stage of submission and acceptance.
  • the system on the judicial institution can conduct a preliminary trial, and the trial process is as described above.
  • the trial has already been conducted, but the actual trial result has not yet been made. At this time, before the output of the trial opinion, it also includes:
  • the trial opinion is adjusted according to the trial record to generate a trial proposal.
  • both parties input case files to the remote On the trial platform, the platform extracts the focus of disputes in case documents by implementing the above methods, and displays them to both parties for review, and then establishes a corresponding legal knowledge map according to the requirements of the trial. After both parties confirm, in the process of the trial, according to A quick review of the atlas.
  • the following is a detailed description of the above-mentioned auxiliary trial methods in combination with specific cases, which are litigation cases based on the business type of lending.
  • the knowledge graph cited in the examples of this application includes the plaintiff, the lawyer, the petition, the court’s argument, the focus of the dispute, the big facts, the small facts, the law, and the evidence. The relationship between these factors is shown in Figure 3. Show:
  • Step S310 Based on the claims made by the plaintiff, the defenses made by the court, and the evidence presented by both parties, combined with the existing legal knowledge graph, the technologies of entity linking, relationship alignment and semantic similarity are used to identify the focus of the dispute. At present, there are seven major disputes about whether the loan relationship is established, whether the loan form is reasonable, whether the contract is effective, whether the contract is valid, whether the contract is normally performed, whether the guarantee relationship is established, and whether the loan is a joint debt of the husband and wife.
  • Step S320 For each dispute focus, an independent reasoning tree is generated according to laws and regulations and historical cases, using the method of automatically producing judgment standards.
  • the reasoning tree is composed of four layers, from top to bottom, it is the focus of dispute, the big facts, the small facts, and the evidence.
  • Each reasoning tree contains a dispute focus, a fact element, multiple small fact elements and multiple evidences
  • step S330 at the node of the dispute focus of the inference tree, mark its child nodes as big fact elements, and the establishment condition is the logical calculation and calculation.
  • Step S340 On the node of the big fact element of the inference tree, mark its child node as a small fact element, and the parent node is the focus of dispute.
  • the establishment condition is based on the logical calculation of the child node, and the automatic production judgment standard will be used here. Method, the calculation logic of the child node is identified here.
  • Step S350 On the node of the small fact element of the inference tree, mark its child node as evidence, and the parent node as the big fact element, and the establishment condition is based on the logical calculation of the child node.
  • Step S360 On the evidence node of the inference tree, mark its parent node as a small fact element, and at the same time, split each kind of evidence supporting (or not supporting) the small fact element according to the legal knowledge graph.
  • Step S370 Each time the user uses this method, the logical result of each layer is set to false. When using this method, the user only needs to check the part of the evidence at the lowest level. It automatically calculates the suggestion whether the focus of the dispute is established in this scenario.
  • the logical calculation will be performed on the parent node of its father_node according to the value of its support_law, and the calculation result will be updated to the result_bool of the parent node.
  • the small fact elements are updated, a new round of logical calculations will be performed based on the operation of their parent node, and the calculation results will be updated to the big fact elements. Finally, based on the results of the big fact elements, whether the focus of the dispute is supported is output.
  • FIG. 5 is a schematic diagram of functional modules of the data processing device provided by the embodiment of the present application.
  • the device includes:
  • the collection module 51 is used to obtain the case files of the plaintiff and the court, and extract key information of the case files through identification technology to identify at least two dispute points in the case files, wherein the case files are at least Including the petition files filed by the plaintiff and the defense files filed by the court;
  • the query module 52 is used to determine the type of the case according to the focus of the dispute, and to query the laws and regulations corresponding to the type, historical case information, and the corresponding legal knowledge graph framework;
  • the graph generation module 53 is configured to construct a logical relationship between at least two of the dispute focal points in the case file according to the laws, regulations and historical case information, and fill in the dispute focal points according to the logical relationship
  • a trial map is formed in the legal knowledge map framework, wherein the trial map is a reasoning tree for the trial of a case generated based on the dispute focus;
  • the trial module 54 is used to review the focus of the dispute according to the trial atlas to obtain a theoretical trial result.
  • the review includes judging whether the reasoning logic of the appeal document and the defense document is reasonable, and the defense document Whether the defense reason in is correct;
  • the output module 55 is configured to output a trial opinion based on the theoretical trial result and use the trial opinion as a reference document for the court judgment.
  • the content of the embodiment is the same as the description of the data processing method in the above embodiment of the present application. Therefore, the content of the embodiment of the data processing device in this embodiment will not be described in detail.
  • the present application also provides a data processing device, including: a memory and at least one processor, the memory stores instructions, the memory and the at least one processor are interconnected through a wire; the at least one processor calls the The instructions in the memory, so that the data processing device executes the steps in the above data processing method.
  • the present application also provides a computer-readable storage medium.
  • the computer-readable storage medium may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium.
  • the computer-readable storage medium stores computer instructions, and when the computer instructions are executed on the computer, the computer executes the following steps:
  • case files of the plaintiff and the court and extract key information from the case files through identification technology to identify at least two dispute points in the case files, where the case files include at least the petition filed by the plaintiff Documents and defense documents submitted by the court;
  • a trial atlas is formed in the framework, wherein the trial atlas is a reasoning tree for the trial of the case generated based on the dispute focus;
  • the dispute focus is reviewed according to the trial graph to obtain the theoretical trial result, wherein the review includes judging whether the reasoning logic of the appeal document and the defense document is reasonable, and whether the defense reason in the defense document is reasonable correct;

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Computational Linguistics (AREA)
  • Tourism & Hospitality (AREA)
  • Economics (AREA)
  • Technology Law (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Human Computer Interaction (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种数据处理方法,涉及大数据技术领域,通过构建庭审案件的法律知识图谱的方式来实现辅助法庭判案,其中,知识图谱由各种节点及节点关系构成,天生具有表示各种复杂关系的优势,其是知识方法与人类认知相结合的一种关系图谱,也是实现人工智能的基础,它是机器理解自然语言和构建知识网络的重要方法;还包括一种数据处理装置、设备及计算机可读存储介质,使用该种图谱来实现辅助判案,大大帮助了从业人员快速地在线检索相关的法务内容,从而提高法院审判工作质量和效率。

Description

数据处理方法、装置、设备及存储介质
本申请要求于2019年10月18日提交中国专利局、申请号为201910992322.1、发明名称为“数据处理方法、装置、设备及存储介质”的中国专利申请的优先权,其全部内容通过引用结合在申请中。
技术领域
本申请涉及数据处理技术领域,尤其涉及一种数据处理方法、装置、设备及计算机可读存储介质。
背景技术
随着人工智能的不断发展,为了降低人力资源的应用成本,目前正在不断地研发和推广人工智能在工作中的应用,尤其是在法律领域上,法律业务是当前可知的消耗人工时长最长的一种工作,这也是由于法律诉讼业务的工作内容相对比较复杂和逻辑性较强的特性,决定了该业务会涉及到比较多的考量因素。
在目前所能实现的法律智能化审判中,大部分都是应用在一些事实、证据、参考案例的检索审判上,以及一些案情简单的案件审判上,而对于案情复杂一点的法律事件,若继续采用智能审判,基于关键字查询的和融合自然语言处理技术来实现时,由于当前智能审判方式的限制,关键字的检索判断效率和准确性都不能满足需求,尤其是在准确性上,必须要考虑到各方面的关联关系,因此,智能审判的准确性就面领着极大的挑战,同时,自然语言处理技术无法对自身产生的结果进行解释,也让其在严肃的法律领域无法让人信服。
发明人意识到,针对上述现有技术中审判过于简单,无法保证审判结果的准确度的问题,目前尚未提出有效的解决方案。
发明内容
本申请的主要目的在于提供一种数据处理方法、装置、设备及计算机可读存储介质,旨在解决现有的人工审判的业务处理方式,其效率较低的技术问题。
为实现上述目的,本申请第一方面提供了一种数据处理方法,包括:获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点,其中,所述案件文件至少包括原告提出的诉请文件和被告提出的抗辩文件;根据所述争议焦点确定案件的类型,并查询出与所述类型对应的法律法规、历史案例信息以及对应的法律知识图谱框架;根据所述法律法规和所述历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中,形成审判图谱,其中,所述审判图谱为基于所述争议焦点生成的案件的审判的推理树;根据所述审判图谱对所述争议焦点进行审核,得到理论审判结果,所述审核包括判断所述诉请文件和所述抗辩文件的推理逻辑是否合理,以及所述抗辩文件中的抗辩理由是否正确;根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文件。
本申请第二方面提供了一种数据处理设备,包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机可读指令,所述处理器执行所述计算机可读指令时实现如下步骤:获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点,其中,所述案件文件至少包括原告提出的诉请文件和被告提出的抗辩文件;根据所述争议焦点确定案件的类型,并查询出与所述类型对应的法律法规、历史案例信息以及对应的法律知识图谱框架;根据所述法律法规和所述历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中,形成审判图谱,其中,所述审判图谱为基于所述争议焦点生成的案件的审判的推理树;根据所述审判图谱对所述 争议焦点进行审核,得到理论审判结果,所述审核包括判断所述诉请文件和所述抗辩文件的推理逻辑是否合理,以及所述抗辩文件中的抗辩理由是否正确;根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文件。
本申请第三方面提供了一种计算机可读存储介质,所述计算机可读存储介质中存储有计算机指令,当所述计算机指令在计算机上运行时,使得计算机执行如下步骤:获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点,其中,所述案件文件至少包括原告提出的诉请文件和被告提出的抗辩文件;根据所述争议焦点确定案件的类型,并查询出与所述类型对应的法律法规、历史案例信息以及对应的法律知识图谱框架;根据所述法律法规和所述历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中,形成审判图谱,其中,所述审判图谱为基于所述争议焦点生成的案件的审判的推理树;根据所述审判图谱对所述争议焦点进行审核,得到理论审判结果,所述审核包括判断所述诉请文件和所述抗辩文件的推理逻辑是否合理,以及所述抗辩文件中的抗辩理由是否正确;根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文件。
本申请第四方面提供了一种数据处理装置,包括:采集模块,用于获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点,其中,所述案件文件至少包括原告提出的诉请文件和被告提出的抗辩文件;查询模块,用于根据所述争议焦点确定案件的类型,并查询出与所述类型对应的法律法规、历史案例信息以及对应的法律知识图谱框架;图谱生成模块,用于根据所述法律法规和历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中形成审判图谱,其中,所述审判图谱为基于所述争议焦点生成的案件的审判的推理树;审判模块,用于根据所述审判图谱对所述争议焦点进行审核,得到理论审判结果,其中,所述审核包括判断所述诉请文件和所述抗辩文件的推理逻辑是否合理,以及所述抗辩文件中的抗辩理由是否正确;输出模块,用于根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文件。
本申请提供的数据处理方法,主要是通过构建庭审案件的法律知识图谱的方式来实现辅助法庭判案,其中,知识图谱由各种节点及节点关系构成,天生具有表示各种复杂关系的优势,其是知识方法与人类认知相结合的一种关系图谱,也是实现人工智能的基础,它是机器理解自然语言和构建知识网络的重要方法。使用该种图谱来实现辅助判案,大大帮助了从业人员快速地在线检索相关的法务内容,从而提高法院审判工作质量和效率。
附图说明
图1为本申请实施例方案涉及的业务系统的结构示意图;
图2为本申请提供的数据处理方法第一实施例的流程示意图;
图3为本申请提供的审判图谱的构建示意图;
图4为本申请提供的数据处理方法第二实施例的流程示意图;
图5为本申请提供数据处理装置的功能模块示意图。
本申请目的的实现、功能特点及优点将结合实施例,参照附图做进一步说明。
具体实施方式
应当理解,此处所描述的具体实施例仅用以解释本申请,并不用于限定本申请。
本申请提供一种业务系统,如图1所示,该系统主要是从原被告提出观点及证据出发,法官只需要对证据的属性进行勾选,即可依据本申请的知识图谱和司法引擎进行判决方案 的输出,同时会以可视化的形式对判决结果进行解释说明,增强法官对本方法的信服度。同时,本申请提出的方法还能结合其他应用,如证据指引和类案推送等,能给予法官更多的有效信息。同时法官使用此方法的过程也是补充知识图谱的过程,能优化底层的法律知识图谱本身,利用本申请的方法,能极大的缩短法官判案的时间,增加效率。
如图1所示,该业务系统包括:处理器101,例如CPU,通信总线102、用户接口103,网络接口104,存储器105。其中,通信总线102用于实现这些组件之间的连接通信。用户接口103可以包括显示屏(Display)、输入单元比如键盘(Keyboard),网络接口104可选的可以包括标准的有线接口、无线接口(如WI-FI接口)。存储器105可以是高速RAM存储器,也可以是稳定的存储器(non-volatile memory),例如磁盘存储器。存储器105可选的还可以是独立于前述处理器101的存储系统。
本领域技术人员可以理解,图1中示出的业务系统的硬件结构并不构成对数据处理装置的限定,可以包括比图示更多或更少的部件,或者组合某些部件,或者不同的部件布置。
如图1所示,作为一种计算机可读存储介质的存储器105中可以包括操作系统、网络通信模块、用户接口模块以及用于实现对庭审案件的诉请文件和抗辩文件进行自动审判的数据处理程序。其中,操作系统是管理和控制数据处理装置、存储器中的软件资源调用的程序,支持数据处理程序以及其它软件和/或程序的运行。
在图1所示的业务系统的硬件结构中,网络接口104主要用于接入网络;用户接口103主要用于检测系统中是否需要进行数据处理操作或者是监控系统中的数据是否存在更新、异常等信息,而处理器101可以用于调用存储器105中存储的数据处理程序,并执行以下数据处理方法的各实施例的操作。
在本申请实施例中,对于图1的实现还可以是由移动终端和服务器组成的连网系统,其中,移动终端作为任务数据表的产生设备,服务器用于同步移动终端上的任务数据表的设备,该服务器的处理器通过读取存储在缓存器或者存储单元中的可以实现数据处理方法的程序代码来分辨出增量任务数据,并实现同步的操作。
进一步的,本申请提供的数据处理方法还可以用于实现多个服务器S与终端TE之间,或者是几个终端TE之间与几个服务器S之间的同步数据库数据处理,其中,同步数据库可广义理解为可以存储任何数据的存储系统。如果要在终端TE或服务器S之间执行同步,则其中终端TE作为待同步的任务数据表的产生设备,服务器S用作同步服务器,服务器S通常是网络服务器或PC。TE通常是移动电话、PC(个人计算机)、膝上型计算机或PDA设备。
基于上述业务系统的硬件结构,本申请提出了一种数据处理方法,如图2所示,本实施例中,所述数据处理方法具体包括以下步骤:
步骤S210,获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点;
在该步骤中,所述案件文件至少包括原告提出的诉请文件和被告提出的抗辩文件;
在实际应用中,原告会通过法律事务机构或者是通过在司法互联网的相应网站上提出控诉请求,而司法机构通过从实时监控法律事务机构或者司法互联网上的通信接口即可实时获取到原告的案件文件。
在该步骤中,在获取到原告提供的案件文件后,会对原告提供的案件文件进行初步审核,具体的是该案件文件包括诉请理由和诉请的证据,而这里的初步审核主要是对诉请理由的合理性审核,即是是否满足法律法规的诉请,以及对证据的真实性的审核,可选的,对于真实性的审核一般是对证据的来源是否合法进行审核即可。
在本实施例中,在对原告的诉请审核通过后,司法机构会根据诉请请求向被告发出要求出庭的庭审通知,被告基于庭审通知提供抗辩的文件,具体是抗辩的理由和抗辩证据, 以及对应的抗辩焦点,同理,司法机构通过监控法律事务机构或者司法互联网上的通信接口即可实时获取到被告的抗辩文件。
在实际应用中,除了通过识别技术来提取关键信息得到争议焦点之外,还可以是通过预设的法律知识图谱来提取,即是根据法律知识图谱中规定的识别算法来对案件文件进行分析处理,而对于法律知识图谱的选择具体是根据原告提交的诉请请求的案件文件中的案件类型信息来确定,而在司法机构的系统中预先设置有多种不同的类型的法律知识图谱,在确定了知识图谱后,根据知识图谱中的争议点识别算法或技术来确定该案件的争议焦点。
在本实施例中,对于被告提交的案件文件中的争议焦点识别来说,采用的识别方式与识别原告的案件文件的识别方式一样。
而在该步骤中,在识别原告和被告的案件文件中的争议焦点之后,还包括将两者识别到的焦点进行比较,选择出相同的或者相对应的争议焦点,从而确定该案件的最终争议焦点。
步骤S220,根据所述争议焦点确定案件的类型,并查询出与所述类型对应的法律法规、历史案例信息以及对应的法律知识图谱框架;
在实际应用中,对于案件的类型可能会根据争议焦点的定位不同会存在一定的类别调整,因此,这里的法律知识图谱构架是初步的争议焦点的识别图谱,而真正适合该案件的图谱,还是需要根据争议焦点来重新确定,而根据争议焦点的相互组合来确定该案件的实际类型,并根据确定的实际类型从司法结构的系统中获取一些相同类型的历史案例信息以及对应的法律法规。
步骤S230,根据所述法律法规和历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中,形成审判图谱;
在该步骤中,所述审判图谱为基于所述争议焦点生成的案件的审判的推理树;对于审判图谱的构建,具体可以通过模型训练学习的方式来获得,具体的可以是随机森林算法可以实现,首先将历史案件信息进行争议焦点的提取,然后根据争议焦点进行小点的划分,例如:一个争议焦点做一个分叉树枝,在该分叉树枝上还设置有多个小树杈,每个树杈对应一个争议小点,同时,在每个小树杈下设置事实证据的存储节点,在训练得到审判图谱后,可以根据该审判图谱将该审判案件的争议焦点、要素和证据分别对应存储到审判图谱的对应节点上,然后直接调用,并执行步骤S240即可。
步骤S240,根据所述审判图谱对所述争议焦点进行审判,得到理论审判结果;
在该步骤中,对于所述审判具体包括判断所述诉请文件和所述抗辩文件的推理逻辑是否合理,以及所述抗辩文件中的抗辩理由是否正确;
在本实施例中,在对争议焦点进行逻辑审判和核算过程中,具体是首先针对一个争议焦点上的各个节点进行计算,每个节点上都存储有审核要素、审核规则和事实证据,而审核规则可以理解为是一些法律法规和一些节点之间的关系逻辑规则;然后是对上一级节点的审核运算,这样依次进行逐级核算。
步骤S250,根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文件。
在本实施例中,这里输出的审判意见书具体可以是一种判断的结果,例如诉讼成功或者是诉讼失败,此外,还可以是一种庭审的判决书,该判决书中包含有诉讼的结果以及诉讼后的一些赔偿、手续、费用等等的通知书。
在本案中,基于知识图谱的形式来实现辅助审判,其底层的知识图谱信息完全来自裁判文书、法理法规及判案手册,在图谱的真实性上无容置疑。同时,本方法的使用场景完 全来自真实的法庭场景,即从原被告提出观点及证据出发,法官只需要对证据的属性进行勾选,即可依据本申请的知识图谱和司法引擎进行判决方案的输出,同时会以可视化的形式对判决结果进行解释说明,增强法官对本方法的信服度。同时,本申请提出的方法还能结合其他应用,如证据指引和类案推送等,能给予法官更多的有效信息。同时法官使用此方法的过程也是补充知识图谱的过程,能优化底层的法律知识图谱本身,利用本申请的方法,能极大的缩短法官判案的时间,增加效率。
在本实施例中,对于在根据识别技术提取案件文件中的争议焦点时,通过根据原告和被告双方的文件来对比提取得到,具体的根据关键字提取规则分别提取所述诉请文件和抗辩文件中焦点关键字;
利用实体链接识别技术提取与所述焦点关键字关联的链接证据,其中,所述链接证据包括合同文件、诉请理由、诉请事实证据、抗辩理由、抗辩证据和案件所涉及的法律法规的司法解释;
根据诉请与抗辩的对齐关系和语义相似度识别技术,对所述链接证据进行分析,以筛选出所述焦点关键字中的争议焦点。
在实际应用中,根据原告提出的诉请、被告提出的辩称以及双方提出的证据,结合已有的法律知识图谱,利用实体链接、关系对齐和语义相似度的技术,识别出其的争议焦点。
例如现在的借贷诉讼案件,其大部分是贷款未能及时还款、合同存在矛盾,借贷人不愿意还利息部分等等的问题,而对于这类的案件的争议焦点往往都是借贷关系是否成立、借贷形式是否合理、合同是否生效、合同是否有效、合同是否正常履行、担保关系是否成立、借款是否为夫妻共同债务等七大争议焦点。这时,在提取争议焦点的过程中,根据原告的诉请文件和被告的抗辩文件来进行相互比对,结合语义的识别来区分出相对应的争议焦点,同时归类焦点的类型,从而提取出该诉讼案件的争议焦点,在后续的庭审中,直接针对这些焦点问题进行庭审即可,这样大的加快了庭审的效率。
进一步的,根据提取到的争议焦点,确定每个争议焦点之间的关系,根据其关系来构建图谱,具体的可以的通过提取所述历史案例信息中的知识图谱架构;
将所述争议焦点以及所述争议焦点对应的链接证据进行分类;
以所述知识图谱架构作为基础,根据所述法律法规的审判方式对所述知识图谱构架进行修改,并以分类后的链接证据作为训练样本,对修改后的知识图谱构架进行训练,以生成所述推理树,其中,所述推理树包含争议焦点节点层和至少一层链接证据层。
在实际应用中,一个诉讼往往都会存在多个争议焦点,并且每个争议焦点其都会存在多个小要素,分为大要素和小要素之分,根据这样要素来构建推理树,得到的推理树也往往是存在多层;同时,每个要素之间会存在对应关系,大要素可能要在小要素的前提下进行审核推算。
例如,推理树由四层构成,从上往下依次为争议焦点、大事实要素、小事实要素、证据。其中每一个推理树包含一个争议焦点、一个事实要素,多个小事实要素和多个证据。
在推理树的争议焦点这一节点,标注其子节点为大事实要素,成立条件是逻辑计算的and计算。
在推理树的大事实要素这一节点上,标注其子节点为小事实要素,父节点为争议焦点,成立条件是依据子节点的逻辑计算,并且此处会利用自动生产判案标准的方法,将子节点的计算逻辑在此处标识出来。
即是说,在构建推理树节点的同时,还需要对节点的之间的关系进行梳理,从而构建一个具有逻辑性、推理流程完整的推理树,其具体实现步骤为:
根据所述案件的案情提取每个争议焦点中的案件关键点,并对所述案件关键点按照父 子关系进行分类;
根据所述案件关键点的分类情况对所述链接证据进行二次分类,并与对应的案件关键点建立引用关系;
所述以所述知识图谱架构作为基础,根据所述法律法规的审判方式对所述构架进行修改,以生成所述推理树包括:
根据所述法律法规的审判方式,对所述案件关键点按照分类完成后的父子关系建立所述推理框图;
将节点上对应的链接证据附加到所述推理框图上对应的节点中,并确定其与父节点或者下级节点之间的逻辑计算关系,以生成所述推理树。
这时,在输出审判结果时,根据所述法律法规的司法解释确定每个节点上的链接证据的核算的推理方式;
根据所述推理方式对所述链接证据中的诉请理由、诉请事实依证据、抗辩理由和抗辩证据进行对齐推理分析,确定所述节点的连接证据是否满足成立条件,所述成立条件包括诉请成功或者抗辩成立。
在本案中,所述审判结果除了包括核算的结果之外,还包括根据审判图谱进行核算推理的过程和得到结果的理由说明,而最终得输出的审判建议书是根据核算的结果进行输出的,其审判建议书的格式可以是根据预设的文档构建格式进行输出,也可以是按照历史案件信息的判决书的模板进行生成。
在本实施例中,在输出审判建议书的过程中,还包括生成审判图谱(即是推理树)中每个节点的推算理由,其推算理由的生成逻辑是根据原告提供的证据、理由和被告提出的抗辩理由、证据进行一一对应的推理说明,并且记录在审判建议书中,以便于审判员的参考理解,同时还提高了该审判意见书的说服力和可信力。
在实际应用中,若所述节点为所述推理树中的父节点时,在所述根据所述推理方式对所述链接证据中的诉请理由、诉请事实依证据、抗辩理由和抗辩证据进行对齐推理分析,确定所述节点的连接证据是否满足成立条件之前,还包括:
检测与所述父节点连接的子节点中的推理结果是否为诉请成功;
若是,则根据所述子节点的推理结果和父节点的推理结果进行与关系的计算;
若否,则结束所述父节点的推理或者所述父节点所在的争议焦点的推理,并跳转至其他争议焦点继续推理审判。
上述的方法可以应用于两种场景中,一种是为进行庭审的,即是刚处于提交受理阶段,这时,司法机构上的系统可以进行初步的审判,其审判过程如上所述。
而对于另一种场景,则是已经进行过庭审的,但是庭审还没有做出实际的审判结果,这时在输出的审判意见书之前,还包括:
检测所述案件是否为已完成庭审案件;
若是,则获取庭审的庭审记录;
根据所述庭审记录对所述审判意见书进行调整,以生成庭审建议书。
对于本申请提高的数据处理方法中,主要是应用于辅助庭审的快速审查和判决,但是也可以应用到网络远程庭审的情况,而对于远程的庭审,其应用则是双方将案件文件输入到远程审理平台上,该平台通过执行上述的方法对案件文件的争议焦点进行提取,并展示给双方查看,然后根据庭审的要求建立对应的法律知识图谱,在双方确认后,在庭审的过程中,根据图谱进行快速审查。下面结合具体的案例对上述的辅助审判方法进行详细的说明,基于借贷的业务类型的诉讼案件。
首先,在正常的法院开庭之前,会有原告提出诉请,被告提出辩称,同时双方会提出 自己观点所需的证据,同时法庭会让双方交换证据,并进行证据真实性的检测。
本申请实施例中所引用的知识图谱,包含有原告、被告、原告诉请、被告辩称、争议焦点、大事实要素、小事实要素、法条、证据等,这些要素的关系如图3所示:
在梳理出上述的关系图后,生成对应的知识图谱,在该知识图谱的辅助下进行审判,如图4所示,其具体流程为:
步骤S310、根据原告提出的诉请、被告提出的辩称以及双方提出的证据,结合已有的法律知识图谱,利用实体链接、关系对齐和语义相似度的技术,识别出其的争议焦点。目前有借贷关系是否成立、借贷形式是否合理、合同是否生效、合同是否有效、合同是否正常履行、担保关系是否成立、借款是否为夫妻共同债务等七大争议焦点。
步骤S320、针对每一个争议焦点,根据法规及历史案件,利用自动生产判案标准的方法,生成一个独立的推理树。推理树由四层构成,从上往下依次为争议焦点、大事实要素、小事实要素、证据。其中每一个推理树包含一个争议焦点、一个事实要素,多个小事实要素和多个证据
步骤S330、在推理树的争议焦点这一节点,标注其子节点为大事实要素,成立条件是逻辑计算的and计算。
步骤S340、在推理树的大事实要素这一节点上,标注其子节点为小事实要素,父节点为争议焦点,成立条件是依据子节点的逻辑计算,并且此处会利用自动生产判案标准的方法,将子节点的计算逻辑在此处标识出来。
步骤S350、在推理树的小事实要素这一节点上,标注其子节点为证据,父节点为大事实要素,成立条件是依据子节点的逻辑计算。
步骤S360、在推理树的证据这一节点上,标注其父节点为小事实要素,同时会根据法律知识图谱,将每一种证据支持(或者不支持)小事实要素进行拆分。
步骤S370、在使用者每次使用本方法之前,都将每一层的逻辑结果置为false,使用者在使用本方法的时候,只需要对最底层的证据这一部分进行勾选操作,即可自动计算出该场景下争议焦点是否成立的建议。
此处以一个简化版的借贷关系是否成立这一争议焦点为例,进行以上步骤的详细说明。
借贷关系是否成立:id=1,child_node=2,operation=and,result_bool=False
借贷主体及关系:id=2,father_node=1,child_node=3,4,5,operation=3,4,5,result_bool=False,support_law=True
是否签订借条/收据/借款合同/欠条:id=3,father_node=2,child_node=6,7,operation=child_node,result_bool=False,support_law=True
是否存在转账凭证:id=4,father_node=2,child_node=8,9,operation=child_node,result_bool=False,support_law=True
出借人与借款人的关系、款项交付情形、出借人经济能力、交易方式、交易地点、交易习惯、财产变动情况是否合理:id=5,father_node=2,child_node=10,11,operation=child_node,result_bool=False,support_law=True
无借款人签名的【借款合同】:id=6,father_node=3,support_law=False
约定借款人、本金、利息及用途的【借款合同】:id=7,father_node=3,support_law=True
转账清单累计金额与诉请金额不符的【转账凭证】:id=8,father_node=4,support_law=False
载明接收人账号的汇款单,且金额符合诉请金额的【转账凭证】:id=9,father_node=4,support_law=True
系当事人之间的真实意思表示的【借款合同】:id=10,father_node=5,support=True
未能证明借款用于经营活动的【借款合同】:id=11,father_node=6,support_law=False
当使用者对证据进行勾选的时候,会根据其support_law的值对其father_node的父节点进行逻辑计算,计算结果会更新到父节点的result_bool上。小事实要素都更新后会根据他们的父节点的operation,进行新一轮的逻辑计算,计算结果更新到大事实要素上,最后根据大事实要素的结果对争议焦点是否支持进行输出。
为了解决上述的问题,本申请实施例还提供了一种数据处理装置,参照图5,图5为本申请实施例提供的数据处理装置的功能模块的示意图。在本实施例中,该装置包括:
采集模块51,用于获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点,其中,所述案件文件至少包括原告提出的诉请文件和被告提出的抗辩文件;
查询模块52,用于根据所述争议焦点确定案件的类型,并查询出与所述类型对应的法律法规、历史案例信息以及对应的法律知识图谱框架;
图谱生成模块53,用于根据所述法律法规和历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中形成审判图谱,其中,所述审判图谱为基于所述争议焦点生成的案件的审判的推理树;
审判模块54,用于根据所述审判图谱对所述争议焦点进行审核,得到理论审判结果,所述审核包括判断所述诉请文件和所述抗辩文件的推理逻辑是否合理,以及所述抗辩文件中的抗辩理由是否正确;
输出模块55,用于根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文件。
基于与上述本申请实施例的数据处理方法相同的实施例说明内容,因此本实施例对数据处理装置的实施例内容不做过多赘述。
本申请还提供一种数据处理设备,包括:存储器和至少一个处理器,所述存储器中存储有指令,所述存储器和所述至少一个处理器通过线路互连;所述至少一个处理器调用所述存储器中的所述指令,以使得所述数据处理设备执行上述数据处理方法中的步骤。
本申请还提供一种计算机可读存储介质,该计算机可读存储介质可以为非易失性计算机可读存储介质,也可以为易失性计算机可读存储介质。计算机可读存储介质存储有计算机指令,当所述计算机指令在计算机上运行时,使得计算机执行如下步骤:
获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点,其中,所述案件文件至少包括原告提出的诉请文件和被告提出的抗辩文件;
根据所述争议焦点确定案件的类型,并查询出与所述类型对应的法律法规、历史案例信息以及对应的法律知识图谱框架;
根据所述法律法规和所述历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中形成审判图谱,其中,所述审判图谱为基于所述争议焦点生成的案件的审判的推理树;
根据所述审判图谱对所述争议焦点进行审核,得到理论审判结果,其中,所述审核包括判断所述诉请文件和所述抗辩文件的推理逻辑是否合理以及所述抗辩文件中的抗辩理由是否正确;
根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文 件。
通过以上的实施方式的描述,本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现,当然也可以通过硬件,但很多情况下前者是更佳的实施方式。基于这样的理解,本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来,该计算机软件产品存储在一个存储介质(如ROM/RAM)中,包括若干指令用以使得一台终端(可以是手机,计算机,服务器或者网络设备等)执行本申请各个实施例所述的方法。
上面结合附图对本申请的实施例进行了描述,但是本申请并不局限于上述的具体实施方式,上述的具体实施方式仅仅是示意性的,而不是限制性的,本领域的普通技术人员在本申请的启示下,在不脱离本申请宗旨和权利要求所保护的范围情况下,还可做出很多形式,凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换,或直接或间接运用在其他相关的技术领域,这些均属于本申请的保护之内。

Claims (20)

  1. 一种数据处理方法,其中,包括:
    获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点,其中,所述案件文件至少包括原告提出的诉请文件和被告提出的抗辩文件;
    根据所述争议焦点确定案件的类型,并查询出与所述类型对应的法律法规、历史案例信息以及对应的法律知识图谱框架;
    根据所述法律法规和所述历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中形成审判图谱,其中,所述审判图谱为基于所述争议焦点生成的案件的审判的推理树;
    根据所述审判图谱对所述争议焦点进行审核,得到理论审判结果,其中,所述审核包括判断所述诉请文件和所述抗辩文件的推理逻辑是否合理以及所述抗辩文件中的抗辩理由是否正确;
    根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文件。
  2. 根据权利要求1所述的数据处理方法,其中,所述获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点的步骤包括:
    根据关键字提取规则分别提取所述诉请文件和抗辩文件中焦点关键字;
    利用实体链接识别技术提取与所述焦点关键字关联的链接证据,其中,所述链接证据包括合同文件、诉请理由、诉请事实证据、抗辩理由、抗辩证据和案件所涉及的法律法规的司法解释;
    根据诉请与抗辩的对齐关系和语义相似度识别技术,对所述链接证据进行分析,以筛选出所述焦点关键字中的争议焦点。
  3. 根据权利要求2所述的数据处理方法,其中,所述根据所述法律法规和历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点提前入所述法律知识图谱框架中形成审判图谱的步骤包括:
    提取所述历史案例信息中的知识图谱架构;
    将所述争议焦点以及所述争议焦点对应的链接证据进行分类;
    以所述知识图谱架构作为基础,根据所述法律法规的审判方式对所述知识图谱构架进行修改,并以分类后的链接证据作为训练样本,对修改后的知识图谱构架进行训练,以生成所述推理树,其中,所述推理树包含争议焦点节点层和至少一层链接证据层。
  4. 根据权利要求3所述的数据处理方法,其中,所述根据所述法律法规和历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中形成审判图谱的步骤,还包括:
    根据所述案件的案情提取每个争议焦点中的案件关键点,并对所述案件关键点按照父子关系进行分类;
    根据所述案件关键点的分类情况对所述链接证据进行二次分类,并与对应的案件关键点建立引用关系;
    所述以所述知识图谱架构作为基础,根据所述法律法规的审判方式对所述构架进行修改,以生成所述推理树包括:
    根据所述法律法规的审判方式,对所述案件关键点按照分类完成后的父子关系建立所述推理框图;
    将节点上对应的链接证据附加到所述推理框图上对应的节点中,并确定其与父节点或者下级节点之间的逻辑计算关系,以生成所述推理树。
  5. 根据权利要求4所述的数据处理方法,其中,所述根据所述审判图谱对所述争议焦点进行审核,得到理论审判结果包括:
    根据所述法律法规的司法解释确定每个节点上的链接证据的核算的推理方式;
    根据所述推理方式对所述链接证据中的诉请理由、诉请事实依证据、抗辩理由和抗辩证据进行对齐推理分析,确定所述节点的连接证据是否满足成立条件,所述成立条件包括诉请成功或者抗辩成立。
  6. 根据权利要求5所述的数据处理方法,其中,若所述节点为所述推理树中的父节点时,在所述根据所述推理方式对所述链接证据中的诉请理由、诉请事实依证据、抗辩理由和抗辩证据进行对齐推理分析,确定所述节点的连接证据是否满足成立条件之前,还包括:
    检测与所述父节点连接的子节点中的推理结果是否为诉请成功;
    若是,则根据所述子节点的推理结果和父节点的推理结果进行与关系的计算;
    若否,则结束所述父节点的推理或者所述父节点所在的争议焦点的推理,并跳转至其他争议焦点继续推理审判。
  7. 根据权利要求1-6中任一项所述的数据处理方法,其中,在所述根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文件的步骤之后,还包括:
    检测所述案件是否为已完成庭审案件;
    根据所述庭审记录对所述审判意见书进行调整,以生成庭审建议书。
    根据所述庭审记录对所述审判意见书进行调整,以生成庭审建议书。
  8. 一种数据处理设备,包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机可读指令,所述处理器执行所述计算机可读指令时实现如下步骤:
    获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点,其中,所述案件文件至少包括原告提出的诉请文件和被告提出的抗辩文件;
    根据所述争议焦点确定案件的类型,并查询出与所述类型对应的法律法规、历史案例信息以及对应的法律知识图谱框架;
    111122221所述历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中形成审判图谱,其中,所述审判图谱为基于所述争议焦点生成的案件的审判的推理树;
    根据所述审判图谱对所述争议焦点进行审核,得到理论审判结果,其中,所述审核包括判断所述诉请文件和所述抗辩文件的推理逻辑是否合理以及所述抗辩文件中的抗辩理由是否正确;
    根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文件。
  9. 根据权利要求8所述的数据处理设备,所述处理器执行所述计算机程序时还实现以下步骤:
    根据关键字提取规则分别提取所述诉请文件和抗辩文件中焦点关键字;
    利用实体链接识别技术提取与所述焦点关键字关联的链接证据,其中,所述链接证据包括合同文件、诉请理由、诉请事实证据、抗辩理由、抗辩证据和案件所涉及的法律法规的司法解释;
    根据诉请与抗辩的对齐关系和语义相似度识别技术,对所述链接证据进行分析,以筛选出所述焦点关键字中的争议焦点。
  10. 根据权利要求9所述的数据处理设备,所述处理器执行所述计算机程序时还实现以下步骤:
    提取所述历史案例信息中的知识图谱架构;
    将所述争议焦点以及所述争议焦点对应的链接证据进行分类;
    以所述知识图谱架构作为基础,根据所述法律法规的审判方式对所述知识图谱构架进行修改,并以分类后的链接证据作为训练样本,对修改后的知识图谱构架进行训练,以生成所述推理树,其中,所述推理树包含争议焦点节点层和至少一层链接证据层。
  11. 根据权利要求8所述的数据处理设备,所述处理器执行所述计算机程序时还实现以下步骤:
    根据所述案件的案情提取每个争议焦点中的案件关键点,并对所述案件关键点按照父子关系进行分类;
    根据所述案件关键点的分类情况对所述链接证据进行二次分类,并与对应的案件关键点建立引用关系;
    所述以所述知识图谱架构作为基础,根据所述法律法规的审判方式对所述构架进行修改,以生成所述推理树包括:
    根据所述法律法规的审判方式,对所述案件关键点按照分类完成后的父子关系建立所述推理框图;
    将节点上对应的链接证据附加到所述推理框图上对应的节点中,并确定其与父节点或者下级节点之间的逻辑计算关系,以生成所述推理树。
  12. 根据权利要求11所述的数据处理设备,所述处理器执行所述计算机程序时还实现以下步骤:
    根据所述法律法规的司法解释确定每个节点上的链接证据的核算的推理方式;
    根据所述推理方式对所述链接证据中的诉请理由、诉请事实依证据、抗辩理由和抗辩证据进行对齐推理分析,确定所述节点的连接证据是否满足成立条件,所述成立条件包括诉请成功或者抗辩成立。
  13. 根据权利要求8所述的数据处理设备,所述处理器执行所述计算机程序时还实现以下步骤:
    检测与所述父节点连接的子节点中的推理结果是否为诉请成功;
    若是,则根据所述子节点的推理结果和父节点的推理结果进行与关系的计算;
    若否,则结束所述父节点的推理或者所述父节点所在的争议焦点的推理,并跳转至其他争议焦点继续推理审判。
  14. 根据权利要求8-13任一项所述的数据处理设备,所述处理器执行所述计算机程序时还实现以下步骤:
    检测所述案件是否为已完成庭审案件;
    若是,则获取庭审的庭审记录;
    根据所述庭审记录对所述审判意见书进行调整,以生成庭审建议书。
  15. 一种计算机可读存储介质,所述计算机可读存储介质中存储计算机指令,当所述计算机指令在计算机上运行时,使得计算机执行如下步骤:
    获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点,其中,所述案件文件至少包括原告提出的诉请文件和被告提出的抗辩文件;
    根据所述争议焦点确定案件的类型,并查询出与所述类型对应的法律法规、历史案例信息以及对应的法律知识图谱框架;
    根据所述法律法规和所述历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中形成审判图谱,其中,所述审判图谱为基于所述争议焦点生成的案件的审判的推理树;
    根据所述审判图谱对所述争议焦点进行审核,得到理论审判结果,其中,所述审核包括判断所述诉请文件和所述抗辩文件的推理逻辑是否合理以及所述抗辩文件中的抗辩理由是否正确;
    根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文件。
  16. 根据权利要求15所述的计算机可读存储介质,当所述计算机指令在计算机上运行时,使得计算机还执行以下步骤:
    根据关键字提取规则分别提取所述诉请文件和抗辩文件中焦点关键字;
    利用实体链接识别技术提取与所述焦点关键字关联的链接证据,其中,所述链接证据包括合同文件、诉请理由、诉请事实证据、抗辩理由、抗辩证据和案件所涉及的法律法规的司法解释;
    根据诉请与抗辩的对齐关系和语义相似度识别技术,对所述链接证据进行分析,以筛选出所述焦点关键字中的争议焦点。
  17. 根据权利要求16所述的计算机可读存储介质,当所述计算机指令在计算机上运行时,使得计算机还执行以下步骤:
    提取所述历史案例信息中的知识图谱架构;
    将所述争议焦点以及所述争议焦点对应的链接证据进行分类;
    以所述知识图谱架构作为基础,根据所述法律法规的审判方式对所述知识图谱构架进行修改,并以分类后的链接证据作为训练样本,对修改后的知识图谱构架进行训练,以生成所述推理树,其中,所述推理树包含争议焦点节点层和至少一层链接证据层。
  18. 根据权利要求15所述的计算机可读存储介质,当所述计算机指令在计算机上运行时,使得计算机还执行以下步骤:
    根据所述案件的案情提取每个争议焦点中的案件关键点,并对所述案件关键点按照父子关系进行分类;
    根据所述案件关键点的分类情况对所述链接证据进行二次分类,并与对应的案件关键点建立引用关系;
    所述以所述知识图谱架构作为基础,根据所述法律法规的审判方式对所述构架进行修改,以生成所述推理树包括:
    根据所述法律法规的审判方式,对所述案件关键点按照分类完成后的父子关系建立所述推理框图;
    将节点上对应的链接证据附加到所述推理框图上对应的节点中,并确定其与父节点或者下级节点之间的逻辑计算关系,以生成所述推理树。
  19. 根据权利要求18所述的计算机可读存储介质,当所述计算机指令在计算机上运行时,使得计算机还执行以下步骤:
    根据所述法律法规的司法解释确定每个节点上的链接证据的核算的推理方式;
    根据所述推理方式对所述链接证据中的诉请理由、诉请事实依证据、抗辩理由和抗辩证据进行对齐推理分析,确定所述节点的连接证据是否满足成立条件,所述成立条件包括诉请成功或者抗辩成立。
  20. 一种数据处理装置,其中,所述数据处理包括:
    采集模块,用于获取原告和被告的案件文件,通过识别技术对所述案件文件进行关键信息的提取,以识别出所述案件文件中的至少两个争议焦点,其中,所述案件文件至少包括原告提出的诉请文件和被告提出的抗辩文件;
    查询模块,用于根据所述争议焦点确定案件的类型,并查询出与所述类型对应的法律法规、历史案例信息以及对应的法律知识图谱框架;
    图谱生成模块,用于根据所述法律法规和历史案例信息,构建所述案件文件中至少两个所述争议焦点之间的逻辑关系,并根据所述逻辑关系,将所述争议焦点填入所述法律知识图谱框架中形成审判图谱,其中,所述审判图谱为基于所述争议焦点生成的案件的审判的推理树;
    审判模块,用于根据所述审判图谱对所述争议焦点进行审核,得到理论审判结果,其中,所述审核包括判断所述诉请文件和所述抗辩文件的推理逻辑是否合理,以及所述抗辩文件中的抗辩理由是否正确;
    输出模块,用于根据所述理论审判结果输出审判意见书并将所述审判意见书作为庭审判决的参考文件。
PCT/CN2020/118273 2019-10-18 2020-09-28 数据处理方法、装置、设备及存储介质 WO2021073409A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910992322.1A CN110929039B (zh) 2019-10-18 2019-10-18 数据处理方法、装置、设备及存储介质
CN201910992322.1 2019-10-18

Publications (1)

Publication Number Publication Date
WO2021073409A1 true WO2021073409A1 (zh) 2021-04-22

Family

ID=69849021

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/118273 WO2021073409A1 (zh) 2019-10-18 2020-09-28 数据处理方法、装置、设备及存储介质

Country Status (2)

Country Link
CN (1) CN110929039B (zh)
WO (1) WO2021073409A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114092119A (zh) * 2021-11-29 2022-02-25 北京金堤科技有限公司 供应关系获取方法、装置、存储介质及电子设备
CN114894498A (zh) * 2022-05-31 2022-08-12 重庆长安汽车股份有限公司 车辆云端检测系统、方法、电子设备及可读存储介质
CN115934702A (zh) * 2023-03-14 2023-04-07 青岛安工数联信息科技有限公司 流程工业中数据处理方法、装置、存储介质及处理器
CN116523036A (zh) * 2023-04-13 2023-08-01 北京鹅厂科技有限公司 多人利用大模型居间决策、建议的方法与装置

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110929039B (zh) * 2019-10-18 2023-09-29 平安科技(深圳)有限公司 数据处理方法、装置、设备及存储介质
CN111753517A (zh) * 2020-06-30 2020-10-09 北京来也网络科技有限公司 基于rpa及ai的文档对比方法、装置、设备及介质
CN111859969B (zh) * 2020-07-20 2024-05-03 航天科工智慧产业发展有限公司 数据分析方法及装置、电子设备、存储介质
CN112329891B (zh) * 2020-11-27 2022-05-31 浙江大学 双向注意力和判案逻辑结合的辅助判案方法、装置、介质
CN113032528B (zh) * 2021-04-09 2022-12-23 平安国际智慧城市科技股份有限公司 案件分析方法、装置、设备及存储介质
CN113222251A (zh) * 2021-05-13 2021-08-06 太极计算机股份有限公司 一种基于案件争议焦点的辅助裁判结果预测方法及系统
CN113377263B (zh) * 2021-05-25 2022-11-22 四川大学 一种法庭虚拟示证方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107633465A (zh) * 2017-08-21 2018-01-26 厦门能见易判信息科技有限公司 智能辅助判案方法
CN108304386A (zh) * 2018-03-05 2018-07-20 上海思贤信息技术股份有限公司 一种基于逻辑规则推断法律文书判决结果的方法及装置
US20190311367A1 (en) * 2015-06-20 2019-10-10 Quantiply Corporation System and method for using a data genome to identify suspicious financial transactions
CN110929039A (zh) * 2019-10-18 2020-03-27 平安科技(深圳)有限公司 数据处理方法、装置、设备及存储介质

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108241621B (zh) * 2016-12-23 2019-12-10 北京国双科技有限公司 法律知识的检索方法及装置
CN108009299A (zh) * 2017-12-28 2018-05-08 北京市律典通科技有限公司 法律审判业务处理方法和装置
CN110232447B (zh) * 2019-04-28 2021-04-06 杭州实在智能科技有限公司 法律案件深度推理方法
CN110175605A (zh) * 2019-05-27 2019-08-27 北京市律典通科技有限公司 基于要素式的电子审判数据处理方法及装置
CN110288495A (zh) * 2019-06-25 2019-09-27 北京市律典通科技有限公司 案件诉讼时效智能审查方法及装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190311367A1 (en) * 2015-06-20 2019-10-10 Quantiply Corporation System and method for using a data genome to identify suspicious financial transactions
CN107633465A (zh) * 2017-08-21 2018-01-26 厦门能见易判信息科技有限公司 智能辅助判案方法
CN108304386A (zh) * 2018-03-05 2018-07-20 上海思贤信息技术股份有限公司 一种基于逻辑规则推断法律文书判决结果的方法及装置
CN110929039A (zh) * 2019-10-18 2020-03-27 平安科技(深圳)有限公司 数据处理方法、装置、设备及存储介质

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114092119A (zh) * 2021-11-29 2022-02-25 北京金堤科技有限公司 供应关系获取方法、装置、存储介质及电子设备
CN114894498A (zh) * 2022-05-31 2022-08-12 重庆长安汽车股份有限公司 车辆云端检测系统、方法、电子设备及可读存储介质
CN114894498B (zh) * 2022-05-31 2023-07-25 重庆长安汽车股份有限公司 车辆云端检测系统、方法、电子设备及可读存储介质
CN115934702A (zh) * 2023-03-14 2023-04-07 青岛安工数联信息科技有限公司 流程工业中数据处理方法、装置、存储介质及处理器
CN115934702B (zh) * 2023-03-14 2023-05-23 青岛安工数联信息科技有限公司 流程工业中数据处理方法、装置、存储介质及处理器
CN116523036A (zh) * 2023-04-13 2023-08-01 北京鹅厂科技有限公司 多人利用大模型居间决策、建议的方法与装置

Also Published As

Publication number Publication date
CN110929039A (zh) 2020-03-27
CN110929039B (zh) 2023-09-29

Similar Documents

Publication Publication Date Title
WO2021073409A1 (zh) 数据处理方法、装置、设备及存储介质
CN109255499B (zh) 投诉、投诉案件处理方法、装置及设备
CN112182246B (zh) 通过大数据分析建立企业画像的方法、系统、介质及应用
CN111967761B (zh) 一种基于知识图谱的监控预警方法、装置及电子设备
CN110704572A (zh) 疑似非法集资风险的预警方法、装置、设备和存储介质
CN110866836B (zh) 计算机执行的医疗保险立案审核方法和装置
CN111859969B (zh) 数据分析方法及装置、电子设备、存储介质
CN116415017B (zh) 基于人工智能的广告敏感内容审核方法及系统
CN113656805A (zh) 一种面向多源漏洞信息的事件图谱自动构建方法及系统
O’Leary Big Data and knowledge management with applications in accounting and auditing: The case of Watson
CN113553444A (zh) 一种基于超边的审计知识图谱表示模型及关联推理方法
CN110544035A (zh) 一种内控检测方法、系统和计算机可读存储介质
CN112328868A (zh) 一种基于信息数据的信用评估与授信申请系统及方法
CN107679977A (zh) 一种基于语义分析的税务管理平台及实现方法
CN113722433A (zh) 一种信息推送方法、装置、电子设备及计算机可读介质
CN112396437A (zh) 一种基于知识图谱的贸易合同验证方法及装置
CN114282498B (zh) 一种应用于电力交易的数据知识处理系统
CN115292352A (zh) 问题查询方法、装置、设备、介质和程序产品
CN111966835B (zh) 基于知识图谱的场景所需功能服务的解析装置及方法
CN111782803A (zh) 一种工单的处理方法、装置、电子设备及存储介质
Randles et al. A vocabulary for describing mapping quality assessment, refinement and validation
CN113407734B (zh) 基于实时大数据的知识图谱系统的构建方法
Anunne et al. Framing analysis of Belt and Road Initiative coverage in major Nigerian, Malaysian, and Vietnamese newspapers
CN110322252B (zh) 风险主体识别方法以及装置
CN114549014A (zh) 基于自动和人工审核结合的供应链金融风控办法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20877344

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20877344

Country of ref document: EP

Kind code of ref document: A1