An Automatic Method for Evaluating the Quality of Construction Engineering Consulting Data

Abstract

This invention discloses an automatic assessment method for the quality of construction engineering consulting data, specifically relating to the field of process linkage control technology. The method involves collecting pricing caliber evidence and registering it as evidence items; extracting caliber elements to generate caliber constraint items and forming a caliber constraint set; calculating the comprehensive quantity of caliber evidence credibility strength and using it as an evidence strength field to associate with evidence items and caliber constraint items, thus forming a pricing caliber evidence set; performing field standardization and consistency of consulting business data for data storage, generating a list item set, constructing a caliber domain set, and establishing a caliber domain, data, and evidence association diagram; using the caliber domain as the smallest reasoning unit to perform pricing caliber consistency reasoning and problem detection, generating caliber quality problem records and evidence chain location information, and calculating the comprehensive quantity of caliber domain gate control risk; determining the caliber domain gate control level based on the gate control risk, outputting review tasks or supplementary evidence work orders, and writing back and triggering incremental reassessment to update the results.

Description

Technical Field

[0001] This invention relates to the field of process linkage control technology, and more specifically, to an automatic evaluation method for the quality of building engineering consulting data. Background Technology

[0002] Engineering cost consulting data quality assessment technology mainly targets the automatic verification and consistency assessment of data such as bill of quantities measurement, pricing elements, fee rules and evidence materials generated during the cost consulting process. It realizes more effective, reliable and traceable closed-loop management of process measurement payment review, change order review and final settlement audit. It is applicable to the whole process cost management of construction projects, construction process cost control, contract price adjustment and audit verification. The data processing method based on evidence-driven pricing caliber verification can implement quota version, tax rate policy, fee base and applicable conditions to specific bill of quantities and scope of application. It is often used to reduce the risk of caliber disputes and rework during review.

[0003] The existing technology has the following shortcomings: In the past, data quality control in cost consulting mainly relied on format checks, arithmetic verification, and threshold screening. Typically, only static checks were performed on quantities, unit prices, total prices, and field completeness. This approach failed to extract pricing evidence such as contract terms, supplementary agreements, and approval notices into calculable structured constraints. It also lacked domain-based classification and consistency reasoning for dimensions such as building, specialty, contract section, and measurement period. As a result, it was difficult to identify in a timely manner issues such as incorrect quota versions, incorrect tax rates or tax items, improper selection of fee bases, mixed use of regulatory fees / management fees, and missing items in the composition of comprehensive unit prices. At the same time, the sources of evidence were scattered and there were issues such as version iterations, differences in signature status, and unclear effective periods. Traditional methods could not form a traceable chain of evidence and explain the location of conflicts. Consequently, disputes were exposed in the payment review or settlement audit stage, resulting in rework, extended cycles, and increased risks.

[0004] To address the above problems, this invention proposes a solution. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide an automatic assessment method for the quality of architectural engineering consulting data to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: An automatic assessment method for the quality of construction engineering consulting data includes the following steps: Collect pricing caliber evidence from construction engineering cost consulting services and register it as evidence items. Perform caliber element extraction and structuring processing on the evidence items to generate caliber constraint items and form a caliber constraint set. Calculate the comprehensive caliber evidence credibility strength based on the evidence items and associate it as an evidence strength field with the evidence items and caliber constraint items. Compile the registered evidence items to form a pricing caliber evidence set. The data from consulting services is processed for field standardization and consistency before being stored in the database to form a structured list of items. Based on the engineering organization structure and the applicable conditions of the scope constraints, a scope domain set is constructed and the list items are mapped to the scope domains. A correlation diagram of scope domains, data, and evidence is constructed to uniformly represent the relationship between scope domains, list items, scope constraints, and evidence items. Using the caliber domain as the smallest reasoning unit, the system performs consistency reasoning and problem detection based on the caliber domain, data, and evidence association diagram to generate caliber quality problem records and associate evidence chain location information. The system aggregates the comprehensive quantity of caliber evidence credibility strength of evidence items within the caliber domain to obtain the caliber domain evidence support degree. The system also combines caliber split degree, payment chain sensitivity, caliber domain evidence support degree, and irresolvable conflict degree to calculate the comprehensive quantity of caliber domain gated risk. Based on the comprehensive risk level of the caliber domain, the caliber domain gate level is determined and the corresponding gate handling is executed to generate review tasks or supplementary certification work orders. The review conclusions and supplementary certification results are written back to the pricing caliber evidence set and caliber constraint set. When an update of evidence or data related to the caliber domain is detected, an incremental reassessment is triggered to update the comprehensive risk level of the caliber domain gate and the caliber domain gate level.

[0007] In a preferred embodiment, the registered evidence items are compiled to form a pricing caliber evidence set. The pricing caliber evidence set includes at least one of the following: evidence item identifier, submitter identifier, and data source system identifier, evidence strength field, and association index with caliber constraint items.

[0008] In a preferred embodiment, the evidence of pricing caliber includes at least one or more of the following: special terms of the contract and their pricing stipulations, supplementary agreements and amendment supplementary terms, pricing caliber descriptions in the tender documents, pricing caliber descriptions in the bid price, audit caliber descriptions or owner-confirmed caliber descriptions, local quotas and their version descriptions, tax rate policy notices or tax rate change documents, fee schedules and rate tables, approval documents for materials supplied by the owner or provisional valuation materials, and signed and sealed visa slips or change orders.

[0009] In a preferred embodiment, when an evidence item is entered into the database, the evidence element information is fixed and a page coordinate index is generated; the evidence element information includes at least the submitter or submission time, document version number or content summary fingerprint, page range, clause paragraph index, and table cell coordinates; the page coordinate index is used to anchor the evidence chain location information to achieve traceable location of the evidence item in the original document.

[0010] In a preferred embodiment, the scope element includes at least the quota version identifier, pricing mode identifier, tax rate type and tax rate value, fee base and rate for regulatory fees, fee base and rate for management fees, fee base and rate for profit, calculation method for provisional fees, and settlement rules for materials supplied by Party A or provisionally valued materials. The scope element and the scope application conditions are jointly encapsulated into a scope constraint item. The scope constraint item carries at least the source evidence identifier and clause anchor positioning information.

[0011] In a preferred embodiment, the comprehensive measure of the credibility strength of caliber evidence is composed of one or more of the following: effectiveness coverage, traceability completeness, extraction stability, and conflict pressure, and is normalized to form a comparable evidence strength field.

[0012] In a preferred embodiment, the consulting business data includes at least monthly measurement list items and change list items, as well as associated quantities, units of measurement, comprehensive unit prices, total prices, cost categories, professional categories, building or area identifiers, bid or contract section identifiers, measurement periods, and measurement period start and end times; the field standardization and consistency data entry process includes at least the unified expression and dimensional verification of units of measurement, unified precision and rounding rules for amount fields, unified coding of tax-inclusive and tax-exclusive identifiers, standardized mapping of cost categories and professional categories, and noise reduction and formatting of text fields.

[0013] In a preferred embodiment, a unique key is generated for each item in the structured list of items. The unique key is generated by combining the bid section or contract section identifier, building or area identifier, measurement period, list item code, and list item name. In the case of materials supplied by the client or materials with provisional prices, the unique key further incorporates material attribute identifiers or cost category identifiers. The source information or version information of the list items is also fixed to form a version chain tracking.

[0014] In a preferred embodiment, constructing the scope domain set includes hierarchical aggregation construction: determining the boundary of the period range based on the bid section or contract section identifier, the measurement period and the measurement period time interval, and dividing the basic scope domain within the period range based on the building or area identifier and professional category; when there are independent pricing rules for materials supplied by the owner, provisional estimated materials or measure fees in the scope constraint set, a special scope domain is generated in the corresponding basic scope domain and a scope domain identifier is generated.

[0015] In a preferred embodiment, the caliber domain gating levels include at least three categories: allowable, gray-scale allowable with supplementary documentation, and blocked review. When the caliber domain gating level is blocked review, a review task is generated and gating is applied to the associated list item to restrict its entry into the payable amount summary link of the measurement payment audit. When the caliber domain gating level is gray-scale allowable with supplementary documentation, a supplementary documentation work order is generated and a risk mark is retained. Incremental reassessment is triggered when new evidence submission, changes in evidence metadata, changes in the caliber constraint set version, or list item version updates are detected.

[0016] The technical effects and advantages of this invention are as follows: This invention uses evidence items, caliber constraint items, and caliber domains as the main framework. It registers pricing caliber evidence as evidence items and extracts caliber elements to generate caliber constraint items. Furthermore, it calculates the comprehensive credibility strength of caliber evidence and solidifies it into an evidence strength field, enabling the structured expression of the evidence's applicability and credibility, which can be directly invoked for subsequent reasoning. Simultaneously, it standardizes and unifies the fields of consulting business data into a list of items, and constructs a caliber domain set by combining the project organization structure and caliber applicability conditions. Through a caliber domain, data, and evidence relationship diagram, it unifies the modeling of the relationships between list items, caliber constraint items, and evidence items, thereby enabling consistent pricing caliber reasoning and problem detection at the caliber domain level. It outputs interpretable caliber quality problem records and provides evidence chain location information, facilitating rapid backtracking and verification by reviewers in the original materials.

[0017] At the disposal level, this invention aggregates evidence within the scope domain to obtain the evidence support degree of the scope domain, and calculates the comprehensive risk of the scope domain gate control by combining the scope split degree, payment chain sensitivity, and irresolvable conflict degree. Then, it determines the scope domain gate control level and automatically generates review tasks or supplementary evidence work orders: For high-risk scope domains, blocking review can be implemented to restrict their entry into the measurement and payment review chain; for medium-risk scope domains, gray-scale release can be adopted, supplementary evidence is required, and risk marks are retained to ensure that progress and risk are controllable; when new evidence is submitted, evidence element information changes, scope constraint set version changes, or list item version is updated, incremental re-evaluation is triggered. Based on the association graph, the local scope is located, the reasoning is rerun, and the gate control results are recalculated, so that the gate control conclusions converge dynamically as the evidence is improved. This avoids repeated communication and concentrated rework caused by one-time judgment, and reduces the risk and cost of disputes such as misapplication of quotas, misuse of tax rates, and inconsistency of fee collection scope that are concentrated in the settlement stage. Attached Figure Description

[0018] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings; Figure 1 This is a flowchart illustrating an automatic assessment method for the quality of architectural engineering consulting data according to the present invention. Figure 2This is a time-series flowchart of an automatic assessment method for the quality of architectural engineering consulting data according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Example 1: An automatic assessment method for the quality of architectural engineering consulting data according to the present invention, the method flow is as follows: Figure 1 As shown, the timing flow is as follows Figure 2 As shown, it includes the following steps: Step 1: In this embodiment, Step 1 focuses on relevant data regarding pricing standards in construction engineering cost consulting services. This involves collecting pricing standard evidence, storing it in a database, extracting structural data, and generating a set of calculable pricing standard constraints. Pricing standard evidence includes: contract special terms and their pricing stipulations, supplementary agreements and amendment supplementary terms, pricing standard descriptions in bidding documents or bid quotations, audit standard descriptions or owner-confirmed standards, local quotas and their version descriptions, tax rate policy notices or tax rate change documents, fee schedules and rate tables, approval documents for materials supplied by the client and provisional price materials, signed and confirmed visa slips or change orders, etc. This evidence can be obtained through interface access, batch import, or scanned upload. Each piece of evidence is registered as a unique evidence entry upon storage, enabling continuous tracking.

[0021] To prevent duplicate versions of evidence from being overwritten later, which could lead to a loss of traceability in legal arguments, evidence metadata is fixed upon entry into the evidence repository. Evidence metadata includes at least the evidence identifier, submitter or submission time, document version number or content summary fingerprint, page range, clause / paragraph index, or table cell coordinates. For scanned documents or image evidence, a page coordinate index can also be generated without modifying the original document, ensuring that evidence entries are locatable at page numbers and anchor points for quick tracing to the corresponding clauses or table cells during subsequent review. Both the aforementioned evidence metadata and page coordinate indexes are written into the evidence repository through the evidence entries, serving as the foundation of the evidence chain.

[0022] After the evidence is solidified, the system performs caliber element extraction and structuring processing on the evidence items. For textual evidence, the system segments the content according to the clause boundaries, identifies key segments related to the pricing caliber, and extracts caliber elements from these key segments. For tabular evidence, the system identifies the table header and field semantics, completes field mapping and numerical normalization, and then extracts caliber elements. The caliber elements include at least the quota version identifier, pricing mode identifier, tax rate type and tax rate value, the fee base and rate for regulatory fees / management fees / profits, the method for calculating measure fees, the settlement rules for materials supplied by Party A and provisionally valued materials, and the applicable conditions of the caliber. The applicable conditions of the caliber are used to describe at least which profession, building or section, and measurement period or time interval the caliber element applies to, and priority information may be attached to handle situations where multiple pieces of evidence coexist. The system encapsulates the extracted criteria elements and their applicable conditions into criteria constraint entries, and retains the source evidence identifier and clause anchor point positioning information for each criteria constraint entry, so that the constraint entry can always be traced back to the original evidence in subsequent reasoning.

[0023] Considering the common issues in engineering consulting, such as inconsistent expression of the same caliber element in different pieces of evidence, significant differences in the authority and timeliness of evidence, and unstable extraction of scanned documents, relying solely on extraction results for reasoning can easily lead to the caliber constraint being skewed by low-credibility evidence, or to wavering judgments due to conflicting evidence. This embodiment calculates a comprehensive measure of the credibility strength of caliber evidence. This is used to quantify the supporting capacity of each piece of evidence for the caliber constraints under the current project and current measurement period, and to write this comprehensive quantity into the caliber constraint item associated with that piece of evidence, so that subsequent steps can be resolved and gating based on the strength of evidence. For any piece of evidence e, the system calculates: ; in, This is a normalization function used to compress the synthesis results into the range of 0 to 1, facilitating comparison and aggregation between different pieces of evidence.

[0024] parameter This indicates the coverage of validity, used to characterize the authority of the evidence and its extent of coverage of the current measurement period. In the engineering implementation, the system assigns a basic authority score to each type of evidence and signature / approval status based on a preset authority scoring table. Combined with timeliness coverage factor Calculated The specific calculation method is as follows: .in, Examples of value rules are as follows: A red-headed document with the official seal of the project legal person is 1.0; meeting minutes signed by the construction unit, supervision unit, and construction unit are 0.8; and a letter of explanation issued unilaterally by the construction unit is 0.5. The time-limited coverage factor is calculated as follows: if the evidence specifies an effective time interval of... The current measurement period is [time interval]. ,but ;in Indicates the interval length. If the evidence does not specify an effective date, the default value is used. .

[0025] parameter This indicates the completeness of traceability, used to characterize whether evidence possesses the traceability elements required for verification. The system calculates a score based on the completeness of the preset traceability elements contained in the evidence item. Specifically, let the preset set of traceability elements be U, such as: version number, content summary fingerprint, page number range, clause paragraph index, submitter, and submission time; the actual set of elements possessed by evidence item e is... ,but That is, the proportion of the number of complete items to the total number of items.

[0026] parameter This represents extraction stability, used to characterize the consistency of key elements extracted from the same evidence under two independent analyses or two preprocessing strategies. The system obtains this stability by performing two independent analyses of the evidence entries and calculating the similarity of the two extraction results on key elements. Specifically, let the sets of key elements obtained from the two analyses be respectively... and The elements include tax rate value, fixed quota version identifier, tax base and rate, etc. This refers to the Jaccard similarity coefficient. If the analysis result is a structured numerical value, then a numerical tolerance comparison is used; for example, a difference of no more than 0.01% is considered consistent.

[0027] parameter The conflict pressure is represented by the following calculation method: For each caliber element key k supported by evidence item e, let its possible values ​​in the evidence set E be... If e supports the following values exist The frequency of occurrence in is Then the conflict coefficient of this element key is . Take the average of the conflict coefficients of all feature keys supported by evidence e. Let the set of caliber feature keys supported by evidence e be . ,but: ; in, Represents a set The number of element keys in the middle, This represents the total number of distinct pieces of evidence supporting element key k in the entire evidence set. This indicates the value that evidence e provides for the feature key k. Frequency of occurrence among all possible values ​​of evidence.

[0028] After completing the above calculations, the system will The evidence strength field is bound to the evidence entry and its associated caliber constraint entry, and the caliber constraint entry, along with the source evidence chain location information, is output as a caliber constraint set. Thus, step one not only obtains a computable set of caliber constraints, but also provides a comparable evidence strength basis for each constraint, enabling subsequent caliber domain construction and consistency reasoning to be carried out under the constraints of evidence credibility strength, thereby reducing the risk of low-credibility evidence dominating the judgment or evidence conflicts leading to unstable conclusions.

[0029] Step Two: In this embodiment, Step Two takes the combined set of caliber constraints and the credibility strength of caliber evidence output from Step One. This system performs unified structured processing on various types of consulting data in cost consulting services. Based on this, it constructs caliber domains and maps bill of quantities items to caliber domains, ultimately forming a relational graph structure of caliber domains, data, and evidence. This allows subsequent consistent reasoning to proceed under clear constraints of scope and strength of evidence. The consulting data includes at least monthly measurement bill of quantities items, change order items, and their associated quantities, units of measurement, comprehensive unit prices, total prices, cost categories, professional categories, building or area identifiers, bid or contract section identifiers, measurement periods, and start and end dates. In cases involving materials supplied by the client or provisional pricing materials, the consulting data may also include material attribute identifiers, approval numbers, or provisional pricing identifiers to include them within an independent caliber scope for constraint management.

[0030] To avoid errors caused by inconsistent field definitions during subsequent processing, consulting business data should be standardized and consistent before being stored in the database. This should include at least: unified expression and dimensional verification of units of measurement; unified precision and rounding rules for monetary amounts; unified coding for tax-inclusive and tax-exclusive indicators; standardized mapping between expense categories and professional categories; and noise reduction and formatting of text fields such as item codes, project names, and specifications. Simultaneously, the source or version information of each item should be fixed, including the data source system or submitter, submission time, batch number or version number, and original document summary fingerprint. When the same item is submitted multiple times or supplemented, version chain tracking can be used to avoid the risk of mixing old and new versions later. For offline table imports or aggregation from multiple systems, the unique key of the item should be prioritized for merging. The unique key of the item can be generated by combining the bid section or contract section identifier, building or area identifier, measurement period, item code, and item name. Material attribute identifiers or expense category identifiers can also be added when necessary to prevent incorrect merging across professions or buildings. After the above processing is completed, a structured list of items can be obtained as the basic data for caliber domain mapping and inference.

[0031] After the structured list of items is formed, the system constructs caliber domains based on the project organization structure and the applicable conditions of the caliber constraint items. These caliber domains express the minimum scope of application of the same set of caliber constraints under the same pricing caliber. Their purpose is to transform caliber determination from a globally mixed approach to a domain-wide consistency, thereby avoiding misjudgment of caliber differences between different buildings, different specialties, or different periods as caliber misuse. The construction of caliber domains preferably adopts a hierarchical aggregation method: first, the period scope boundary is determined based on the bid section or contract section identifier, the measurement period, and the measurement period time interval; within the period, basic caliber domains are then divided based on the building or area identifier and professional category; when there are independent pricing rules for special objects such as materials supplied by the client, provisional price materials, and measure fees in the caliber constraint set, the system further generates dedicated caliber domains within the corresponding basic caliber domains, distinguishing them from ordinary list items in terms of constraint scope. Through the above method, the system obtains a set of caliber domains and generates a caliber domain identifier for each caliber domain to facilitate the establishment of stable reference relationships later.

[0032] Once the scope domain set is determined, the structured list item set is mapped to the scope domain. The mapping process preferably employs deterministic matching with a fallback candidate approach: for list items with complete fields and clear boundaries, they are definitively assigned to a unique scope domain based on the section or contract segment identifier, building or area identifier, professional category, measurement period or cost category, or material attribute identifier; for list items with missing fields, unclear boundaries, or potential multiple assignments, no forced classification is used, forming a candidate scope domain set and recording mapping confidence markers to avoid mismapping that could lead to overestimation of scope splitting or misuse of scopes in subsequent calculations. Mapping confidence markers can be obtained from results such as field matching completeness, coding rule matching, and neighborhood consistency of similar items, and also serve as a reference for subsequent reasoning, triggering manual review or supplementary information backfilling.

[0033] After mapping the list items to the scope domains, the system further loads the scope constraint set output in step one into each scope domain, so that each scope domain obtains the set of scope constraint items it should follow. Specifically, the system matches the applicable conditions of the scope constraint items with the domain attributes of the scope domain: when the applicable profession, applicable building or section, and applicable time interval of a scope constraint item are consistent with or cover the attributes of the scope domain, the scope constraint item is associated with that scope domain; if the same scope element has parallel candidates in multiple constraint items, the system does not directly eliminate them in this step, but retains multiple constraints and carries their source evidence and The value is then handled in subsequent steps during in-domain reasoning, taking into account the strength of evidence and the state of conflict. This is because each caliber constraint item is bound to its source evidence item and its... Therefore, the caliber domain, after being constrained, simultaneously possesses both calculable constraints and comparable evidence strength, laying the necessary input for the subsequent calculation of the risk composite quantity.

[0034] Finally, a relationship diagram is constructed to represent the relationships between caliber domains, list items, caliber constraint items, and evidence items. This diagram includes at least caliber domain nodes, list item nodes, caliber constraint nodes, and evidence nodes. Caliber domain nodes and list item nodes are connected by attribution edges to indicate which caliber domain the list item belongs to. Caliber domain nodes and caliber constraint nodes are connected by application edges to indicate which caliber constraints should be followed by that caliber domain. Caliber constraint nodes and evidence nodes are connected by source edges to indicate which evidence item the caliber constraint originates from. Evidence nodes carry the information calculated in step one. The value and scope constraint nodes carry scope elements and applicable conditions, while the list item nodes carry amount, unit, tax rate, quota version identifier, and expense category. The association diagram integrates the original business data, scope rules, and evidence chains into a single calculation object. This allows for direct specification of evidence sources and the introduction of evidence strength fields when performing consistency inference, evidence support aggregation, and gated risk calculation in the scope domain. This ensures that the assessment process is interpretable, verifiable, and consistent. After step two is completed, the system will carry... The output of the association graph, scope domain set, and list item mapping relationship is used for the consistency reasoning and risk comprehensive quantity calculation in step three.

[0035] Step 3: In this embodiment, Step 3 uses the caliber domain, data, and evidence association diagram output from Step 2 as input. It performs pricing caliber consistency reasoning and problem detection in the caliber domain, calculating the caliber quality problem set and the comprehensive caliber domain gating risk quantity. The system converts evidence, caliber constraints, and list item relationships into interpretable and verifiable quality problem records, and converges the severity, scope of impact, and evidence support capabilities of the problem into an executable risk quantity for step four to handle in a graded gating closed-loop manner.

[0036] First, the system uses the scope domain as the smallest reasoning unit to verify the consistency between scope constraints and list items. For any scope domain, the system reads the set of list items associated with that scope domain, the set of scope constraint items, and the source evidence items for the constraint items from the association graph, and merges the constraint items using the scope element key as an index. The scope element key is used to uniquely characterize a certain type of pricing scope element, such as a tax rate element key, a fixed quota version element key, or a fee base element key, and may include the applicable time interval and the scope of applicable objects to prevent cross-domain confusion. Since the same scope element key may give different values ​​in different evidence, the system maintains multiple candidate constraints within the domain and introduces an evidence strength field for support ranking: for each candidate constraint, the system reads the comprehensive quantity of the credibility strength of the scope evidence of its source evidence. This allows subsequent reasoning to proceed under the constraints of the strength of the evidence's credibility, rather than simply adopting the latest submission or fixed priority rules, thereby reducing the risk of low-credibility evidence dominating the conclusion.

[0037] During the inference process within the domain, the system generates at least four types of records for quality issues related to the scope of the data. The first type is the issue of mixed use within the domain: the system performs distribution checks on the key scope fields of the list items within the scope scope, such as statistically analyzing the different values ​​of the tax rate field, the fixed quota version field, and the fee base category field within the domain; when two or more mutually exclusive values ​​appear within the same scope scope, and the difference cannot be explained by the applicable conditions in the scope constraint set, the system determines that there is mixed use within the domain, and writes the scope of the list items involved in the mixed use, the corresponding set of values, and the proportion of their amounts into the issue record.

[0038] Secondly, there is the issue of missing certification: The system checks whether there are valid constraint entries that can cover the applicable conditions of the caliber domain for each type of key caliber element key. If a certain caliber element value appears in the list item, but there is no corresponding evidence to support that covers the domain in the set of caliber constraint entries, or there is only evidence with obviously low strength, the system marks it as missing certification and gives the missing caliber element key, the scope of the list item involved, and the recommended type of supplementary certification in the problem record so that a supplementary certification work order can be generated in step four.

[0039] Thirdly, there is the issue of conflicting interpretations: When a certain interpretation element key has multiple candidate constraints with different values, the system attempts to resolve them based on the strength of evidence and the information of the evidence elements. If there is a clear difference in the strength of evidence among the candidate constraints and the timeliness coverage of the high-strength evidence meets the measurement period requirements, the system will use the value corresponding to the high-strength evidence as the preferred interpretation and retain the eliminated candidate and its reasons in the problem record. If the strength of evidence of the candidate constraints is close or all are high-strength and the signed documents at the same level contradict each other, making it impossible to form a single interpretation, the system will mark the element key as an irresolvable conflict and output the location information of the conflicting evidence chain. The judgment criteria for the conflict of high-strength evidence at the same level are as follows: At the same time, the system determines that the authority level of these evidence items is the same. The authority level can be determined based on the evidence's... Divide into basic scores, for example: It is the first level (high authority). It is at the second level (intermediate authority). It is classified as Level 3 (low authority). If conflicting evidence... If the scores fall within the same preset level range, they are determined to be evidence of the same level. The high-intensity threshold... (For example, 0.8) and the rules for dividing the level range can be set in the system configuration. Such conflicts cannot be resolved by automatic evidence strength comparison and require manual review.

[0040] Fourthly, there is the issue of hidden missing items: The system performs completeness checks on the elements that should be included in the comprehensive unit price or fee composition. In particular, for items that should appear under different calibers, such as regulatory fees, management fees, profits, and taxes, when it is found that some list items are missing in terms of constituent elements but their total price or unit price still falls within the statistically reasonable range, the system does not directly regard them as normal. Instead, it makes a prompt mark by comparing the distribution of similar items within the caliber domain and outputs comparison evidence to avoid the hidden risk of missing items in the caliber even though the values ​​are not outliers.

[0041] After obtaining the caliber quality problem set, the system further converges the problem detection results into a comprehensive caliber domain gating risk quantity. This is to support subsequent tiered gating. Since this embodiment emphasizes evidence-driven caliber quality assessment, the system first calculates the degree of evidence support within the caliber domain. The evidence support level is used to characterize the overall support level of currently available evidence for the caliber constraint in that caliber domain. For caliber domain d, the system retrieves the evidence set for that domain from the association graph. And to comprehensively assess the credibility and strength of each piece of evidence within the set. Taking the average, we get: ; in, The degree of evidence support in the caliber ranges from 0 to 1; The set of evidence entries associated with the caliber domain d; This is the comprehensive measure of the credibility strength of the evidence calculated in step one.

[0042] After the evidence support level is calculated, the system calculates the comprehensive risk assessment of the caliber domain d. This factor is used to comprehensively reflect the degree of fragmentation in the field's reporting, the impact of the payment chain, the degree of insufficient evidence support, and the degree of irreconcilable conflict. The calculation method is as follows: ; in, The split degree is used to characterize the consistency of key elements within a caliber domain. In engineering implementation, the system calculates the proportion covered by the value with the largest amount for each preset key element category, thus obtaining the split degree. Specifically, for each key element category k, the distribution of its value across all list items within the caliber domain d is statistically analyzed, and the total price of the list item with value v is denoted as... Then the maximum consistency ratio of element k is Then the split degree of this feature category is: . Take the arithmetic mean of the split degrees of all key element categories, i.e. , where N is the total number of key element categories (e.g., N=3).

[0043] Payment chain sensitivity is used to characterize the impact of issues in this domain on the current period's payable amount; the system calculates it by combining the proportion of this domain's amount with the concentration of key expense items. The specific calculation method is as follows: .in, It equals the total price of all items in the list within the scope d divided by the amount payable in the current period; It equals the sum of the amounts of key expense items within the scope d divided by the total price of the list items within the scope d; and Let be the weight coefficient, and satisfy... In this embodiment, it is possible to take... , .

[0044] The degree of irresolvable conflict is used to characterize the extent to which there are equally strong evidentiary conflicts within the scope of the data and that these conflicts cannot be resolved. The system calculates this based on the percentage of list items with irresolvable conflicts. Specifically, let the total amount of all list items within scope d be... Among them, due to irresolvable conflicts (i.e., the existence of two or more conflicts) All are higher than the preset high-intensity threshold. (e.g., 0.8), and these pieces of evidence are mutually exclusive in value on a certain element) and the total price of the items marked as conflicting is ,but If there are no irresolvable conflicts, then .

[0045] After completing this step, output the quality problem set and the evidence support level for each caliber domain. Comprehensive Risk Measurement of Gating Domain The relationship between problem records and evidence chain location information enables step four to be generated in hierarchical gating and review tasks, and supplementary evidence tasks. After supplementary evidence or review, the evidence set and evidence strength can be updated.

[0046] Step 4: In this embodiment, Step 4 follows from Step 3, which outputs the caliber quality problem set and the caliber domain evidence support level. Comprehensive risk assessment of caliber domain gating The process transforms quality evaluation conclusions into actionable business procedures and writes the results back into the pricing caliber evidence and caliber constraints, forming a traceable and verifiable closed loop. This step processes the caliber domain and its associated list of items, caliber quality issue records, and the corresponding evidence chain location information. The output includes the caliber domain classification judgment results, review tasks or supplementary documentation work orders and their status, the gating results of the measurement and payment summary chain, and update trigger information for subsequent incremental re-evaluation.

[0047] To ensure that the gating strategy can control risks without causing payment node shutdowns, the system performs gating on the caliber domain d in this step. Perform hierarchical mapping and generate corresponding gating results. The system presets at least two gating thresholds, namely the grayscale threshold. With blocking threshold And satisfy .in, Used to define the boundary between release that can be permitted and release that requires supplementary documentation under gray-scale conditions. Used to define the boundary between requiring supplementary documentation for gray-scale release and blocking for review. System based on... The comparison with the above thresholds generates the aperture domain gating level: when When the system determines the aperture range to be open, when The system determines the caliber range as grayscale and requires supplementary certification for release; when When this occurs, the system will classify the scope as blocked from review. The above thresholds can be configured or calibrated based on project risk appetite, audit rigor, or historical review results, but they must remain consistent within the same measurement period to ensure that the gated scope is stable and reviewable.

[0048] Once the gating level of the scope domain is determined, the consultation process is handled according to the gating level. Scope domains blocked from review are deemed unusable, and their associated list items are removed or frozen in the payable amount summary chain of the measurement and payment audit. A review task is generated and bound to the scope domain and issue record. The review task contains the key issue type being blocked, the scope of the list items, the amount and percentage of the amount, and evidence chain information directly locating the evidence page number and clause anchor point, enabling reviewers to quickly trace back the scope element keys, candidate evidence, and evidence strength that led to the scope risk. To avoid the risk of scope changes or arbitrary modifications during the review process, the review task for this type of scope domain retains the reasoning path summary formed in step three, such as the scope domain's scope split degree. The corresponding splitting elements and the degree of irresolvable conflict The corresponding conflicting evidence pairs and location information provide an interpretable basis for the review conclusion.

[0049] For data categories deemed to require supplementary documentation under gray-scale release, they are not directly blocked from entering the accounts payable summary. Instead, a risk marker is added, and a report is generated. The supplementary documentation work order for this data category is sent to the responsible role or data provider. The supplementary documentation work order lists the corresponding data category element keys for missing or insufficient documentation and recommends the type of documentation to be supplemented, such as supplementing tax rate change notices, provisional valuation approval documents, or signed versions of fee schedules. The supplementary documentation work order records the current data category... and It also includes evidence chain location information, enabling those requesting supplementary evidence to understand the reasons for insufficient existing evidence and avoid repeatedly submitting invalid materials. The gray-scale release range can participate in the current period's payment calculation, but retains risk visibility in the payment review interface or output reports, so that business decision-makers can identify the credibility status of this portion of the amount and manage it by controlling the amount limit or suspending payment when necessary.

[0050] For data categories deemed acceptable for release, the system marks them as consistent and allows their associated list items to enter subsequent settlement analysis or payment review processes. Even in the acceptable state, the system retains audit traces with zero or low-risk issue records and saves the corresponding chain of evidence to ensure that the source and strength of evidence supporting the consistency of data categories can be explained during subsequent spot checks or audit tracing.

[0051] After completing the gate control process, the system writes the review conclusion and supplementary evidence results back to the pricing caliber evidence set and caliber constraint set to form a convergent closed loop. Specifically, when the reviewer confirms during the review task that the validity status of a certain evidence item has changed, such as the addition of a signature leading to an increase in authority, or the confirmation of a modification to the effective period of a clause, the system updates the metadata and applicable time period of the corresponding evidence item, and recalculates the comprehensive caliber evidence credibility strength of that evidence item under the same calculation caliber in step one. When a supplementary evidence work order submits new evidence items, the system follows the process in step one to complete the data entry and data extraction of the new evidence. Calculate and associate the new evidence with the corresponding caliber constraints. This is because the caliber constraint in step three involves assessing the support level of evidence within the caliber domain. It is within the caliber domain evidence set The aggregation results show that when the evidence set is updated or the strength of the evidence changes, the system automatically updates the scope of that caliber. This triggers the comprehensive risk assessment of the caliber domain. The recalculation allows the gating results to converge dynamically as the evidence becomes more complete, rather than remaining at a one-time determination.

[0052] To ensure project progress and business timeliness, when new evidence submissions, changes in evidence metadata, changes in definition constraints, or updates to list item versions are detected, the entire definition domain will not be re-run for inference. Instead, the scope of the definition domain and its associated list items will be located based on the association graph. Only a localized scope will be re-run for inference to ensure definition domain consistency and issue detection. , When the incremental reassessment satisfies Under certain conditions, the system will automatically remove the blocking or close the gray-scale risk marker for that caliber domain, and record the reason for removal and the triggering evidence chain in the audit log. If the threshold conditions are not met, the system will maintain the original gating level and update the pending status of the review task or supplementary certificate work order, ensuring that the handling process remains traceable and verifiable. Through hierarchical gating, write-back, and incremental re-evaluation, the automatic assessment of the pricing caliber quality will move from problem discovery to executable handling and close the loop until automatic convergence after the evidence is complete. The caliber domain has interpretable, traceable, and continuously updatable data quality assurance before payment summary and settlement analysis, reducing the rework costs and dispute risks when problems such as incorrect quotas, incorrect tax rates, and inconsistent fee collection calibers are exposed in the later stages.

[0053] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0054] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatus and methods can be implemented in other ways.

[0055] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An automatic assessment method for the quality of architectural engineering consulting data, characterized in that, Includes the following steps: Collect pricing caliber evidence from construction engineering cost consulting services and register it as evidence items. Perform caliber element extraction and structuring processing on the evidence items to generate caliber constraint items and form a caliber constraint set. Calculate the comprehensive caliber evidence credibility strength based on the evidence items and associate it as an evidence strength field with the evidence items and caliber constraint items. Compile the registered evidence items to form a pricing caliber evidence set. The data from consulting services is processed for field standardization and consistency before being stored in the database to form a structured list of items. Based on the engineering organization structure and the applicable conditions of the scope constraints, a scope domain set is constructed and the list items are mapped to the scope domains. A correlation diagram of scope domains, data, and evidence is constructed to uniformly represent the relationship between scope domains, list items, scope constraints, and evidence items. Using the caliber domain as the smallest reasoning unit, the system performs consistency reasoning and problem detection based on the caliber domain, data, and evidence association diagram to generate caliber quality problem records and associate evidence chain location information. The system aggregates the comprehensive quantity of caliber evidence credibility strength of evidence items within the caliber domain to obtain the caliber domain evidence support degree. The system also combines caliber split degree, payment chain sensitivity, caliber domain evidence support degree, and irresolvable conflict degree to calculate the comprehensive quantity of caliber domain gated risk. Based on the comprehensive risk level of the caliber domain, the caliber domain gate level is determined and the corresponding gate handling is executed to generate review tasks or supplementary certification work orders. The review conclusions and supplementary certification results are written back to the pricing caliber evidence set and caliber constraint set. When an update of evidence or data related to the caliber domain is detected, an incremental reassessment is triggered to update the comprehensive risk level of the caliber domain gate and the caliber domain gate level.

2. The automatic assessment method for the quality of architectural engineering consulting data according to claim 1, characterized in that: The registered evidence items are compiled into a pricing caliber evidence set. The pricing caliber evidence set shall include at least one of the following: evidence item identifier, submitter identifier, and data source system identifier, evidence strength field, and association index with caliber constraint items.

3. The automatic assessment method for the quality of architectural engineering consulting data according to claim 2, characterized in that: Evidence of pricing caliber shall include at least one or more of the following: special terms of the contract and their pricing stipulations, supplementary agreements and amendment supplementary terms, pricing caliber descriptions in the tender documents, pricing caliber descriptions in the bid price, audit caliber descriptions or owner-confirmed caliber descriptions, local quotas and their version descriptions, tax rate policy notices or tax rate change documents, fee schedules and rate tables, approval documents for materials supplied by the owner or provisional valuation materials, and signed and sealed visa slips or change orders.

4. The automatic assessment method for the quality of architectural engineering consulting data according to claim 2, characterized in that: When evidence items are entered into the database, evidence element information is fixed and a page coordinate index is generated. Evidence element information includes at least the submitter or submission time, document version number or content summary fingerprint, page range, clause paragraph index, and table cell coordinates. The page coordinate index is used to anchor the evidence chain location information to achieve traceable location of evidence items in the original document.

5. The automatic assessment method for the quality of architectural engineering consulting data according to claim 4, characterized in that: The scope elements should at least include the quota version identifier, pricing mode identifier, tax rate type and tax rate value, fee base and rate for regulatory fees, fee base and rate for management fees, fee base and rate for profit, calculation method for provisional fees, and settlement rules for materials supplied by Party A or materials with provisional valuation. The scope elements and applicable conditions should be encapsulated together as scope constraint items. The scope constraint items should at least carry source evidence identifiers and clause anchor positioning information.

6. The automatic assessment method for the quality of architectural engineering consulting data according to claim 1, characterized in that: The comprehensive measure of the credibility strength of caliber evidence is composed of one or more of the following: validity coverage, traceability completeness, extraction stability, and conflict pressure. After normalization, a comparable evidence strength field is formed.

7. The automatic assessment method for the quality of architectural engineering consulting data according to claim 1, characterized in that: Consulting business data should include at least monthly measurement list items and change list items, as well as associated quantities, units of measurement, comprehensive unit prices, total prices, cost categories, professional categories, building or area identifiers, bid or contract section identifiers, measurement periods, and measurement period start and end times; field standardization and consistency in data entry should include at least the unified expression and dimensional verification of units of measurement, unified precision and rounding rules for amount fields, unified coding of tax-inclusive and tax-exclusive identifiers, standardized mapping of cost categories and professional categories, and noise reduction and formatting of text fields.

8. The automatic assessment method for the quality of architectural engineering consulting data according to claim 7, characterized in that: For each item in the structured list of items, a unique key is generated. The unique key is generated by combining the bid section or contract section identifier, building or area identifier, measurement period, list item code and list item name. In the case of materials supplied by the owner or materials with provisional prices, the unique key further incorporates material attribute identifiers or cost category identifiers. The source information or version information of the list items is also fixed to form a version chain tracking.

9. The automatic assessment method for the quality of architectural engineering consulting data according to claim 7, characterized in that: The construction of the scope domain set includes hierarchical aggregation construction: the boundary of the period range is determined based on the bid section or contract section identifier, the measurement period and the measurement period time interval, and the basic scope domain is divided within the period based on the building or area identifier and professional category; when there are independent pricing rules for materials supplied by the owner, provisional estimated materials or measure fees in the scope constraint set, a special scope domain is generated in the corresponding basic scope domain and a scope domain identifier is generated.

10. The automatic assessment method for the quality of architectural engineering consulting data according to claim 9, characterized in that: The threshold levels for the scope of payment should include at least three categories: allowable, gray-scale allowable with supplementary documentation, and blocked review. When the threshold level for the scope of payment is blocked review, a review task is generated and the related list items are subject to thresholding to restrict their entry into the payable amount summary link of the measurement and payment audit. When the threshold level for the scope of payment is gray-scale allowable with supplementary documentation, a supplementary documentation work order is generated and a risk mark is retained. Incremental reassessment is triggered when new evidence submissions, changes in evidence metadata, changes in the set of constraints, or updates to the list of items are detected.

Images