Bid inviting and tendering list data processing method and device based on large model

Abstract

The invention discloses a bidding and tendering list data processing method and device based on a large model. The method comprises the steps of obtaining a to-be-evaluated list of target construction engineering; inputting the basic attribute description of the to-be-evaluated list into a pre-trained large language model, and obtaining a semantic feature vector of the to-be-evaluated list output by the large language model; comparing the semantic feature vector of the to-be-evaluated list with the semantic feature vector of each historical list group in a preset database to determine a target historical list group matched with the to-be-evaluated list; wherein the historical list group comprises a plurality of historical lists with similar semantic feature vectors; obtaining a reasonable price interval associated with the target historical list group, and evaluating the rationality of the bidding quotation in the to-be-evaluated list by using the reasonable price interval; according to the method, the bidding quotation of the to-be-evaluated list in the current construction engineering can be efficiently and accurately evaluated.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and apparatus for processing bidding list data based on a large model. Background Technology

[0002] In the field of construction project bidding, conducting a reasonable analysis of bid prices is crucial to ensuring investment efficiency and project quality. Currently, the most commonly used method for analyzing bill of quantities (BOQ) prices in the industry is the single-benchmark comparison method. This method relies solely on the rigid matching of BOQ codes and names, using the tender control price published by the tendering party or the arithmetic mean of all valid bid prices as the benchmark price, and then simply comparing each bidder's BOQ item price with the benchmark price. However, relying solely on matching codes and names results in low data utilization and a lack of objective historical project data analysis for relevant projects.

[0003] Therefore, how to efficiently and reasonably evaluate the bid prices of the bill of quantities in current construction projects has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide a method and apparatus for processing bidding list data based on a large model, which can efficiently and reasonably evaluate the bid price of the current construction project list by combining historical data.

[0005] According to one aspect of the present invention, a method for processing bidding list data based on a large model is provided, the method comprising: Obtain the list of target construction projects to be evaluated; wherein the list of projects to be evaluated includes: basic attribute descriptions and bid prices; The basic attribute description of the list to be evaluated is input into a pre-trained large language model, and the semantic feature vector of the list to be evaluated is obtained from the output of the large language model. The semantic feature vector of the list to be evaluated is compared with the semantic feature vector of each historical list group in the preset database to determine the target historical list group that matches the list to be evaluated; wherein, the historical list group includes multiple historical lists with similar semantic feature vectors. Obtain a reasonable price range associated with the target historical list group, and use the reasonable price range to evaluate the reasonableness of the bid prices in the list to be evaluated; wherein the reasonable price range is calculated based on the bid prices of each historical list in the target historical list group.

[0006] Optionally, before obtaining the list of target construction projects to be evaluated, the method further includes: Obtain historical lists of multiple historical building projects from multiple data sources; The basic attribute descriptions of each historical list are input into the large language model, and the semantic feature vectors of the corresponding historical lists output by the large language model are obtained. All historical lists are clustered by calculating the similarity between the semantic feature vectors of each historical list, forming multiple historical list groups; Based on the bid prices of each historical list in each historical list group, calculate the reasonable price range for the corresponding historical list group; Each historical list group and its corresponding reasonable price range are associated and stored in the preset database.

[0007] Optionally, the step of calculating the reasonable price range for each historical list group based on the bid prices of each historical list in each historical list group includes: The bid prices of each historical list in the historical list group are statistically analyzed using the assumption of normal distribution, and bid prices falling within two standard deviations are selected to form an effective sample set; Based on the effective sample set, the reasonable price range of the historical list group is calculated; wherein, the reasonable price range includes: reasonable minimum value, reasonable maximum value, reasonable average value and median value.

[0008] Optionally, obtaining a reasonable price range associated with the target historical list group and using the reasonable price range to assess the reasonableness of bid prices in the list to be evaluated includes: Obtain the current price of the consumables corresponding to the list to be evaluated, and parse the consumable usage information from the list to be evaluated; The cost price is calculated based on the current price of the consumables and the usage information of the consumables. A comprehensive reasonableness assessment is conducted based on the bid prices in the list to be evaluated, the reasonable price range, and the cost price.

[0009] Optionally, before obtaining the list of target construction projects to be evaluated, the method further includes: Based on the basic information of the historical building projects to which each historical list belongs, corresponding feature tags are set for each historical list; wherein, the dimensions of the feature tags include at least one of the following: cost type, project location, quotation timestamp, and professional engineering category; The feature tags of each historical list are stored in the preset database.

[0010] Optionally, obtaining a reasonable price range associated with the target historical list group and using the reasonable price range to assess the reasonableness of bid prices in the list to be evaluated includes: When the list of items to be evaluated is deemed unreasonable, feature tags are generated for the list of items to be evaluated based on the basic information of the target construction project. Filter out matching historical lists whose feature tags match the feature tags of the list to be evaluated from the target historical list group; The pricing details used to form the bid price are parsed from the list to be evaluated, and the corresponding pricing details are parsed from each of the matching historical lists; An analysis of unreasonable pricing is conducted by comparing the pricing details of the list to be evaluated with the pricing details of each matching historical list.

[0011] To achieve the above objectives, the present invention also provides a data processing device for bidding and tendering lists based on a large model, the device comprising: The first acquisition module is used to acquire the list of target construction projects to be evaluated; wherein, the list of projects to be evaluated includes: basic attribute descriptions and bid prices; The first parsing module is used to input the basic attribute description of the list to be evaluated into a pre-trained large language model and obtain the semantic feature vector of the list to be evaluated output by the large language model. The semantic matching module is used to compare the semantic feature vector of the list to be evaluated with the semantic feature vector of each historical list group in the preset database to determine the target historical list group that matches the list to be evaluated; wherein, the historical list group includes multiple historical lists with similar semantic feature vectors. The quotation evaluation module is used to obtain a reasonable price range associated with the target historical list group, and to use the reasonable price range to evaluate the reasonableness of the bid quotations in the list to be evaluated; wherein, the reasonable price range is calculated based on the bid quotations of each historical list in the target historical list group.

[0012] Optionally, the device further includes: The second acquisition module is used to obtain a historical list of multiple historical building projects from multiple data sources; The second parsing module is used to input the basic attribute description of each historical list into the large language model, and obtain the semantic feature vector of the corresponding historical list output by the large language model. The clustering module is used to cluster all historical lists by calculating the similarity between the semantic feature vectors of each historical list, forming multiple historical list groups; The calculation module is used to calculate the reasonable price range for each historical list group based on the bid price of each historical list in each historical list group. The storage module is used to associate each historical list group with its corresponding reasonable price range and store it in the preset database.

[0013] To achieve the above objectives, the present invention also provides a computer device, which specifically includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for processing bidding list data based on a large model.

[0014] To achieve the above objectives, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method for processing bidding list data based on a large model.

[0015] This invention provides a method and apparatus for processing bidding list data based on a large model. It acquires the basic attribute descriptions and bid prices of the list to be evaluated in a target construction project; inputs the basic attribute descriptions of the list to be evaluated into a pre-trained large language model, and obtains the semantic feature vectors of the list output by the large language model. This addresses the natural language differences in list description methods across different regions and projects, achieving standardization and accurate mapping of list items. The semantic feature vectors of the list to be evaluated are compared with the semantic feature vectors of various historical list groups in a preset database to determine the target historical list group that matches the list to be evaluated, solving the matching problem caused by non-standard list descriptions and providing objective data for subsequent price analysis. Finally, it acquires a reasonable price range associated with the target historical list group and uses this reasonable price range to evaluate the reasonableness of the bid prices in the list to be evaluated. Combined with historical bid prices, it efficiently and accurately analyzes the bid prices of the list to be evaluated in the current construction project. Attached Figure Description

[0016] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a schematic diagram of an optional process for processing bidding list data based on a large model, as provided in Example 1. Figure 2 This is a schematic diagram of another optional process for the bidding list data processing method based on a large model provided in Example 1; Figure 3 This is a schematic diagram illustrating the effect of the evaluation of the reasons for unreasonable pricing provided in Example 1; Figure 4 This is a schematic diagram of an optional component structure of the bidding list data processing device based on a large model provided in Embodiment 2. Figure 5 This is a schematic diagram of an optional hardware structure for the computer device provided in Embodiment 3. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0018] Example 1 This invention provides a method for processing bidding list data based on a large model, such as... Figure 1 As shown, the method specifically includes the following steps: Step S101: Obtain the list of target construction projects to be evaluated; wherein the list of projects to be evaluated includes: basic attribute descriptions and bid prices.

[0019] In this embodiment, the construction project includes multiple sub-items, and each sub-item includes multiple lists. During the bidding process, each list to be evaluated is used as the minimum bid unit, and all lists to be evaluated across all sub-items form the total bid price for the entire construction project. Each list to be evaluated includes a basic attribute description and a bid price. The basic attribute description includes code, name, characteristics, quantity, and unit, while the bid price includes unit price and total price. Characteristics are natural language text associated with the name of the list item to be evaluated, used to detail the technical parameters, material specifications, construction techniques, and installation requirements of the list. For example, if the name of the list to be evaluated is "cast-in-place structural reinforcement," the corresponding characteristics are: the type of steel reinforcement in the cast-in-place structural reinforcement is HRB400E, the specification is Φ25mm, the construction location is a second-floor frame beam, the connection method is a straight threaded sleeve mechanical connection, and the specification requirements comply with GB50204.

[0020] Step S102: Input the basic attribute description of the list to be evaluated into the pre-trained large language model, and obtain the semantic feature vector of the list to be evaluated output by the large language model.

[0021] In this embodiment, the pre-trained large language model, fine-tuned with a massive corpus of construction engineering terminology, is able to identify professional terms in construction engineering. For the current list to be evaluated, the units of all lists are first standardized. Then, the names and features of the lists are input into the pre-trained large language model. The large language model parses the input names and features to identify semantic features and standardizes the semantic features to obtain semantic feature vectors. The dimensions of the semantic feature vectors include name and features. For example, the same engineering object may be described very differently in different projects and by different cost estimators: "C30 commercial concrete, pumped", "poured commercial concrete with strength grade C30", "commercial concrete (pumped) C30", "pumped concrete C30 grade". The features of these four lists to be evaluated are input into the large language model respectively. The semantics are parsed and standardized to obtain the semantic feature vectors of the four lists to be evaluated as follows: "Name: pumped commercial concrete; Features: strength grade: C30, construction process: pumping, construction location: floor slab". By using a large language model to parse the names and features of the list to be evaluated, the standardized conversion of names and features is achieved, which solves the problem of natural language differences in the description of the list in different regions and projects.

[0022] Step S103: Compare the semantic feature vector of the list to be evaluated with the semantic feature vector of each historical list group in the preset database to determine the target historical list group that matches the list to be evaluated; wherein, the historical list group includes multiple historical lists with similar semantic feature vectors.

[0023] In this embodiment, a pre-set database stores bid prices for historical building construction projects. These historical lists have been standardized in terms of both data and semantics, and are labeled in the database in the form of semantic feature vectors. The list to be evaluated is the smallest unit of bid price in the current building construction project, and the historical list is the smallest unit of bid price in historical building construction projects. Multiple historical lists constitute a historical list group. The user matches the semantic feature vector obtained after semantic parsing the current list to be evaluated with the semantic feature vectors of the historical lists. When a historical list has a similarity to the semantic feature vector of the current list to be evaluated that is greater than a preset value, it is extracted and a target historical list group is formed, preparing for subsequent bid price evaluation. By matching semantic feature vectors with a large amount of historical data, the limitations of a single data source and the limitations of traditional keyword-based mechanical matching are overcome, significantly improving the recall and accuracy of the matching.

[0024] Step S104: Obtain a reasonable price range associated with the target historical list group, and use the reasonable price range to evaluate the reasonableness of the bid prices in the list to be evaluated; wherein, the reasonable price range is calculated based on the bid prices of each historical list in the target historical list group.

[0025] In this embodiment, the user searches for a reasonable price range that has been pre-calculated and stored in the database from the target historical list group. Then, from the historical lists within this reasonable price range, further filtering is performed to extract those with the same feature tags as the current list to be evaluated. The reasonableness of the current list to be evaluated is analyzed based on the unit price, total price, and component price of the bid quotations from these extracted historical lists. The component price is a detailed breakdown of the bid quotation, composed of multiple unit prices, including labor costs, material costs, construction equipment usage fees, management fees, and profit. Combining reasonable prices from historical data with the analysis of the current list to be evaluated makes the evaluation results more accurate and objective.

[0026] In this embodiment, by acquiring the basic attribute descriptions and bid prices of the list to be evaluated in the target construction project; inputting the basic attribute descriptions of the list to be evaluated into a pre-trained large language model, and obtaining the semantic feature vectors of the list to be evaluated output by the large language model, it is possible to cope with the natural language differences in the description methods of the list in different regions and projects, and achieve standardization and accurate mapping of list items; comparing the semantic feature vectors of the list to be evaluated with the semantic feature vectors of each historical list group in a preset database to determine the target historical list group that matches the list to be evaluated, solving the matching problem caused by non-standard list descriptions, and providing objective data for subsequent price analysis; obtaining the reasonable price range associated with the target historical list group, and using the reasonable price range to evaluate the reasonableness of the bid prices in the list to be evaluated, and combining historical bid prices, the bid prices of the list to be evaluated in the current construction project can be analyzed efficiently and accurately.

[0027] Specifically, prior to obtaining the list of target construction projects to be evaluated in step S101, the method further includes: Step A1: Obtain a historical list of multiple historical building projects from multiple data sources; This involves API integration with multiple data sources, including internal enterprise databases and industry data platforms (such as Indicator.com and Guangcai.com), and parsing structured and semi-structured cost files in formats such as GBQ and Excel to construct a multi-dimensional cost database. The data in this multi-dimensional cost database then undergoes standardization processes such as data cleaning, format conversion, and schema mapping to obtain a standardized historical bill of quantities database. The elements of the historical bill of quantities data in this database include codes, names, characteristics, quantities, units, and prices.

[0028] Step A2: Input the basic attribute description of each historical list into the large language model, and obtain the semantic feature vector of the corresponding historical list output by the large language model; Step A3: Cluster all historical lists by calculating the similarity between the semantic feature vectors of each historical list to form multiple historical list groups; Step A4: Calculate the reasonable price range for each historical list group based on the bid prices of each historical list in each historical list group; Step A5: Associate each historical list group with its corresponding reasonable price range and store them in the preset database.

[0029] In this embodiment, all historical list data in the historical list database are acquired, and the name and features of each historical list are sequentially input into a trained large language model to obtain the semantic feature vector corresponding to each historical list. After the large language model processes all historical lists, it performs cluster analysis by calculating the similarity between each semantic feature vector using a similarity algorithm. When the similarity is greater than a preset value, the corresponding historical list is grouped into multiple historical list groups. During clustering, redundant semantic feature vectors are merged, and lists with similar features in the same project are aggregated to reduce the total processing volume. Subsequently, a batch parallel processing strategy is adopted, and the task is split into multiple sub-tasks according to the unit project, which are executed concurrently using multi-threading technology. The entire clustering process adopts an asynchronous non-blocking mode, and users can view the processed grouping results in real time without waiting for all tasks to complete. For each historical list group, the bid price of each historical list is calculated sequentially, and the bid prices in each group are formed into a normal distribution to calculate a reasonable price range within the standard deviation. This reasonable price range is linked and annotated with the current historical list group and stored in the database for users to use when evaluating and analyzing bid prices for construction projects. This solves the problem of focusing solely on current project data, failing to incorporate multi-source information such as historical projects, industry indicators, and market prices, and making the analysis results susceptible to extreme values ​​or project-specific interference, lacking objectivity and comprehensiveness. This makes subsequent bid price analysis more accurate and reasonable.

[0030] In this embodiment, as Figure 2As shown, the roles of each module in processing historical list data are as follows: Data Standardization Module: First, standardize the historical list data and store the standardized historical list data in the database; Standard List Matching Module: Input the basic attribute descriptions of each historical list into the large language model to obtain language feature vectors; Feature Extraction Module: Add feature labels to the historical lists to facilitate clustering; Clustering Analysis Module: Cluster and group according to the similarity of semantic feature vectors. This feature extraction module uses asynchronous non-blocking processing, ultimately grouping historical lists with high semantic feature vector similarity into one group.

[0031] Furthermore, step A4, which involves calculating the reasonable price range for each historical list group based on the bid prices of each historical list within that group, includes: Step A41: Statistically analyze the bid prices of each historical list in the historical list group using the assumption of normal distribution, and select bid prices that fall within two standard deviations to form an effective sample set; Step A42: Based on the effective sample set, calculate the reasonable price range for the historical list group; wherein the reasonable price range includes: reasonable minimum value, reasonable maximum value, reasonable average value, and median value.

[0032] In this embodiment, a normal distribution assumption is made for the price data in each historical list group. The standard deviation and mean of these price data are calculated, and the price data falling between (mean - twice the standard deviation) and (mean + twice the standard deviation) are used as the effective sample set. Multiple reference indicators, such as reasonable minimum value, reasonable maximum value, reasonable average value, and median value, are calculated from the effective sample set. The reasonable minimum value and reasonable maximum value form the two ends of the reasonable price range. The reasonable price range refers to the price range calculated from the historical prices in the historical list group using statistical algorithms and according to preset rules. For example, after removing outliers through normal distribution analysis, the range of values ​​determined by the minimum and maximum prices in the effective sample set constitutes the reasonable price range. Preferably, the minimum price in the reasonable price range is based on the cost price. Users can customize or have the system automatically generate dynamic warning thresholds based on these reference indicators, providing a quantitative and objective basis for judging the reasonableness of bid prices.

[0033] Specifically, step S104, which involves obtaining a reasonable price range associated with the target historical list group and using the reasonable price range to assess the reasonableness of bid prices in the list to be evaluated, includes: Step B1: Obtain the current price of the consumables corresponding to the list to be evaluated, and parse the consumable usage information from the list to be evaluated; In particular, in construction projects, consumables specifically refer to labor, materials, and machinery, including the amount and price of the required labor, materials, and machinery.

[0034] Step B2: Calculate the cost price based on the current price of the consumables and the usage information of the consumables; Step B3: Conduct a comprehensive reasonableness assessment based on the bid prices in the list to be evaluated, the reasonable price range, and the cost price.

[0035] In this embodiment, the bid price in each list to be evaluated consists of multiple parts, with the cost price of consumables accounting for a large proportion. Each consumable part corresponds to a unit price. The cost price of the corresponding consumable is obtained by multiplying the consumption quantity by the unit price. The cost prices of all consumables are then summed to obtain the bid price for the material item. The combination ratio and cost price of each consumable in the bid price are evaluated separately based on the reasonable price range in historical lists with the same corresponding feature tags. The reasonableness of the bid prices in the list to be evaluated is analyzed. If unreasonable, the reasons for the unreasonableness are analyzed in detail.

[0036] Specifically, before obtaining the list of target construction projects to be evaluated in step S101, the method further includes: Step C1: Based on the basic information of the historical building projects to which each historical list belongs, set corresponding feature tags for each historical list; wherein, the dimensions of the feature tags shall include at least one of the following: cost type, project location, quotation timestamp, and professional engineering category; Step C2: Store the feature tags of each historical list into the preset database.

[0037] In this embodiment, feature tags are added to the standardized historical lists in the historical list database. The construction project corresponding to each historical list is obtained, and the basic information of the construction project to which each historical list belongs is parsed. At least one feature tag is added to the historical list based on the basic information, and the list is stored in a preset database. The basic information of the construction project includes time information, location information, engineering specialty information, cost stage information, and structural information.

[0038] In addition, the method also includes: After obtaining the list of projects to be evaluated, feature tags are added based on the basic information of the construction projects to which the list belongs, in preparation for subsequent bid price analysis; the dimensions of the feature tags shall include at least one of the following: cost type, project location, bid timestamp, and professional engineering category.

[0039] Furthermore, the historical lists within a group are further clustered based on their feature labels. Clustering can be performed according to at least one of the following dimensions: region, year, cost price, and market price. For example, historical lists might show a cost price of 18 yuan and a market price of 23 yuan in Beijing in 2024; a cost price of 20 yuan in Beijing in 2025; and a cost price of 30 yuan and a market price of 35 yuan in Shanxi in 2024. When grouping by feature labels, the historical lists from Beijing and Shanxi in 2024 are grouped by year; the historical lists from Beijing in 2024 and 2025 are grouped by region; and the historical lists from Beijing in 2024 and 2025 are grouped by market price. Adding feature labels and performing clustering facilitates accurate analysis during subsequent bid price evaluation.

[0040] Specifically, step S104, which involves obtaining a reasonable price range associated with the target historical list group and using the reasonable price range to evaluate the reasonableness of bid prices in the list to be evaluated, further includes: Step D1: When the list to be evaluated is deemed unreasonable, feature tags are generated for the list to be evaluated based on the basic information of the target construction project. Step D2: Select matching historical lists from the target historical list group whose feature tags match the feature tags of the list to be evaluated; Step D3: Parse the pricing details used to form the bid price from the list to be evaluated, and parse the corresponding pricing details from each of the matching historical lists; Step D4: By comparing the pricing details of the list to be evaluated with the pricing details of each of the matching historical lists, an analysis of unreasonable pricing is performed.

[0041] In this embodiment, when the bid price in the list to be evaluated is not within the reasonable price range of the historical lists, the bid price of the list to be evaluated is analyzed and the reasons are found. Feature tags are added to the list to be evaluated, and historical lists with the same feature tags as the list to be evaluated are searched from the historical list group matched by semantic feature vectors. The bid prices of the found historical lists are then analyzed. For example, if the list to be evaluated has a cost price of 28 yuan and a market price of 36 yuan in Beijing in 2025, a precise match is made from the historical lists that have been clustered within the reasonable price range. Further analysis is performed on the bid prices of the list to be evaluated based on the historical lists of Beijing in 2025. The differences in costs of each part of the pricing are compared and analyzed, with detailed evaluations conducted on labor costs, material costs, and the usage fees of labor, materials, and machinery. The analysis mainly focuses on three levels: Price anomaly: Significant deviations between the quota application combination and conventional practices. Consumption anomaly: Deviations in the consumption of resources such as labor, materials, and machinery from quota standards or industry practices. Abnormal unit prices of labor, materials, and machinery: Deviations in the unit prices of the main materials or equipment used from the market prices of the same period. Horizontal market comparison: Current bids are statistically analyzed against historical winning bids to calculate percentile values ​​and assess market levels. Vertical cost comparison: Profit margins are analyzed by comparing current bids with average social costs (calculated based on current material prices and quotas). Bids in the pending evaluation list are compared sequentially with those in historical lists and will be evaluated as follows... Figure 3 The analysis reveals the reasons for any irrationality, and the results will automatically generate a diagnostic report with clear attribution labels, providing direct and in-depth insights for review and decision-making.

[0042] In this embodiment, the overall solution achieves the following effects: It aggregates historical enterprise data, industry market data, and document data to form a dynamic, massive cost database, breaking through the limitations of a single data source; Intelligent list matching: By parsing natural language descriptions through a large model, it achieves standardization and accurate mapping of list items, improving the efficiency and accuracy of cross-project data matching, solving matching problems caused by non-standard list descriptions, and fully leveraging the value of historical data; In-depth source analysis: Based on identifying unreasonable quotations, it decomposes pricing elements layer by layer, locating anomalies in detailed items such as personnel, materials, and machinery, providing quantitative basis for determining reasonableness.

[0043] Example 2 This invention provides a data processing device for bidding and tendering lists based on a large model, such as... Figure 4 As shown, the device specifically includes the following components: The first acquisition module 401 is used to acquire the list of target construction projects to be evaluated; wherein, the list of projects to be evaluated includes: basic attribute descriptions and bid prices; The first parsing module 402 is used to input the basic attribute description of the list to be evaluated into a pre-trained large language model, and obtain the semantic feature vector of the list to be evaluated output by the large language model. The semantic matching module 403 is used to compare the semantic feature vector of the list to be evaluated with the semantic feature vector of each historical list group in the preset database to determine the target historical list group that matches the list to be evaluated; wherein, the historical list group includes multiple historical lists with similar semantic feature vectors. The quotation evaluation module 404 is used to obtain a reasonable price range associated with the target historical list group, and to use the reasonable price range to evaluate the reasonableness of the bid quotations in the list to be evaluated; wherein, the reasonable price range is calculated based on the bid quotations of each historical list in the target historical list group.

[0044] Specifically, the device further includes: The second acquisition module is used to obtain a historical list of multiple historical building projects from multiple data sources; The second parsing module is used to input the basic attribute description of each historical list into the large language model, and obtain the semantic feature vector of the corresponding historical list output by the large language model. The clustering module is used to cluster all historical lists by calculating the similarity between the semantic feature vectors of each historical list, forming multiple historical list groups; The calculation module is used to calculate the reasonable price range for each historical list group based on the bid price of each historical list in each historical list group. The storage module is used to associate each historical list group with its corresponding reasonable price range and store it in the preset database.

[0045] Specifically, the computing module is used for: The bid prices of each historical list in the historical list group are statistically analyzed using the assumption of normal distribution, and bid prices falling within two standard deviations are selected to form an effective sample set; Based on the effective sample set, the reasonable price range of the historical list group is calculated; wherein, the reasonable price range includes: reasonable minimum value, reasonable maximum value, reasonable average value and median value.

[0046] Specifically, the quotation evaluation module 404 is used for: Obtain the current price of the consumables corresponding to the list to be evaluated, and parse the consumable usage information from the list to be evaluated; The cost price is calculated based on the current price of the consumables and the usage information of the consumables. A comprehensive reasonableness assessment is conducted based on the bid prices in the list to be evaluated, the reasonable price range, and the cost price.

[0047] Specifically, the device further includes a feature tag module, used for: Based on the basic information of the historical building projects to which each historical list belongs, corresponding feature tags are set for each historical list; wherein, the dimensions of the feature tags include at least one of the following: cost type, project location, quotation timestamp, and professional engineering category; The feature tags of each historical list are stored in the preset database.

[0048] Specifically, the quotation evaluation module 404 is also used for: When the list of items to be evaluated is deemed unreasonable, feature tags are generated for the list of items to be evaluated based on the basic information of the target construction project. Filter out matching historical lists whose feature tags match the feature tags of the list to be evaluated from the target historical list group; The pricing details used to form the bid price are parsed from the list to be evaluated, and the corresponding pricing details are parsed from each of the matching historical lists; An analysis of unreasonable pricing is conducted by comparing the pricing details of the list to be evaluated with the pricing details of each matching historical list.

[0049] Example 3 This embodiment also provides a computer device, such as a smartphone, tablet computer, laptop computer, desktop computer, rack server, blade server, tower server, or cabinet server (including a standalone server or a server cluster composed of multiple servers), etc., capable of executing programs. Figure 5 As shown, the computer device 50 in this embodiment includes, but is not limited to, a memory 501 and a processor 502 that are communicatively connected to each other via a system bus. It should be noted that... Figure 5 Only a computer device 50 with components 501-502 is shown; however, it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0050] In this embodiment, the memory 501 (i.e., the readable storage medium) includes flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 501 may be an internal storage unit of the computer device 50, such as the hard disk or memory of the computer device 50. In other embodiments, the memory 501 may also be an external storage device of the computer device 50, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 50. Of course, the memory 501 may include both the internal storage unit and the external storage device of the computer device 50. In this embodiment, the memory 501 is typically used to store the operating system and various application software installed on the computer device 50. In addition, the memory 501 may also be used to temporarily store various types of data that have been output or will be output.

[0051] In some embodiments, processor 502 may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. This processor 502 is typically used to control the overall operation of computer device 50.

[0052] Specifically, in this embodiment, the processor 502 is used to execute the program of the bidding list data processing method based on the large model stored in the memory 501. When the program of the bidding list data processing method based on the large model is executed, it performs the following steps: Obtain the list of target construction projects to be evaluated; wherein the list of projects to be evaluated includes: basic attribute descriptions and bid prices; The basic attribute description of the list to be evaluated is input into a pre-trained large language model, and the semantic feature vector of the list to be evaluated is obtained from the output of the large language model. The semantic feature vector of the list to be evaluated is compared with the semantic feature vector of each historical list group in the preset database to determine the target historical list group that matches the list to be evaluated; wherein, the historical list group includes multiple historical lists with similar semantic feature vectors. Obtain a reasonable price range associated with the target historical list group, and use the reasonable price range to evaluate the reasonableness of the bid prices in the list to be evaluated; wherein the reasonable price range is calculated based on the bid prices of each historical list in the target historical list group.

[0053] For a detailed description of the above method steps, please refer to Example 1. This example will not be repeated here.

[0054] Example 4 This embodiment also provides a computer-readable storage medium, such as flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, server, app store, etc., which stores a computer program. When the computer program is executed by a processor, it implements the following method steps: Obtain the list of target construction projects to be evaluated; wherein the list of projects to be evaluated includes: basic attribute descriptions and bid prices; The basic attribute description of the list to be evaluated is input into a pre-trained large language model, and the semantic feature vector of the list to be evaluated is obtained from the output of the large language model. The semantic feature vector of the list to be evaluated is compared with the semantic feature vector of each historical list group in the preset database to determine the target historical list group that matches the list to be evaluated; wherein, the historical list group includes multiple historical lists with similar semantic feature vectors. Obtain a reasonable price range associated with the target historical list group, and use the reasonable price range to evaluate the reasonableness of the bid prices in the list to be evaluated; wherein the reasonable price range is calculated based on the bid prices of each historical list in the target historical list group.

[0055] For a detailed description of the above method steps, please refer to the first embodiment. This embodiment will not repeat the details here.

[0056] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0057] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0058] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method.

[0059] The above are merely preferred embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for processing bidding list data based on a large model, characterized in that, The method includes: Obtain the list of target construction projects to be evaluated; wherein the list of projects to be evaluated includes: basic attribute descriptions and bid prices; The basic attribute description of the list to be evaluated is input into a pre-trained large language model, and the semantic feature vector of the list to be evaluated is obtained from the output of the large language model. The semantic feature vector of the list to be evaluated is compared with the semantic feature vector of each historical list group in the preset database to determine the target historical list group that matches the list to be evaluated; wherein, the historical list group includes multiple historical lists with similar semantic feature vectors. Obtain a reasonable price range associated with the target historical list group, and use the reasonable price range to evaluate the reasonableness of the bid prices in the list to be evaluated; wherein the reasonable price range is calculated based on the bid prices of each historical list in the target historical list group.

2. The method for processing bidding list data based on a large model according to claim 1, characterized in that, Prior to obtaining the list of target construction projects to be evaluated, the method further includes: Obtain historical lists of multiple historical building projects from multiple data sources; The basic attribute descriptions of each historical list are input into the large language model, and the semantic feature vectors of the corresponding historical lists output by the large language model are obtained. All historical lists are clustered by calculating the similarity between the semantic feature vectors of each historical list, forming multiple historical list groups; Based on the bid prices of each historical list in each historical list group, calculate the reasonable price range for the corresponding historical list group; Each historical list group and its corresponding reasonable price range are associated and stored in the preset database.

3. The method for processing bidding list data based on a large model according to claim 2, characterized in that, The calculation of the reasonable price range for each historical list group, based on the bid prices of each historical list within that group, includes: The bid prices of each historical list in the historical list group are statistically analyzed using the assumption of normal distribution, and bid prices falling within two standard deviations are selected to form an effective sample set; Based on the effective sample set, the reasonable price range of the historical list group is calculated; wherein, the reasonable price range includes: reasonable minimum value, reasonable maximum value, reasonable average value and median value.

4. The method for processing bidding list data based on a large model according to claim 3, characterized in that, The step of obtaining a reasonable price range associated with the target historical list group and using the reasonable price range to evaluate the reasonableness of bid prices in the list to be evaluated includes: Obtain the current price of the consumables corresponding to the list to be evaluated, and parse the consumable usage information from the list to be evaluated; The cost price is calculated based on the current price of the consumables and the usage information of the consumables. A comprehensive reasonableness assessment is conducted based on the bid prices in the list to be evaluated, the reasonable price range, and the cost price.

5. The method for processing bidding list data based on a large model according to claim 4, characterized in that, Prior to obtaining the list of target construction projects to be evaluated, the method further includes: Based on the basic information of the historical building projects to which each historical list belongs, corresponding feature tags are set for each historical list; wherein, the dimensions of the feature tags include at least one of the following: cost type, project location, quotation timestamp, and professional engineering category; The feature tags of each historical list are stored in the preset database.

6. The method for processing bidding list data based on a large model according to claim 5, characterized in that, The step of obtaining a reasonable price range associated with the target historical list group and using the reasonable price range to evaluate the reasonableness of bid prices in the list to be evaluated includes: When the list of items to be evaluated is deemed unreasonable, feature tags are generated for the list of items to be evaluated based on the basic information of the target construction project. Filter out matching historical lists whose feature tags match the feature tags of the list to be evaluated from the target historical list group; The pricing details used to form the bid price are parsed from the list to be evaluated, and the corresponding pricing details are parsed from each of the matching historical lists; An analysis of unreasonable pricing is conducted by comparing the pricing details of the list to be evaluated with the pricing details of each matching historical list.

7. A data processing device for bidding and tendering lists based on a large model, characterized in that, The device includes: The first acquisition module is used to acquire the list of target construction projects to be evaluated; wherein, the list of projects to be evaluated includes: basic attribute descriptions and bid prices; The first parsing module is used to input the basic attribute description of the list to be evaluated into a pre-trained large language model and obtain the semantic feature vector of the list to be evaluated output by the large language model. The semantic matching module is used to compare the semantic feature vector of the list to be evaluated with the semantic feature vector of each historical list group in the preset database to determine the target historical list group that matches the list to be evaluated; wherein, the historical list group includes multiple historical lists with similar semantic feature vectors. The quotation evaluation module is used to obtain a reasonable price range associated with the target historical list group, and to use the reasonable price range to evaluate the reasonableness of the bid quotations in the list to be evaluated; wherein, the reasonable price range is calculated based on the bid quotations of each historical list in the target historical list group.

8. The bidding list data processing device based on a large model according to claim 7, characterized in that, The device further includes: The second acquisition module is used to obtain a historical list of multiple historical building projects from multiple data sources; The second parsing module is used to input the basic attribute description of each historical list into the large language model, and obtain the semantic feature vector of the corresponding historical list output by the large language model. The clustering module is used to cluster all historical lists by calculating the similarity between the semantic feature vectors of each historical list, forming multiple historical list groups; The calculation module is used to calculate the reasonable price range for each historical list group based on the bid price of each historical list in each historical list group. The storage module is used to associate each historical list group with its corresponding reasonable price range and store it in the preset database.

9. A computer device, the computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the steps of the method according to any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images