Construction project case element type intelligent auxiliary judgment method

Abstract

The invention discloses a construction project case element type intelligent auxiliary judgment method, and the method comprises the steps: carrying out the feature analysis of historical cases, and building a construction project case trial element and content extraction model through BERT + BiLSTM + CRF; the method comprises the following steps: firstly, based on a case appeal and trial element index system of an action cause carding construction contract dispute, realizing trial element extraction of a litigation request, and establishing a case appeal and element index relation model; secondly, in combination with specific contents of case appeals and elements, extracting trial element contents, appeal processing, referee statements, applicable law articles and referee results of the cases, and establishing a construction project case element type intelligent auxiliary trial knowledge base; key elements in construction project contract dispute case trial can be clarified, so that trial personnel can efficiently perform case trial; according to the invention, science and technology can be endowed to the trial work, and the judge is assisted to improve the trial efficiency of construction cases.

Description

Technical Field

[0001] The present invention relates to an element-based intelligent assisted trial method for construction cases, belonging to the field of tool technology. Background Art

[0002] Construction engineering contract disputes are characterized by large numbers, large amounts in dispute, complex legal relationships, high difficulty in adjudication, long trial periods, and multiple parties involved, which seriously restrict the development of the industry. Existing technologies have three core deficiencies: 1. Case review level The legal relationship is complex: it involves multiple legal issues such as qualification affiliation, illegal subcontracting, and illegal subcontracting, which often overlap with administrative supervision and criminal offenses, making it difficult to define rights and responsibilities.

[0003] The trial period is too long: Engineering evidence is highly professional and large in volume, requiring reliance on cost / quality appraisals. However, the appraisal procedures are complex and regulations are imperfect, making it difficult to conclude cases within the statutory period. Inconsistent adjudication standards: The legal system is complex, courts at all levels have differing understandings of judicial interpretations, and the adjudication results for similar cases vary significantly.

[0004] 2. Industry Roots Irregular operations are common: false bidding, malicious price-cutting, lending of qualifications and other problems are prominent, and construction units often violate legal procedures. 3. Dispute Causes The main reasons are delays in construction schedule, manifested in construction defects of the contractor and insufficient funds of the contracting party; quality defects, manifested in substandard materials and construction technology defects; and contract changes, which are caused by three major issues, and the unclear division of rights and responsibilities further aggravates the disputes. The existing trial method is highly dependent on the personal abilities of judges and lacks professional and standardized tools, resulting in inefficiency. It is in urgent need of empowerment with intelligent trial methods.

[0005] That is, there is a need for an element-based intelligent trial method for construction cases, which can empower the trial work with science and technology and assist judges in improving the trial efficiency of construction cases. Summary of the Invention

[0006] In view of this, the purpose of the present invention is to provide an element-based intelligent assisted trial method for construction cases, so as to empower the trial work with science and technology, assist judges in improving the trial efficiency of construction cases, and overcome the shortcomings of the existing technology.

[0007] The purpose of the present invention is achieved through the following technical solutions: The present invention discloses a factor-based intelligent assisted trial method for construction cases, which is characterized in that the method analyzes the characteristics of historical cases, establishes a construction case trial factor and content extraction model based on BERT+BiLSTM+CRF, extracts the trial factors of the litigation request according to the case cause, and forms a case claim and trial factor indicator system and relationship model for construction contract disputes; secondly, based on the specific content of the case claim and the factors, extracts the case trial factor content and its claim processing, judgment reasoning, applicable laws, and judgment results, and establishes a factor-based intelligent assisted trial knowledge base for construction cases, which helps to clarify the key factors in the trial of construction contract disputes and support judges in handling cases efficiently.

[0008] As mentioned above, the elements of the trial of construction contract disputes are sorted out, including the sorting out of litigation requests and case element indicators, sorting out whether each element belongs to the subject issue, jurisdiction issue, contract conclusion issue, contract performance issue, project quantity, project payment and interest issue and breach of contract liability, and making good classification of the litigation issues.

[0009] The above-mentioned case claims and trial factor indicators are constructed based on the litigation materials of the case. The BERT language model is used to extract text features to obtain the word-granularity vector matrix, and BiLSTM is used to extract contextual information. At the same time, the CRF model is combined to extract the global optimal sequence, and finally the factor indicators of the case trial are obtained.

[0010] The above-mentioned tests were conducted using actual desensitized defense statements and judgments, and an algorithm model was used to quickly extract and form a set of factor indicators. The newly added factor indicators were manually verified and labeled by industry experts to verify the validity of the factor indicators. Finally, a relationship model between the case claims and trial factor indicators of construction contract disputes was obtained, and the accuracy of the relationship model was optimized through labeling.

[0011] As mentioned above, the BERT language model is used to extract text features to obtain the word-granularity vector matrix, BiLSTM is used to extract contextual information, and the CRF model is combined to extract the global optimal sequence to establish a factor indicator content extraction model; at the same time, a subject relationship extraction model based on SpERT is constructed to identify the subjects of the construction contracts signed by the cooperative development parties and the contractors in the cooperative development and construction projects, determine the responsibilities of the cooperative development parties, identify the subjects in the multiple contracting contracts generated by subcontracting in the joint development and construction projects, distinguish between joint contracting and subcontracting, and identify the subjects in the cases of internal contracting and affiliation to distinguish between internal contracting and affiliation, so as to solve the problem of identifying the contract subjects in complex construction project contract disputes.

[0012] The above-mentioned content of litigation requests and trial factor indicators is automatically identified and extracted through the BERT+BiLSTM+CRF algorithm and the SpERT model.

[0013] The above-mentioned tests were conducted using actual desensitized defense statements and judgment documents. Based on the test results, it is planned to use the actual desensitized defense statements and judgment documents to construct the case trial elements and their claim processing, judgment reasoning, applicable laws, and judgment results as training data sets.

[0014] As mentioned above, an integrated expert mechanism is set up to combine actual trial experience, sort out the characteristic factors that affect the trial of construction project cases, and build a characteristic database of construction project contract dispute cases.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention is based on the research on the element-based extraction and analysis model technology of construction project contracts. Through the steps of automatic element extraction, element analysis, and adjudication reasoning, it can push complex case information, evidence, dispute focus, applicable laws, etc. in a typological manner, empowering trial work with science and technology, and assisting judges in improving the trial efficiency of construction cases.

[0016] 2. Based on the research on the element extraction and analysis model technology of construction project contracts, for difficult and complicated construction cases, through automatic element extraction and intelligent push of judgment reasoning, complex cases can be simplified, the time spent by judges on judgment reasoning can be reduced, and the trial time of construction contract dispute cases can be shortened.

[0017] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings, in which: Figure 1 It is a content relationship logic diagram.

[0019] Figure 2 Organize the diagram for the elements.

[0020] Figure 3 This is a partial schematic diagram of the feature analysis table.

[0021] Figure 4 This is a partial schematic diagram of the feature analysis table.

[0022] Figure 5Automatically identify model diagrams for features.

[0023] Figure 6 A flowchart for legal interpretation based on a knowledge base. DETAILED DESCRIPTION

[0024] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the preferred embodiments are only for illustrating the present invention, and are not intended to limit the scope of protection of the present invention.

[0025] like Figures 1-6 As shown, the method disclosed in this invention analyzes historical case characteristics and establishes a model for extracting trial elements and content for construction cases based on BERT, BiLSTM, and CRF. It extracts trial elements from litigation requests based on the cause of action, forming an indicator system and relationship model for case claims and trial elements in construction contract disputes. Secondly, by combining the specific content of the case claims and elements, it extracts the trial elements, including the handling of the claims, the reasoning of the judgment, the applicable laws, and the judgment results, to establish an element-based intelligent assisted trial knowledge base for construction cases. This helps clarify the key elements in the trial of construction contract disputes, supporting judges in efficiently handling cases.

[0026] Furthermore, the sorting out of the elements of construction contract dispute trials includes sorting out the litigation requests and case element indicators, forming a tree diagram, an element analysis table and an element questionnaire, so that it can be used in the element analysis table to sort out whether each element belongs to the subject issue, jurisdiction issue, contract conclusion issue, contract performance issue, project quantity, project payment and interest issue and breach of contract liability, and make good classification of the litigation issues.

[0027] Furthermore, the case claims and trial element indicators are constructed based on the litigation materials of the case. The BERT language model is used to extract text features to obtain the word-granularity vector matrix, and BiLSTM is used to extract contextual information. At the same time, the CRF model is combined to extract the global optimal sequence, and finally the element indicators of the case trial are obtained.

[0028] Furthermore, we used actual desensitized defense statements and judgments for testing, and adopted an algorithm model to quickly extract and form a set of factor indicators. We also had industry experts manually verify and label the newly added factor indicators to verify the validity of the factor indicators. Finally, we obtained a relationship model between the case claims and trial factor indicators of construction contract disputes, and optimized the accuracy of the relationship model through labeling.

[0029] The above-mentioned BERT language model is called Bidirectional Encoder Representations from Transformers (Bidirectional Transformer Model). BiLSTM stands for Bidirectional Long Short-Term Memory (Bidirectional Long Short-Term Memory). The full name of the CRF model is (Conditional Random Field, Conditional Random Field) Furthermore, the BERT language model is used to extract text features to obtain a word-granularity vector matrix, and BiLSTM is used to extract contextual information. Combined with the CRF model, the globally optimal sequence is extracted, and a factor indicator content extraction model is established. Furthermore, a SpERT-based subject relationship extraction model is constructed to identify the parties involved in construction contracts between co-development parties and contractors in joint development and construction projects, determine the responsibilities of the co-development parties, identify the parties involved in multiple subcontracting contracts in joint development and construction projects, and distinguish between joint contracting and subcontracting. In cases of internal contracting and subcontracting, the parties involved are identified and distinguished, resolving the issue of identifying contract parties in complex construction contract disputes.

[0030] The content of litigation requests and trial element indicators is automatically identified and extracted using the BERT+BiLSTM+CRF algorithm and the SpERT model.

[0031] The test was conducted using actual desensitized defense statements and judgment documents. Based on the test results, it is planned to construct a training data set of case trial elements and their claim handling, judgment reasoning, applicable laws, and judgment results through actual desensitized defense statements and judgment documents.

[0032] SpERT stands for (Span-based Entity and Relation Transformer, joint entity and relationship extraction model) At the same time, an expert integration mechanism has been set up to combine actual trial experience to sort out the characteristic factors affecting the trial of construction project cases and build a characteristic database of construction project contract dispute cases.

[0033] The present invention is specifically described as follows: 1. Task Requirements Analysis Construction contract disputes generally have longer trial periods than other contract disputes due to the high degree of professionalism in construction, the involvement of multiple litigants, the complexity of legal relationships, and the large amount of evidence. To ascertain the facts of a case, judges often need to have a certain level of professional knowledge in construction, based on their familiarity with the relevant laws and regulations, and make a comprehensive determination based on the evidence submitted by the parties. However, it is common to encounter practical issues in the case handling process that are left blank in the laws and regulations. After summary, the demand for intelligent assistance is mainly proposed in the following aspects: 1. Assisted identification of the parties to a construction project contract. The most common parties to a construction project contract are the contractor and the subcontractor. However, due to the existence of subcontracting, sub-contracting, and affiliation, there are often other parties to a case such as illegal subcontractors, subcontractors, and actual construction personnel. Among the contractors, there are situations where two or more companies cooperate to develop real estate or set up project companies, project departments, or consortiums to bid. Among the contractors, there are problems such as internal contracting, subcontracting, or affiliation. Therefore, the correct distinction and identification of each litigation subject is the first step in characterizing the legal relationship of the case and determining the liability of the parties. Through intelligent analysis of case materials, this project can extract the elements of the subject, and for problems with multiple litigation subjects / complex legal relationships, it can enhance the intuitiveness of the subject relationship by constructing a portrait of the subject relationship and visually assisting judges in identifying the subject.

[0034] 2. Assist in determining the types of dispute focus in construction contract dispute cases. In the relevant cases accepted by the People's Court, the main causes of disputes include project payment, project quality, breach of contract, warranty / repair / labor insurance funds and other non-direct project funds. The focus of the dispute is the key issues of evidence, facts and legal application summarized by the judge and recognized by the parties. It is not only the main content of the trial, but also the main line of the preparation of judgment documents, which facilitates the organization of evidence identification, fact identification and reasoning. This project obtains case data and collects case characteristics. In view of the large amount of evidence in the case and the problems existing in the judge's later preparation of judgment documents, it explores and assists in determining the focus of the dispute. On the one hand, it assists in evidence identification, and on the other hand, it assists the judge in summarizing the "central idea" of the judgment document.

[0035] 3. Legal Clarification for Construction Contract Dispute Cases. Due to the lack of legal knowledge and litigation experience among parties, which can lead to confusion about the content of the litigation and a lack of understanding or clarity about the outcome, this project designs a knowledge-based reasoning model for legal clarification for parties. This model deeply explores the inherent logic of case information and infers legal clarifications from the legal and judicial knowledge base. This model assists judges in providing legal clarifications relevant to their cases, enabling them to understand the legal knowledge involved in their cases. This helps judges in their work, reduces their burden, and accelerates the case turnover.

[0036] Research on intelligent trials for construction cases based on a factor-based approach meets the realities of judicial practice, the public's need for justice, and the demands for AI judicial applications. It addresses three practical requirements: 1) improving the quality and efficiency of judicial trials and effectively resolving backlogs; 2) promoting uniform application of law; and 3) optimizing resource allocation. Regarding the public's need for justice, it can enable efficient dispute resolution and enhance a sense of fairness and justice.

[0037] II. Overall Objectives of the Project The construction industry is one of my country's important industries, and construction contract disputes have always been difficult and complex cases in the civil cases heard by the People's Courts. They encounter many difficult issues, and in judicial practice, courts across the country have different views on this. This project addresses the difficulties in construction contract disputes in terms of dispute focus, contract party identification and relationship building, and legal interpretation. Through in-depth analysis of the characteristic elements of historical cases and based on deep learning, statistical classification and other algorithmic technologies, it can achieve contract party identification and relationship building, automatic judgment of dispute focus type and legal interpretation in individual construction contract dispute cases. It clarifies the key elements in the trial of construction contract disputes, solves practical problems, improves trial efficiency, promotes the healthy and orderly development of the construction market, and maintains social stability. It also guarantees the right of both parties to the construction contract to independently dispose of the rights, and can also be supplemented by the guidance of judges to ensure that construction contract disputes can be correctly heard and adjudicated in accordance with the law.

[0038] 1.1 Technical Route In order to solve the problems in construction contract dispute cases in terms of dispute focus, contract party identification and relationship building, and legal interpretation, and to achieve the research objectives, this project will be carried out from five aspects: service-oriented characteristic elements of construction contract dispute cases, dispute focus type identification technology based on case factors, contract party identification and relationship building technology, legal interpretation technology for case legal analysis, and intelligent auxiliary system for construction contract dispute case trials. The research content relationship logic diagram is as follows: Figure 1 shown.

[0039] 3. The technical contents of the research are as follows: 1. Review of trial elements Based on laws and regulations, legal experts have sorted out the elements of 9 types of cases in construction engineering contract disputes, such as construction engineering survey contract disputes and construction engineering design contract disputes, and formed 39 types of litigation requests and 600+ case elements, forming element tree diagrams, element analysis tables and element questionnaires, such as Figure 2 Cause of action: Summary of elements of construction engineering survey contract disputes.

[0040] Based on the element analysis table, sort out whether each element belongs to the subject issue, jurisdiction issue, contract conclusion issue, contract performance issue, project quantity, project payment and interest issue, breach of contract liability, etc., and then make a judgment based on the type of issue and element content. See the figure for part of the element analysis table. Figure 3 and Figure 4 .

[0041] 2. Identification of contracting parties in construction contract disputes under special circumstances In construction contract disputes, different contracting methods and different contracting parties in the contracts make the relationships in the cases complicated. Among them, the existence of subcontracting, sub-contracting, and affiliation leads to multiple roles such as subcontractors, sub-contractors, and actual construction personnel. There are special circumstances such as unclear legal relationships and unclear rights and responsibilities in construction contracts, mainly including the identification and responsibility assumption of the contractor in cooperative development construction projects, the identification and responsibility assumption of the contractor in joint development construction projects, and the identification and responsibility assumption of internal contracting and affiliation. Based on the litigation materials of the case, the BERT language model is first used to extract text features to obtain the word-granularity vector matrix, and BiLSTM is used to extract contextual information. At the same time, the CRF model is combined to extract the global optimal sequence, and finally the contract subject is obtained; then, role information is annotated according to historical cases, and a subject relationship extraction model based on SpERT is constructed to identify the subjects of the construction contracts signed by the various cooperative development parties and contractors in the cooperative development and construction projects, determine the responsibilities of the cooperative development parties, identify the subjects in the multiple contracting contracts generated by subcontracting in the joint development and construction projects, and distinguish between joint contracting and subcontracting. In the case of internal contracting and affiliation, the subjects are identified and distinguished between internal contracting and affiliation, solving the problem of identifying the contract subjects in construction contract disputes under special circumstances. The automatic element recognition model is shown in the figure. Figure 5 .

[0042] 3. Construction of a case feature database for construction contract disputes In actual engineering construction, construction engineering contract disputes often occur, and they are characterized by a large number of cases, large dispute amounts, complex legal relationships, long trial difficulties, long trial cycles, and a large number of parties involved. Analyzing the characteristics of such cases involves a wide range of aspects and requires high professionalism. It is also necessary to distinguish between the effectiveness and management provisions of laws and regulations, while taking into account the quality of the project, balancing the interests of both parties, and adhering to the order of the construction market. For contract disputes, it is also necessary to consider the signing of the contract, the performance of the contract, the legal effect of the project subcontracting, the contractor's priority right to payment of the project price, and the attribution principles and constituent elements of the contractual breach of contract liability from the perspective of the identification subject. This project research integrates the expert mechanism and combines actual trial experience to sort out the characteristic factors that affect the trial of construction engineering cases and construct a characteristic database of construction engineering contract dispute cases.

[0043] IV. Model Application The research of this project is of great reference value for the adjudication of cases involving multiple subjects and the relationship between the subjects. However, the analysis of a single contract evidence often does not fully display the relationship between the subjects. This project aims to address the current gap in intelligent assisted analysis of such cases in society. Based on the contract evidence of construction contract disputes, it deeply studies NLP technology and relationship building technology, deeply explores the subjects involved, constructs a relationship diagram of the subjects involved, and realizes subject relationship knowledge reasoning, providing data-level support for downstream tasks.

[0044] During the trial process, the focus of the dispute is paramount. The key to this research is how to automatically and quickly extract the types of dispute points from a large amount of case text information. This project first summarizes the types of dispute points in construction contract disputes from a large number of judicial documents. These documents cover the disagreements between the parties in most cases. This project then merges millions of dispute points based on different representations of similar disagreements. Within a given dispute type, feature labels are established to automatically recommend dispute types based on the identification of case characteristics.

[0045] 5. Fine-tuning training This project mainly focuses on the difficult problems in construction contract dispute cases in terms of dispute focus, contract party identification and relationship building, and legal interpretation. Through in-depth analysis of the characteristic elements of historical cases, based on deep learning, statistical classification and other algorithmic technologies, it realizes the identification and relationship building of contract parties in individual construction contract dispute cases, automatic judgment of dispute focus types and legal interpretation, clarifies the key elements in the trial of construction contract dispute cases, solves practical problems and improves trial efficiency. Specifically, it carries out five aspects, including service-oriented research on characteristic elements of construction contract dispute cases, research on dispute focus type identification technology based on case factors, research on contract party identification and relationship building technology, research on legal interpretation technology for case legal analysis, and research and development of intelligent auxiliary systems for the trial of construction contract dispute cases. The above research content requires not only proficiency in laws and regulations, but also familiarity with intelligent research and application of such cases.

[0046] The amount of relevant data used initially was insufficient, and the algorithms and models require further optimization. First, we will optimize the accuracy of automatic identification of dispute focus types. We plan to test this using actual desensitized defense statements. Based on the test results, we intend to construct a training dataset from these desensitized defense statements to optimize the accuracy of automatic identification of dispute focus types. Second, we will optimize the accuracy of extracting and identifying contractual relationships. We plan to test this using actual desensitized indictments. Based on the test results, we intend to construct a training dataset from these desensitized indictments to optimize the accuracy of extracting and identifying contractual relationships.

[0047] 6. Usage Scenarios Based on the technical research on the element-based extraction and analysis model of construction engineering contracts, for difficult and complex construction cases, the automatic extraction of elements and the intelligent push of judicial reasoning can simplify complex cases, reduce the time judges spend on judicial reasoning, and shorten the trial time of construction contract dispute cases. At the same time, the timely update and iterative upgrade of the algorithm library and reasoning library can promote the standardization and regularization of case trials. Based on the technical research on the element-based extraction and analysis model of construction engineering contracts, through the steps of automatic extraction of elements, analysis of elements, and judicial reasoning, complex case information, evidence, dispute focus, applicable laws, etc. can be pushed in a typological manner, realizing the empowerment of technology to trial work and assisting judges in improving the trial efficiency of construction cases.

[0048] The above description is only a preferred embodiment of the present invention and does not constitute any form of confidentiality restriction on the present invention. Any simple modification, equivalent change and modification of the above embodiment that does not deviate from the content of the technical solution of the present invention and is based on the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A method for intelligent assisted trial of construction cases based on elements, characterized by: This method analyzes the characteristics of historical cases and establishes a construction case trial factor and content extraction model based on BERT+BiLSTM+CRF. It extracts the trial factors of litigation requests according to the case causes, and forms a case claim and trial factor indicator system and relationship model for construction contract disputes. Secondly, based on the specific content of the case claims and factors, it extracts the trial factor content of the case and its claim handling, judgment reasoning, applicable laws, and judgment results, and establishes an element-based intelligent assisted trial knowledge base for construction cases, which helps to clarify the key factors in the trial of construction contract disputes and support judges in handling cases efficiently.

2. The element-based intelligent assisted trial method for construction cases according to claim 1 is characterized in that: Sort out the elements of the trial of construction contract disputes, including the sorting out of litigation requests and case element indicators, sort out whether each element belongs to the subject issue, jurisdiction issue, contract conclusion issue, contract performance issue, project quantity, project payment and interest issue and breach of contract liability, and do a good job in classifying the litigation issues.

3. The element-based intelligent assisted trial method for construction cases according to claim 1 is characterized in that: The case pleadings and trial element indicators are constructed based on the litigation materials of the case. The BERT language model is used to extract text features to obtain the word-granularity vector matrix, and BiLSTM is used to extract contextual information. At the same time, the CRF model is combined to extract the global optimal sequence, and finally the element indicators of the case trial are obtained.

4. The element-based intelligent assisted trial method for construction cases according to claim 3 is characterized in that: We used actual desensitized defense statements and judgments for testing, and adopted an algorithm model to quickly extract and form a set of factor indicators. Industry experts then manually verified and labeled the newly added factor indicators to verify their validity. Ultimately, we obtained a relationship model between case claims and trial factor indicators for construction contract disputes, and optimized the accuracy of the relationship model through labeling.

5. The element-based intelligent assisted trial method for construction cases according to claim 3 is characterized in that: The BERT language model is used to extract text features to obtain the word-granularity vector matrix, and BiLSTM is used to extract contextual information. At the same time, the CRF model is combined to extract the global optimal sequence and establish a factor indicator content extraction model. At the same time, a subject relationship extraction model based on SpERT is constructed to identify the subjects of the construction contracts signed by the cooperative development parties and the contractors in the cooperative development and construction projects, determine the responsibilities of the cooperative development parties, identify the subjects in the multiple contracting contracts generated by subcontracting in the joint development and construction projects, distinguish between joint contracting and subcontracting, and identify the subjects in the cases of internal contracting and affiliation, so as to solve the problem of identifying the contract subjects in complex construction project contract disputes.

6. The element-based intelligent assisted trial method for construction cases according to claim 5 is characterized in that: The content of the litigation request and trial element indicators are automatically identified and extracted using the BERT+BiLSTM+CRF algorithm and the SpERT model.

7. The element-based intelligent assisted trial method for construction cases according to claim 6 is characterized in that: The test was conducted using actual desensitized defense statements and judgment documents. Based on the test results, it is planned to use the actual desensitized defense statements and judgment documents to construct the case trial elements and their claim processing, judgment reasoning, applicable laws, and judgment results as a training data set.

8. The element-based intelligent assisted trial method for construction cases according to any one of claims 1 to 7, characterized in that: Establish an integrated expert mechanism, combine actual trial experience, sort out the characteristic factors that affect the trial of construction project cases, and build a characteristic database of construction project contract dispute cases.

Images