Engineering contract management method based on artificial intelligence

Abstract

The invention provides a project contract management method based on artificial intelligence, and relates to the technical field of data processing, and the method comprises the steps: obtaining a sample project contract; performing key feature identification on the sample engineering contract to obtain a plurality of first key features included in the sample engineering contract; presetting a first combination relationship among the plurality of first key features; constructing a structural equation model according to the first combination relationship; obtaining pre-collected historical data; verifying the structural equation model by adopting historical data, and verifying whether the first path coefficient meets preset accuracy or not; if it is determined that a first path coefficient output by the structural equation model meets preset accuracy, the contract to be processed is input to the structural equation model, an output second path coefficient is obtained, and the second path coefficient is a path coefficient corresponding to a second combination relation among a plurality of second key features contained in the contract to be processed. According to the invention, the combination relationship between factors contained in the engineering contract can be quantified.

Description

Technical Field

[0001] The present application relates to the technical field of data processing, and specifically to an engineering contract management method based on artificial intelligence. Background Art

[0002] Engineering contract management refers to the entire process of systematically managing the signing, execution, supervision and performance of contracts during the implementation of engineering projects, covering the preparation and review of contract terms, cost and progress control of contract execution, contract change management, dispute and risk management, etc. Its core goal is to ensure the fulfillment of the rights and obligations of all parties to the contract through standardized management, promote the completion of projects on schedule, according to quality and within budget, reduce legal and economic risks, and improve project management efficiency and benefits.

[0003] Engineering contracts usually involve a variety of complex causal relationships, which are not only affected by explicit factors such as contract terms, project scale, and frequency of changes, but also by implicit factors such as external environment, management capabilities, and resource allocation. For example, complex factors such as project scale, technical requirements, and construction environment increase the difficulty of contract execution, and frequent design or construction changes directly lead to budget increases and uncontrollable costs. The causal relationships in engineering contracts are complex, involving multiple dimensions such as planning and execution, terms and performance, and funding and quality. These relationships need to be quantified and verified through systematic analytical methods in order to identify key drivers, optimize contract terms, reduce project risks, and improve the efficiency and effectiveness of contract management. Summary of the invention

[0004] The present application provides an engineering contract management method based on artificial intelligence, which can quantify the combined relationship between the factors included in the engineering contract.

[0005] In a first aspect of the present application, a construction contract management method based on artificial intelligence is provided, the method comprising:

[0006] Get a sample engineering contract;

[0007] Performing key feature identification on the sample engineering contract to obtain a plurality of first key features included in the sample engineering contract;

[0008] Presetting a first combination relationship between a plurality of the first key features;

[0009] According to the first combination relationship, a structural equation model is constructed;

[0010] Acquire pre-collected historical data, where the historical data is feature data extracted from historical engineering contracts;

[0011] Verifying the structural equation model using the historical data to verify whether a first path coefficient meets a preset accuracy, wherein the first path coefficient is a path coefficient corresponding to the first combination relationship obtained by quantifying the frequency of the first combination relationship by the structural equation model;

[0012] If it is determined that the first path coefficient output by the structural equation model meets the preset accuracy, the contract to be processed is input into the structural equation model to obtain the output second path coefficient, where the second path coefficient is the path coefficient corresponding to the second combination relationship between multiple second key features contained in the contract to be processed.

[0013] On the basis of the above technical solution, preferably, the key feature identification of the sample engineering contract is performed to obtain a plurality of first key features contained in the sample engineering contract, specifically including:

[0014] Based on word segmentation and part-of-speech tagging, extract the noun phrases contained in the sample engineering contract to obtain a plurality of first pre-selected features;

[0015] Based on named entity recognition, specific entities included in the sample engineering contract are identified to obtain a plurality of second pre-selected features;

[0016] Merging a plurality of the first pre-selected features and a plurality of the second pre-selected features to obtain a pre-selected feature set;

[0017] According to a predefined key feature dictionary, multiple pre-selected features in the pre-selected feature set are searched and screened in the text in a regular expression manner to extract multiple first key features.

[0018] On the basis of the above technical solution, preferably, the key feature identification of the sample engineering contract is performed to obtain a plurality of first key features contained in the sample engineering contract, specifically including:

[0019] Based on word segmentation and part-of-speech tagging, extract the noun phrases contained in the sample engineering contract to obtain a plurality of first pre-selected features;

[0020] Based on named entity recognition, specific entities included in the sample engineering contract are identified to obtain a plurality of second pre-selected features;

[0021] Merging a plurality of the first pre-selected features and a plurality of the second pre-selected features to obtain a pre-selected feature set;

[0022] According to a predefined key feature dictionary, multiple pre-selected features in the pre-selected feature set are searched and screened in the text in a regular expression manner to extract multiple first key features.

[0023] On the basis of the above technical solution, preferably, constructing a structural equation model according to the first combination relationship specifically includes:

[0024] Dividing the plurality of the first key features into first causal features and first result features according to the first combination relationship, wherein the first causal features are influencing features and the first result features are affected features;

[0025] The first causal feature is set as the observed indicator of the first result feature, and a structural equation is defined for the first result feature. The structural equation is used to describe the relationship between the first result feature and the first causal feature. The structural equation is expressed as follows:

[0026] Y=βX+ε

[0027] Among them, Y is the vector corresponding to the first outcome feature, β is the first path coefficient, X is the vector corresponding to the first causal feature, and ε is the error term.

[0028] On the basis of the above technical solution, preferably, the use of the historical data to verify the structural equation model to verify whether the first path coefficient meets the preset accuracy specifically includes:

[0029] The first key feature corresponding to the first combination relationship is standardized to obtain a first standardized feature. The standardization process is shown as follows:

[0030]

[0031] Wherein, X' is the vector corresponding to the first causal feature after standardization, X is the vector corresponding to the first causal feature, μx is the mean of the first causal feature, σx is the standard deviation of the first causal feature, Y' is the vector corresponding to the first result feature after standardization, Y is the vector corresponding to the first result feature, μy is the mean of the first result feature, and σy is the standard deviation of the first result feature;

[0032] The first covariance of the first standardized feature is calculated, specifically by the following formula:

[0033]

[0034] Among them, Cov(X',Y') is the first covariance, N is the sample size, X i ' is the vector corresponding to the first causal feature after the i-th standardized processing, Y i ' is the vector corresponding to the first result feature after the i-th normalization process;

[0035] Calculating the first path coefficient based on the first covariance and the variance of the first standardized feature;

[0036] The T value is calculated according to the first path coefficient, specifically by the following formula:

[0037]

[0038] Wherein, T is the T value, β is the first path coefficient, and P is the standard error;

[0039] It is determined whether the T value is greater than or equal to a preset threshold value, thereby determining whether the first path coefficient meets the preset accuracy.

[0040] On the basis of the above technical solution, preferably, the step of inputting the contract to be processed into the structural equation model and obtaining the output second path coefficient specifically includes:

[0041] Standardizing the second causal feature and the second result feature, wherein the second causal feature and the second result feature are the results of dividing the plurality of the second key features according to the second combination relationship, the second causal feature is an influencing feature, and the second result feature is an affected feature;

[0042] Calculate the second covariance according to the standardized second causal feature and the standardized second outcome feature;

[0043] The second path coefficient is calculated according to the second covariance and the variance of the second standardized feature, and the second standardized vector includes the second causal feature after standardization and the second result feature after standardization.

[0044] On the basis of the above technical solution, preferably, the second path coefficient is calculated according to the second covariance and the variance of the second standardized feature, specifically by the following formula:

[0045]

[0046] Among them, β ij is the second path coefficient, Cov(X i ,Y i ) is the second covariance, Var(X i ) is the variance of the second causal characteristic after standardization, Var(Y i ) is the variance of the second result feature after standardization.

[0047] In a second aspect of the present application, an artificial intelligence-based engineering contract management device is provided, the device comprising an acquisition module, an identification module, a processing module, a construction module and an output module, wherein:

[0048] The acquisition module is used to acquire sample engineering contracts;

[0049] The identification module is used to identify key features of the sample engineering contract to obtain a plurality of first key features included in the sample engineering contract;

[0050] The processing module is used to preset a first combination relationship between the plurality of the first key features;

[0051] The construction module is used to construct a structural equation model according to the first combination relationship;

[0052] The acquisition module is used to acquire pre-collected historical data, where the historical data is feature data extracted from historical engineering contracts;

[0053] The construction module is used to verify the structural equation model using the historical data to verify whether a first path coefficient meets a preset accuracy, wherein the first path coefficient is a path coefficient corresponding to the first combination relationship obtained by quantifying the frequency of the first combination relationship by the structural equation model;

[0054] The output module is used to input the contract to be processed into the structural equation model to obtain the output second path coefficient if it is determined that the first path coefficient output by the structural equation model meets the preset accuracy. The second path coefficient is the path coefficient corresponding to the second combination relationship between multiple second key features contained in the contract to be processed.

[0055] On the basis of the above technical solution, preferably, the recognition module is used to extract the noun phrases contained in the sample engineering contract based on word segmentation and part-of-speech tagging to obtain a plurality of first pre-selected features;

[0056] The recognition module is used to recognize specific entities included in the sample engineering contract based on named entity recognition to obtain a plurality of second pre-selected features;

[0057] The processing module is used to merge the plurality of the first pre-selected features and the plurality of the second pre-selected features to obtain a pre-selected feature set;

[0058] The processing module is used to search and filter multiple pre-selected features in the pre-selected feature set in the text in a regular expression manner according to a predefined key feature dictionary, and extract multiple first key features.

[0059] On the basis of the above technical solution, preferably, the construction module is used to construct a plurality of key feature combinations based on a plurality of the first key features, wherein any one of the key feature combinations contains at least two different first key features;

[0060] The processing module is used to calculate the occurrence frequency of each of the key feature combinations in the historical data, wherein if a plurality of first key features included in the key feature combination appear simultaneously in a historical engineering contract, the occurrence frequency of the key feature combination in the historical data is increased by one;

[0061] The processing module is used to screen out frequent feature combinations whose occurrence frequency is higher than a preset frequency threshold from the multiple key feature combinations;

[0062] The construction module is used to connect in series with the multiple first key features included in the frequent feature combination to construct a combination chain to obtain the first combination relationship.

[0063] On the basis of the above technical solution, preferably, the processing module is used to divide the plurality of the first key features into a first causal feature and a first result feature according to the first combination relationship, wherein the first causal feature is an influencing feature and the first result feature is an affected feature;

[0064] The construction module is used to set the first causal feature as an observation indicator of the first result feature, and define a structural equation for the first result feature. The structural equation is used to describe the relationship between the first result feature and the first causal feature. The structural equation is expressed as follows:

[0065] Y=βX+ε

[0066] Among them, Y is the first outcome feature, β is the first path coefficient, X is the first causal feature, and ε is the error term.

[0067] On the basis of the above technical solution, preferably, the processing module is used to perform standardization processing on the first key feature corresponding to the first combination relationship to obtain a first standardized feature, and the standardization processing process is expressed as follows:

[0068]

[0069] Wherein, X' is the vector corresponding to the first causal feature after standardization, X is the vector corresponding to the first causal feature, μx is the mean of the first causal feature, σx is the standard deviation of the first causal feature, Y' is the vector corresponding to the first result feature after standardization, Y is the vector corresponding to the first result feature, μy is the mean of the first result feature, and σy is the standard deviation of the first result feature;

[0070] The processing module is used to calculate the first covariance of the first standardized feature, specifically by the following formula:

[0071]

[0072] Among them, Cov(X',Y') is the first covariance, N is the sample size, X i ' is the vector corresponding to the first causal feature after the i-th standardized processing, Y i ' is the vector corresponding to the first result feature after the i-th normalization process;

[0073] The processing module is used to calculate the first path coefficient according to the first covariance and the variance of the first standardized feature;

[0074] The processing module is used to calculate the T value according to the first path coefficient, specifically by the following formula:

[0075]

[0076] Wherein, T is the T value, β is the first path coefficient, and P is the standard error;

[0077] The output module is used to determine whether the T value is greater than or equal to a preset threshold, so as to determine whether the first path coefficient meets the preset accuracy.

[0078] On the basis of the above technical solution, preferably, the processing module is used to perform standardization processing on the second causal feature and the second result feature, wherein the second causal feature and the second result feature are the results of dividing the plurality of the second key features according to the second combination relationship, the second causal feature is an influencing feature, and the second result feature is an affected feature;

[0079] The processing module is used to calculate the second covariance according to the second causal feature after the standardization processing and the second result feature after the standardization processing;

[0080] The processing module is used to calculate the second path coefficient according to the second covariance and the variance of the second standardized feature, and the second standardized vector includes the second causal feature after standardization and the second result feature after standardization.

[0081] On the basis of the above technical solution, preferably, the processing module is used to calculate the second path coefficient according to the second covariance and the variance of the second standardized feature, and is specifically calculated by the following formula:

[0082]

[0083] Among them, β ij is the second path coefficient, Cov(X i ,Y i ) is the second covariance, Var(X i ) is the variance of the second causal characteristic after standardization, Var(Y i ) is the variance of the second result feature after standardization.

[0084] In the third aspect of the present application, an electronic device is provided, including a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes any one of the methods described above.

[0085] In a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores instructions, and when the instructions are executed, any of the methods described above is executed.

[0086] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0087] 1. This application constructs a causal analysis framework based on the structural equation model, uses historical data to verify the combination relationship between key features and quantify its impact intensity. The direct, indirect and total impacts between the features are quantified through the path coefficient. By verifying the first path coefficient, the model is ensured to accurately describe the frequency and causal effect of the historical combination relationship; after the model is verified, the model is used to calculate the second path coefficient to quantify the combination relationship between the key features in the contract to be processed, and capture the interaction between variables and their potential impact on project execution.

[0088] 2. Word segmentation and part-of-speech tagging identify the core noun phrases in the contract and mine specific entities such as amount, time, and location, while predefined dictionaries and regular expression matching ensure that the recognition results are highly relevant to domain requirements. The final output key feature set comprehensively covers the factors that have a significant impact on contract management and analysis, providing a high-quality data foundation for subsequent combined relationship modeling and analysis.

[0089] 3. By standardizing key features, calculating covariance, path coefficients and T values, the frequency of combined relationships in the structural equation model is scientifically quantified. The significance of the path coefficient is evaluated through the T value test to ensure that the analysis results are statistically significant. At the same time, combined with the preset accuracy threshold, significant and reliable causal relationships can be screened out, providing a solid data foundation for subsequent model optimization and decision support. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 It is a flowchart of an engineering contract management method based on artificial intelligence disclosed in an embodiment of the present application;

[0091] Figure 2 It is a module schematic diagram of an engineering contract management device based on artificial intelligence disclosed in an embodiment of the present application;

[0092] Figure 3 It is a structural schematic diagram of an electronic device disclosed in an embodiment of the present application.

[0093] Explanation of the reference numerals: 201, acquisition module; 202, identification module; 203, processing module; 204, construction module; 205, output module; 301, processor; 302, communication bus; 303, user interface; 304, network interface; 305, memory. DETAILED DESCRIPTION

[0094] In order to enable technicians in this field to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0095] In the description of the embodiments of the present application, words such as "for example" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "for example" or "for example" is intended to present related concepts in a specific way.

[0096] In the description of the embodiments of the present application, the meaning of the term "multiple" refers to two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "include", "comprise", "have" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0097] Engineering contract management is the systematic management of the entire process of contract signing, execution, supervision and performance, aiming to ensure that all parties fulfill their rights and obligations, complete the project on time and with quality, and reduce risks. The complex causal relationships in the contract are caused by explicit factors (such as terms, scale, and frequency of changes) and implicit characteristics (such as external environment and management capabilities). For example, project complexity and frequency of changes will significantly increase costs and risks. These multi-dimensional relationships need to be quantified and verified through systematic analysis to identify key drivers, optimize contract terms, and improve management efficiency and project effectiveness.

[0098] This embodiment discloses an engineering contract management method based on artificial intelligence, referring to Figure 1 , including the following steps S110-S170:

[0099] S110, obtain a sample engineering contract.

[0100] The embodiment of the present application discloses an artificial intelligence-based engineering contract management method applied to a server, which includes but is not limited to electronic devices such as mobile phones, tablet computers, wearable devices, PCs (Personal Computers), etc., and can also be a background server running an artificial intelligence-based engineering contract management method. The server can be implemented as an independent server or a server cluster consisting of multiple servers.

[0101] Obtaining sample engineering contracts is the first step in building an engineering contract management system. Its goal is to collect sufficiently representative and high-quality contract data to provide basic support for subsequent analysis and modeling. Sample contracts should cover different types of engineering projects (such as construction, energy, infrastructure, etc.), contract size, complexity, and regional differences to ensure the wide applicability of the model. Contracts should contain key elements (such as contract terms, project parameters, payment methods, risk terms, change terms, etc.) to avoid data loss affecting subsequent analysis.

[0102] You can obtain historical contract data of existing projects from the company's contract management department. These contract data are usually recorded in the internal contract management system or archive. Or you can use resources such as public engineering contract templates and sample contracts issued by the government or industry organizations. For example, government engineering bidding websites or standard contracts of industry associations.

[0103] S120, identifying key features of the sample engineering contract to obtain a plurality of first key features included in the sample engineering contract.

[0104] In a possible implementation, key features of a sample engineering contract are identified to obtain multiple first key features contained in the sample engineering contract, specifically including: based on word segmentation and part-of-speech tagging, extracting noun phrases contained in the sample engineering contract to obtain multiple first pre-selected features; based on named entity recognition, identifying specific entities contained in the sample engineering contract to obtain multiple second pre-selected features; merging multiple first pre-selected features and multiple second pre-selected features to obtain a pre-selected feature set; and according to a predefined key feature dictionary, searching and filtering multiple pre-selected features in the pre-selected feature set in the text in a regular expression manner to extract multiple first key features.

[0105] Specifically, use NLP word segmentation tools (such as Python's jieba, spaCy or NLTK) to segment the contract text and divide the sentences into separate words or phrases. For example, the word segmentation results for "the total amount of the project is 5 million yuan" are: ["project", "total amount", "for", "5 million yuan"]. Then perform part-of-speech tagging on the word segmentation results (such as noun NN, verb VB, numeral CD, etc.), and focus on extracting noun phrases (NP). For example: input word segmentation results: ["project / NN", "total amount / NN", "for / VB", "5 million yuan / CD"], extraction results: ["project", "total amount"]. Finally, combine the part-of-speech tagging rules to extract multi-word noun phrases, such as [noun + quantity]: such as "project contract amount" + "5 million yuan" to get "project contract amount 5 million yuan", [adjective + noun]: such as "key contract terms".

[0106] Furthermore, the NER model (Python's spaCy, stanza or BERT pre-trained model) is used to identify specific entity categories involved in the contract text. For example: Original text: Party A XX Group promises to pay 5 million yuan for the project before June 30, 2023. Entity recognition result: ["XX Group / ORG", "2023 June 30 / DATE", "5 million yuan / MONEY"]. For common entities in engineering contract texts (such as "project number", "change clause", "liquidated damages"), the entity recognition effect in specific fields can be improved by annotating sample data and fine-tuning the NER model.

[0107] Merge the noun phrases extracted by word segmentation and part-of-speech tagging (the first pre-selected feature) and the specific entities extracted by NER (the second pre-selected feature) to obtain a pre-selected feature set, and remove duplicates from the merged feature set to ensure that each feature appears only once. Store the pre-selected feature set in the form of a list or table, including the feature name and source label (such as "word segmentation extraction" or "NER extraction").

[0108] Combined with the actual needs in the field of engineering contract management, a dictionary containing common key features is pre-defined. For example: project related: ["engineering project", "contract amount", "construction period", "quality requirements"]; risk management: ["liquidated penalty", "contract change", "progress delay"]; payment terms: ["payment ratio", "payment time", "final payment"]. Finally, regular expressions are used to search for content matching the key feature dictionary in the pre-selected feature set, and features matching the dictionary are extracted as the first key feature set. Example regular expression: Matching amount: \d+10,000 yuan.

[0109] S130, presetting a first combination relationship between a plurality of first key features.

[0110] In a possible implementation, a first combination relationship between multiple first key features is preset, specifically including: based on the multiple first key features, constructing multiple key feature combinations, wherein any key feature combination contains at least two different first key features; calculating the frequency of occurrence of each key feature combination in historical data, wherein, if multiple first key features contained in the key feature combination appear simultaneously in a historical engineering contract, the frequency of occurrence of the key feature combination in the historical data is increased by one; screening out frequent feature combinations whose occurrence frequency is higher than a preset frequency threshold from the multiple key feature combinations; and constructing a combination chain by connecting in series the multiple first key features contained in the frequent feature combination to obtain a first combination relationship.

[0111] Specifically, first, the first key feature set (such as contract amount, construction period, change frequency, etc.) extracted from the sample engineering contract is used to generate a combination of any two or more first key features using the permutation and combination method. For example, for the first key feature set {X1, X2, X3, X4}, the generated feature combination may include (X1, X2), (X1, X3), (X1, X3, X4), (X2, X3), (X2, X4), (X1, X2, X3, X4). The feature combination contains at least two different first key features. According to specific needs, the maximum length of the feature combination is set (such as up to 3 key features) to control the computational complexity.

[0112] According to the input amount of historical engineering contract data and key feature combinations, the historical engineering contract data is traversed to check whether each contract contains all the key features of a certain feature combination. If all the features of a certain key feature combination appear in a contract, its occurrence frequency is increased by one.

[0113] Then calculate the frequency of occurrence of each key feature combination in historical data, frequency of occurrence = number of occurrences of key feature combination / total number of historical engineering contracts. During the briefing, screen out frequent feature combinations with a frequency higher than the preset frequency threshold, and set the preset frequency threshold according to business needs, such as 0.3 (indicating that a certain feature combination must appear in more than 30% of historical contracts). Traverse the frequency results of each feature combination, retain the feature combinations with a frequency greater than or equal to the preset frequency threshold, and discard the rest.

[0114] Finally, by connecting the key features in the frequent feature combinations in series, it is possible to generate possible combination chains, and then determine the possible combination order between features based on business logic or feature semantics. For example, the assumed combination relationship of "project complexity", "contract change frequency" and "cost overrun" is: the more complex the project, the higher the frequency of contract changes, which in turn leads to cost overruns. By connecting the frequent feature combinations in series, the related combinations are merged into a combination chain.

[0115] S140, constructing a structural equation model according to the first combination relationship.

[0116] First, clarify which first key features are causal features and which first key features are result features. Divide features according to the first combination relationship (such as causal chain) mined in the early stage, and causality becomes a feature that affects other features, usually an independent feature (such as "project complexity" and "contract change frequency"). Result features are features affected by other features, usually cause features (such as "cost overrun" and "construction delay"). Ensure that the causal relationship logic after division is reasonable. For example: "project complexity" should be used as a causal feature to affect "contract change frequency"; "contract change frequency" should be used as a causal feature to affect "cost overrun".

[0117] Establish a measurement equation to describe the direct impact of the first causal feature on the first result feature. Select observation indicators for the causal feature. The source of the observation indicator can be a single feature value, such as the direct value of "contract change frequency", or an aggregated feature, such as "project complexity" may be measured by indicators such as "number of clauses" and "frequency of design changes".

[0118] Then, the structural equation is defined for the first outcome feature. The structural equation is used to describe the relationship between the first outcome feature and the first causal feature. The structural equation is expressed as follows:

[0119] Y=βX+ε

[0120] Among them, Y is the vector corresponding to the first outcome feature, β is the first path coefficient, which indicates the strength of the influence of the causal feature on the outcome feature. X is the vector corresponding to the first causal feature, and ε is the error term, which indicates the unexplained fluctuation.

[0121] If the outcome characteristic is affected by multiple causal characteristics, the extended structural equation is:

[0122] Y=β i X i +ε i

[0123] Among them, Y is the vector corresponding to the first result feature, β i is the first path coefficient of the ith path, X i is the vector corresponding to the first causal feature of the i-th order, ε i is the ith error term.

[0124] Through structural equations, the influence of causal characteristics on outcome characteristics is quantified. For example, the cost overrun (outcome characteristic) in a certain engineering contract is affected by the following contract change frequency (causal characteristic), project complexity (causal characteristic) and resource allocation efficiency (causal characteristic). Another example is that the total cost (causal characteristic) in a certain engineering contract is affected by various sub-costs (causal characteristics).

[0125] S150, obtaining pre-collected historical data.

[0126] Obtaining pre-collected historical data and extracting feature data from historical engineering contracts is an important step in building an engineering contract management system and subsequent model analysis. According to the needs of contract management, the key features that need to be extracted are clearly defined, such as: Contract terms: project amount, payment method, construction period, breach of contract clauses. Project execution: change frequency, cost overruns, and schedule delays. Risk factors: force majeure events, quality issues.

[0127] Historical engineering contract data can be obtained from a variety of channels, such as project management systems: such as ERP (Enterprise Resource Planning System), contract management systems, to obtain structured data, or archive management systems to search for historical contract texts and related attachments.

[0128] S160, using historical data to verify the structural equation model to verify whether the first path coefficient meets the preset accuracy.

[0129] First, the first key feature is standardized to eliminate the impact of different dimensions or numerical ranges of the feature and the result feature, so as to facilitate subsequent calculations. The original values ​​of the first causal feature (such as "contract change frequency") and the first result feature (such as "cost overrun") are obtained from the historical contract data, and the mean and standard deviation of the first causal feature and the first result feature are calculated:

[0130]

[0131] The first key feature corresponding to the first combination relationship is standardized to obtain a first standardized feature. The standardization process is shown as follows:

[0132]

[0133] Among them, X' is the vector corresponding to the first causal feature after standardization, X is the vector corresponding to the first causal feature, μx is the mean of the first causal feature, σx is the standard deviation of the first causal feature, Y' is the vector corresponding to the first result feature after standardization, Y is the vector corresponding to the first result feature, μy is the mean of the first result feature, and σy is the standard deviation of the first result feature.

[0134] The goal of standardization is to eliminate the numerical differences between features due to different dimensions (such as "contract amount" in ten thousand yuan, "construction delay" in days), so that each feature can be compared on the same scale.

[0135] Traverse the sample data, for each pair of eigenvalues ​​X i ′,Y i 'Perform a product operation, then sum the products of all samples and divide by the total number of samples N to calculate the first covariance of the first standardized feature, which is specifically calculated by the following formula:

[0136]

[0137] Among them, Cov(X',Y') is the first covariance, N is the sample size, X i ' is the vector corresponding to the first causal feature after the i-th standardized processing, Y i ' is the vector corresponding to the first result feature after the i-th standardized processing. The covariance value range is [-1,1], positive values ​​indicate positive correlation, and negative values ​​indicate negative correlation.

[0138] According to the variance of the first covariance and the first standardized feature, the first path coefficient is calculated by dividing the first covariance by the variance of the first covariance and the first standardized feature, and the first path coefficient is calculated. The path coefficient is a core indicator for measuring the direct impact of the causal feature on the result feature. β>0 indicates that the causal feature has a positive impact on the result feature, β<0 indicates that the causal feature has a negative impact on the result feature, and the larger |β| is, the higher the impact of the causal feature is.

[0139] In order to verify whether the path coefficient is significant, the T value test is used to determine whether the path coefficient meets the preset accuracy. First, the standard error of the path coefficient is calculated, which is specifically calculated using the following formula:

[0140]

[0141] Among them, σε 2 is the variance of the difference term, which can be obtained from the model fitting, and N is the number of samples.

[0142] Then the T value is calculated according to the first path coefficient, which is calculated by the following formula:

[0143]

[0144] Where T is the T value, β is the first path coefficient, and P is the standard error. Through the T value test, confirm whether the path coefficient is statistically significant. According to business needs, preset the T value threshold. For example, the T value threshold is 1.96 (corresponding to the significance level of 0.05). If T ≥ threshold, the path coefficient meets the preset accuracy and the influence relationship is significant. If T < threshold, the path coefficient does not meet the preset accuracy and needs to be readjusted or the data needs to be verified.

[0145] S170: If it is determined that the first path coefficient output by the structural equation model meets the preset accuracy, the contract to be processed is input into the structural equation model to obtain the output second path coefficient.

[0146] First, based on the second combination relationship, the second key feature of the contract to be processed is divided into the second causal feature and the second result feature, and the acquisition process of the first causal feature and the first result feature can be referred to. In order to eliminate the dimensional difference and facilitate the subsequent calculation of covariance and path coefficients, the second causal feature and the second result feature are standardized.

[0147] Referring to the calculation process of the first path coefficient, the data of the pending contracts are analyzed using the structural equation model, and the second path coefficient is calculated and obtained to quantify the frequency of the combination relationship between the second key features. First, the second covariance is calculated based on the second causal feature after standardization and the second result feature after standardization. Then, the second path coefficient is calculated based on the second covariance and the variance of the second standardized feature (the second standardized vector includes the second causal feature after standardization and the second result feature after standardization), which is specifically calculated by the following formula:

[0148]

[0149] Among them, β ij is the second path coefficient, Cov(X i ,Y i ) is the second covariance, Var(X i ) is the variance of the second causal characteristic after standardization, Var(Y i) is the variance of the second result feature after standardization. It should be noted that the calculation principle of the second path coefficient has been introduced in detail in the calculation process of the first path coefficient, so it will not be further described here.

[0150] This embodiment also discloses an engineering contract management device based on artificial intelligence, referring to Figure 2 The device includes an acquisition module 201, an identification module 202, a processing module 203, a construction module 204 and an output module 205, wherein:

[0151] The acquisition module 201 is used to acquire a sample engineering contract.

[0152] The identification module 202 is used to identify key features of the sample engineering contract and obtain a plurality of first key features included in the sample engineering contract.

[0153] The processing module 203 is used to preset a first combination relationship between multiple first key features.

[0154] The construction module 204 is used to construct a structural equation model according to the first combination relationship.

[0155] The acquisition module 201 is used to acquire pre-collected historical data, where the historical data is feature data extracted from historical engineering contracts.

[0156] The construction module 204 is used to verify the structural equation model using historical data to verify whether the first path coefficient meets the preset accuracy. The first path coefficient is the path coefficient corresponding to the first combination relationship obtained by quantifying the frequency of the first combination relationship by the structural equation model.

[0157] The output module 205 is used to input the contract to be processed into the structural equation model to obtain the output second path coefficient if it is determined that the first path coefficient output by the structural equation model meets the preset accuracy. The second path coefficient is the path coefficient corresponding to the second combination relationship between multiple second key features included in the contract to be processed.

[0158] In a possible implementation, the identification module 202 is used to extract noun phrases contained in the sample engineering contract based on word segmentation and part-of-speech tagging to obtain a plurality of first pre-selected features.

[0159] The recognition module 202 is used to recognize specific entities included in the sample engineering contract based on named entity recognition, and obtain a plurality of second pre-selected features.

[0160] The processing module 203 is used to merge the multiple first pre-selected features and the multiple second pre-selected features to obtain a pre-selected feature set.

[0161] The processing module 203 is used to search and filter multiple pre-selected features in the pre-selected feature set in the text in a regular expression manner according to a pre-defined key feature dictionary, and extract multiple first key features.

[0162] In a possible implementation, the construction module 204 is configured to construct a plurality of key feature combinations based on a plurality of first key features, wherein any key feature combination includes at least two different first key features.

[0163] The processing module 203 is used to calculate the occurrence frequency of each key feature combination in the historical data, wherein if multiple first key features included in the key feature combination appear simultaneously in a historical engineering contract, the occurrence frequency of the key feature combination in the historical data is increased by one.

[0164] The processing module 203 is used to select frequent feature combinations whose occurrence frequency is higher than a preset frequency threshold from the multiple key feature combinations.

[0165] The construction module 204 is used to construct a combination chain by connecting multiple first key features included in the frequent feature combination in series to obtain a first combination relationship.

[0166] In a possible implementation, the processing module 203 is used to divide the multiple first key features into first causal features and first result features according to the first combination relationship, wherein the first causal features are influencing features and the first result features are affected features.

[0167] The construction module 204 is used to set the first causal feature as an observation indicator of the first result feature, and define a structural equation for the first result feature. The structural equation is used to describe the relationship between the first result feature and the first causal feature. The structural equation is expressed as follows:

[0168] Y=βX+ε

[0169] Among them, Y is the vector corresponding to the first outcome feature, β is the first path coefficient, X is the vector corresponding to the first causal feature, and ε is the error term.

[0170] In a possible implementation, the processing module 203 is used to perform standardization processing on the first key feature corresponding to the first combination relationship to obtain a first standardized feature. The standardization processing process is shown as follows:

[0171]

[0172] Among them, X' is the vector corresponding to the first causal feature after standardization, X is the vector corresponding to the first causal feature, μx is the mean of the first causal feature, σx is the standard deviation of the first causal feature, Y' is the vector corresponding to the first result feature after standardization, Y is the vector corresponding to the first result feature, μy is the mean of the first result feature, and σy is the standard deviation of the first result feature.

[0173] The processing module 203 is used to calculate the first covariance of the first standardized feature, which is specifically calculated by the following formula:

[0174]

[0175] Among them, Cov(X',Y') is the first covariance, N is the sample size, X i ' is the vector corresponding to the first causal feature after the i-th standardized processing, Y i ' is the vector corresponding to the first result feature after the i-th normalization process.

[0176] The processing module 203 is used to calculate a first path coefficient according to the first covariance and the variance of the first standardized feature.

[0177] The processing module 203 is used to calculate the T value according to the first path coefficient, specifically by the following formula:

[0178]

[0179] Where T is the T value, β is the first path coefficient, and P is the standard error.

[0180] The output module 205 is used to determine whether the T value is greater than or equal to a preset threshold, so as to determine whether the first path coefficient meets a preset accuracy.

[0181] In a possible implementation, the processing module 203 is used to perform standardization processing on the second causal feature and the second result feature, wherein the second causal feature and the second result feature are the results of dividing the multiple second key features according to the second combination relationship, the second causal feature is the influencing feature, and the second result feature is the affected feature.

[0182] The processing module 203 is used to calculate the second covariance according to the second causal feature after the standardization processing and the second result feature after the standardization processing.

[0183] The processing module 203 is used to calculate the second path coefficient according to the second covariance and the variance of the second standardized feature, and the second standardized vector includes the second causal feature after standardization and the second result feature after standardization.

[0184] In a possible implementation, the processing module 203 is used to calculate the second path coefficient according to the second covariance and the variance of the second standardized feature, specifically by the following formula:

[0185]

[0186] Among them, β ij is the second path coefficient, Cov(X i ,Y i ) is the second covariance, Var(X i ) is the variance of the second causal characteristic after standardization, Var(Y i ) is the variance of the second result feature after standardization.

[0187] It should be noted that: when the device provided in the above embodiment realizes its function, only the division of the above functional modules is used as an example. In actual application, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0188] This embodiment also discloses an electronic device, referring to Figure 3 The electronic device may include: at least one processor 301 , at least one communication bus 302 , a user interface 303 , a network interface 304 , and at least one memory 305 .

[0189] The communication bus 302 is used to realize the connection and communication between these components.

[0190] The user interface 303 may include a display screen (Display) and a camera (Camera). The optional user interface 303 may also include a standard wired interface and a wireless interface.

[0191] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0192] Among them, the processor 301 may include one or more processing cores. The processor 301 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 305, and calling data stored in the memory 305. Optionally, the processor 301 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 301 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 301, and it can be implemented separately through a chip.

[0193] Among them, the memory 305 may include a random access memory (RAM) or a read-only memory (Read-Only Memory). Optionally, the memory includes a non-transitory computer-readable storage medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may also be at least one storage device located away from the aforementioned processor 301. The memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface 303 module, and an application program for an engineering contract management method based on artificial intelligence.

[0194] exist Figure 3In the electronic device shown, the user interface 303 is mainly used to provide an input interface for the user and obtain data input by the user; and the processor 301 can be used to call an application program stored in the memory 305 for an engineering contract management method based on artificial intelligence. When executed by one or more processors 301, the electronic device executes one or more methods in the above-mentioned embodiments.

[0195] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the order of the actions described, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for the present application.

[0196] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0197] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0198] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0199] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0200] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory 305 and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned memory 305 includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk.

[0201] The present application also discloses a computer-readable storage medium, wherein the computer-readable storage medium stores instructions. When executed by one or more processors 301, the electronic device executes one or more of the methods described in the above embodiments.

[0202] The above is only an exemplary embodiment of the present disclosure, and the scope of the present disclosure cannot be limited thereto. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification and the truth of practice, it will be easy for those skilled in the art to think of other embodiments of the present disclosure. This application is intended to cover any modification, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the technical field that are not recorded in the present disclosure. The description and examples are regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. An engineering contract management method based on artificial intelligence, characterized in that: The method comprises: Get a sample engineering contract; Performing key feature identification on the sample engineering contract to obtain a plurality of first key features included in the sample engineering contract; Presetting a first combination relationship between a plurality of the first key features; According to the first combination relationship, a structural equation model is constructed; Acquire pre-collected historical data, where the historical data is feature data extracted from historical engineering contracts; Verifying the structural equation model using the historical data to verify whether a first path coefficient meets a preset accuracy, wherein the first path coefficient is a path coefficient corresponding to the first combination relationship obtained by quantifying the frequency of the first combination relationship by the structural equation model; If it is determined that the first path coefficient output by the structural equation model meets the preset accuracy, the contract to be processed is input into the structural equation model to obtain the output second path coefficient, where the second path coefficient is the path coefficient corresponding to the second combination relationship between multiple second key features contained in the contract to be processed.

2. According to claim 1, the method for engineering contract management based on artificial intelligence is characterized in that: The key feature identification of the sample engineering contract to obtain a plurality of first key features contained in the sample engineering contract specifically includes: Based on word segmentation and part-of-speech tagging, extract the noun phrases contained in the sample engineering contract to obtain a plurality of first pre-selected features; Based on named entity recognition, specific entities included in the sample engineering contract are identified to obtain a plurality of second pre-selected features; Merging a plurality of the first pre-selected features and a plurality of the second pre-selected features to obtain a pre-selected feature set; According to a predefined key feature dictionary, multiple pre-selected features in the pre-selected feature set are searched and screened in the text in a regular expression manner to extract multiple first key features.

3. The method for engineering contract management based on artificial intelligence according to claim 1, characterized in that: The preset first combination relationship between the first plurality of key features specifically includes: Based on the plurality of the first key features, construct a plurality of key feature combinations, wherein any one of the key feature combinations contains at least two different first key features; Calculating the frequency of occurrence of each of the key feature combinations in the historical data, wherein if a plurality of first key features included in the key feature combination simultaneously appear in a historical engineering contract, the frequency of occurrence of the key feature combination in the historical data is increased by one; Filtering out frequent feature combinations whose occurrence frequency is higher than a preset frequency threshold from the plurality of key feature combinations; The first key features included in the frequent feature combination are connected in series to build a combination chain to obtain the first combination relationship.

4. According to claim 1, the method for engineering contract management based on artificial intelligence is characterized in that: The constructing of a structural equation model according to the first combination relationship specifically includes: Dividing the plurality of the first key features into first causal features and first result features according to the first combination relationship, wherein the first causal features are influencing features and the first result features are affected features; The first causal feature is set as the observed indicator of the first result feature, and a structural equation is defined for the first result feature. The structural equation is used to describe the relationship between the first result feature and the first causal feature. The structural equation is expressed as follows: Y=βX+ε Among them, Y is the vector corresponding to the first outcome feature, β is the first path coefficient, X is the vector corresponding to the first causal feature, and ε is the error term.

5. The method for engineering contract management based on artificial intelligence according to claim 4 is characterized in that: The using of the historical data to verify the structural equation model and verifying whether the first path coefficient meets the preset accuracy specifically includes: The first key feature corresponding to the first combination relationship is standardized to obtain a first standardized feature. The standardization process is shown as follows: Wherein, X' is the vector corresponding to the first causal feature after standardization, X is the vector corresponding to the first causal feature, μx is the mean of the first causal feature, σx is the standard deviation of the first causal feature, Y' is the vector corresponding to the first result feature after standardization, Y is the vector corresponding to the first result feature, μy is the mean of the first result feature, and σy is the standard deviation of the first result feature; The first covariance of the first standardized feature is calculated, specifically by the following formula: Among them, Cov(X',Y') is the first covariance, N is the sample size, X i ' is the vector corresponding to the first causal feature after the i-th standardized processing, Y i ' is the vector corresponding to the first result feature after the i-th normalization process; Calculating the first path coefficient based on the first covariance and the variance of the first standardized feature; The T value is calculated according to the first path coefficient, specifically by the following formula: Wherein, T is the T value, β is the first path coefficient, and P is the standard error; It is determined whether the T value is greater than or equal to a preset threshold value, thereby determining whether the first path coefficient meets the preset accuracy.

6. The method for engineering contract management based on artificial intelligence according to claim 1, characterized in that: The step of inputting the contract to be processed into the structural equation model and obtaining the output second path coefficient specifically includes: Standardizing the second causal feature and the second result feature, wherein the second causal feature and the second result feature are the results of dividing the plurality of the second key features according to the second combination relationship, the second causal feature is an influencing feature, and the second result feature is an affected feature; Calculate the second covariance according to the standardized second causal feature and the standardized second outcome feature; The second path coefficient is calculated according to the second covariance and the variance of the second standardized feature, and the second standardized vector includes the second causal feature after standardization and the second result feature after standardization.

7. The method for engineering contract management based on artificial intelligence according to claim 6 is characterized in that: The second path coefficient is calculated according to the second covariance and the variance of the second standardized feature, specifically by the following formula: Among them, β ij is the second path coefficient, Cov(X i ,Y i ) is the second covariance, Var(X i ) is the variance of the second causal characteristic after standardization, Var(Y i ) is the variance of the second result feature after standardization.

8. An engineering contract management device based on artificial intelligence, characterized in that: The device comprises an acquisition module (201), an identification module (202), a processing module (203), a construction module (204) and an output module (205), wherein: The acquisition module (201) is used to acquire a sample engineering contract; The identification module (202) is used to identify key features of the sample engineering contract to obtain a plurality of first key features included in the sample engineering contract; The processing module (203) is used to preset a first combination relationship between a plurality of the first key features; The construction module (204) is used to construct a structural equation model according to the first combination relationship; The acquisition module (201) is used to acquire pre-collected historical data, wherein the historical data is feature data extracted from historical engineering contracts; The construction module (204) is used to verify the structural equation model using the historical data to verify whether the first path coefficient meets the preset accuracy, wherein the first path coefficient is the path coefficient corresponding to the first combination relationship obtained by quantifying the frequency of the first combination relationship by the structural equation model; The output module (205) is used to input the contract to be processed into the structural equation model to obtain the output second path coefficient if it is determined that the first path coefficient output by the structural equation model meets the preset accuracy. The second path coefficient is the path coefficient corresponding to the second combination relationship between multiple second key features included in the contract to be processed.

9. An electronic device, characterized in that: The electronic device comprises a processor (301), a communication bus (302), a memory (305), a user interface (303), a network interface (304) and a memory (305), wherein the memory (305) is used to store instructions, the user interface (303) and the network interface (304) are both used to communicate with other devices, and the processor (301) is used to execute the instructions stored in the memory (305) so that the electronic device executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 7 is performed.

Images