CN117495298A - Bid bid review supervision method and system - Google Patents
Bid bid review supervision method and system Download PDFInfo
- Publication number
- CN117495298A CN117495298A CN202311535768.4A CN202311535768A CN117495298A CN 117495298 A CN117495298 A CN 117495298A CN 202311535768 A CN202311535768 A CN 202311535768A CN 117495298 A CN117495298 A CN 117495298A
- Authority
- CN
- China
- Prior art keywords
- bid
- bidding
- bidder
- expressed
- evaluation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 81
- 238000012552 review Methods 0.000 title claims description 43
- 238000011156 evaluation Methods 0.000 claims abstract description 102
- 230000008569 process Effects 0.000 claims abstract description 36
- 230000008901 benefit Effects 0.000 claims abstract description 20
- 230000006399 behavior Effects 0.000 claims abstract description 14
- 238000000691 measurement method Methods 0.000 claims abstract description 11
- 238000013210 evaluation model Methods 0.000 claims abstract description 10
- 239000013598 vector Substances 0.000 claims description 56
- 238000012797 qualification Methods 0.000 claims description 44
- 238000012216 screening Methods 0.000 claims description 26
- 230000008520 organization Effects 0.000 claims description 24
- 239000002131 composite material Substances 0.000 claims description 12
- 210000002569 neuron Anatomy 0.000 claims description 12
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000007689 inspection Methods 0.000 claims description 6
- 239000000463 material Substances 0.000 claims description 6
- 230000005856 abnormality Effects 0.000 claims description 5
- GOLXNESZZPUPJE-UHFFFAOYSA-N spiromesifen Chemical compound CC1=CC(C)=CC(C)=C1C(C(O1)=O)=C(OC(=O)CC(C)(C)C)C11CCCC1 GOLXNESZZPUPJE-UHFFFAOYSA-N 0.000 claims description 5
- 238000010276 construction Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 3
- 230000003872 anastomosis Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000002474 experimental method Methods 0.000 description 5
- 238000012854 evaluation process Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 208000001613 Gambling Diseases 0.000 description 1
- 206010033307 Overweight Diseases 0.000 description 1
- 102100040160 Rabankyrin-5 Human genes 0.000 description 1
- 101710086049 Rabankyrin-5 Proteins 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002457 bidirectional effect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000000275 quality assurance Methods 0.000 description 1
- 238000011524 similarity measure Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/08—Auctions
Landscapes
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Engineering & Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of bid evaluation supervision, in particular to a bid evaluation supervision method and system. Firstly, the invention provides a bidding anomaly analysis model for analyzing bidding parties with string bidding behaviors; the analysis of the string bidding behavior mainly adopts a bidding measurement method to measure the similarity of the collected bidding, and the method can accurately compare the similarity of the bidding, so that the fairness in the bidding process is improved; secondly, the invention provides a bid evaluation model which consists of an expert scoring module and a machine scoring module; the expert scoring module establishes an evaluation scheme through weighting evaluation, optimal value and advantage selection, so that fairness of bidder evaluation is guaranteed; the machine scoring module is used for assisting an expert in evaluating to avoid subjective factors, and performing estimated scoring on a plurality of relevant factors of a bidder, so that the evaluating is more reasonable.
Description
Technical Field
The invention relates to the technical field of bid evaluation supervision, in particular to a bid evaluation supervision method and system.
Background
Bidding is a common method of purchasing widely used in government institutions, businesses and organizations to purchase products, services or engineering items. It is a public, transparent purchasing process aimed at ensuring the realization of fair competition and optimal value.
The qualification and compliance of each bid is often assessed by the bid party organization review board during the bidding process. The main evaluation factors comprise technical evaluation, financial evaluation, qualification check and the like. Scoring decisions are made by the review board members during the evaluation process, and unfair evaluation conditions are likely to occur to bidders due to the strong subjectivity of the easy doping during the process. Furthermore, the whole bidding process is not transparent due to the participation of human factors, and the phenomenon of manipulating the bidding result is easy to occur.
In order to reduce the phenomenon of unfair assessment caused by human factors in the bidding process, the invention provides a bidding review supervision method and system.
Disclosure of Invention
A bid and bid review supervision method and system are used for solving the technical problems. The method and the device send out from the fairness of bidding, compare the similarity of the bidding documents, reject the bidding documents with high similarity, avoid the situation of stringing the bidding among bidders, and ensure the fairness in the bidding process; in order to minimize unfair phenomenon caused by human factors in the bid evaluation process, the invention provides a bid evaluation model; the bid evaluation model consists of an expert scoring module and a machine scoring module, and the comprehensive score of the bidder is obtained through bidirectional scoring, so that certain fairness is ensured.
In order to achieve the above purpose, the present invention provides the following technical solutions:
a bid review supervision method comprising:
acquiring tender books and information data of a plurality of bidders, and respectively constructing a tender book set and an information set; wherein the bidding book set includes a plurality of bidding documents; the information set includes: bidder performance, project organization plans, project management capabilities, and bid quotes;
wherein the bidder performance includes enterprise grade, reputation and credit, similar engineering experience, near five years of quality records and near three years of major security incidents; the project organization plan comprises an engineering project scheme, an address organization structure, a project network plan table and a quality insurance system; the project management capabilities include primary equipment and labor schemes, safety precautions, and delivered project quality; the bid offers include engineering offers, manual in bid price, machine rationality, and prime material cost rationality.
Pre-processing the information before constructing the information set; wherein the preprocessing includes normalizing the information data; wherein, the standardized formula is:
wherein N () is the normalization formula; z is Z i,j A j-th information index expressed as an i-th bidder; z is Z j,B An optimal value expressed as a bidder jth information index; z is Z i,w Expressed as the worst value of the bidder jth information indicator.
Further, in order to avoid ticket stringing among bidders, searching the information set of the tender book again;
constructing a bidding anomaly analysis model to screen the bidding book and reserving the bidders meeting the requirements; the specific process of the bidding anomaly analysis model is as follows:
extracting features of the first bidding information and the second bidding from the bidding set by using the bidding anomaly analysis model to obtain a first bidding feature vector and a second bidding feature vector;
calculating similarity values of the first bidding feature vector and the second bidding feature vector by adopting a bidding measurement method; the calculation formula of the similarity value is as follows:
wherein S () is represented as a function of the bid metric method; d, d 1 Expressed as the first bid feature vector; d, d 2 Represented as the second bid feature vector; d, d 1j Expressed as j features in the first bidding feature vector; d, d 2j A j-th feature represented as the second bidding feature vector; n (N) * () Expressed as a feature variability function of the bid; n (N) ∪ () Represented as a feature addition function of a bid amount；
Wherein the N is ∪ () The calculation formula of (2) is as follows:
wherein the N is * () The calculation formula of (2) is as follows:
wherein λ is represented as a constant value; sigma (sigma) j A hyper-parameter denoted as the j-th feature; n (N) ∩ () Expressed as a feature phase function of the bid;multiplying the feature of the bid to obtain a function;
wherein the N is ∩ () The calculation formula of (2) is as follows:
wherein the saidThe calculation formula of (2) is as follows:
sorting a plurality of bidding information in a descending order according to the similarity, grading, and eliminating the bidders with string bidding behaviors to obtain candidate bidding sets;
wherein the rating is to consider a highly similar bid to have a string behavior, and the rating criteria include-1, 0 and 1; wherein, -1 is represented as the bidder string; 0 is represented as the bid Fang Yishi string; 1 indicates that the bidder is normal.
Further, inputting the candidate bid set and the information set into a bid evaluation model; wherein the bid evaluation model comprises an expert scoring module and a machine scoring module;
the expert scoring module obtains the score of the optimal winning candidate by adjusting the comprehensive combination of the weighted evaluation, the optimal value and the advantage selection of multiple factors; the expert scoring module comprises the following specific processes:
Randomly extracting n experts from an expert database; wherein n is an integer greater than 5; the expert database consists of personnel of a bidding department, an discipline inspection and supervision department, a project charging department and a bidding agency;
selecting an optimal bidder according to the weighted evaluation, the optimal value and the advantage;
wherein the weighted items of the weighted evaluation include bidder performance, project organization plan and project management capability, respectively denoted as P r1 、P r2 And P r3 R is more than or equal to 1 and less than or equal to n, r is expressed as the number of bidders, and n is expressed as the number of bidders; the best value is expressed as a bid price, recorded as TP r The method comprises the steps of carrying out a first treatment on the surface of the The advantage selection includes the expert formulating a bid by varying the evaluation factors, noted as ITD k K is equal to or more than 1 and equal to or less than m, k is the number of the evaluation factors, and m is the number of the evaluation factors;
calculating a composite qualification score for the weighted evaluation of bidders; the calculation mode of the comprehensive qualification score is as follows:
wherein CQS r The composite qualification score expressed as bidder r; w (w) i A weight denoted as the i-th said weighted term;
the bidder is ordered in descending order according to the qualification comprehensive score, and a qualification ordering list is obtained;
The expert reviews the bid according to the evaluation factors and obtains a bid evaluation error value; the calculation formula of the average error value of the bid evaluation is as follows:
wherein ES r The tender review average error value expressed as the bidder r; mu (mu) k The influence coefficient of the bidder r on the kth evaluation factor is expressed; TDV (time domain reflectometer) r A review score representing the tender book of the bidder r; ITDV (International traffic light vector) k Represented as the ITD k A set score for (2); EP (EP) r Representing additional scores obtained by the bidder r according to the qualification ordering list;
and scoring the bidder according to the average error value of the bid evaluation, and outputting a bidder scoring list.
The machine scoring module inputs the candidate bidder set and the information set for scoring;
the specific process of the machine scoring module comprises the following steps:
initializing the weight of the machine scoring module;
the weight is a bidding factor set by a bidding party; the weight adopts w 0 ＝{w 1 ,...,w n -representation;
inputting information data of bidder after preprocessing, namely x i The method comprises the steps of carrying out a first treatment on the surface of the Wherein said x i For the information data, i is an arbitrary natural number of 1 to n;
calculating a value obtained by the information data through shallow neurons by the machine scoring module; wherein, the calculation formula of the value is:
Wherein Z is j A value calculated for the j-th said shallow neuron; f () is a calculation function of the machine scoring module;represented as w at initialization i Is the j-th weight value of (2);
after forward propagation for many times, an output layer outputs scores; the score calculation formula is:
wherein Y is represented as the score;a weight value representing n-1 layers; />Expressed as n layers of output values;
updating the weight value when each forward propagation occurs; the updating formula of the weight value is as follows:
wherein,representing the weight value updated by n times of calculation; η is expressed as a bidding factor gain factor of the bidder; sigma (sigma) ij Expressed as the difference between the bidder and the i-th bidder's expected value for j factors;
wherein the sigma ij The calculation formula of (2) is as follows:
wherein y is ij A score for the ith factor for the ith said bidder;expressed as a desired score by the bidding party for the ith bidding party with respect to the jth factor.
Comprehensively calculating the output of the expert scoring module and the output of the machine scoring module to obtain the bid marking numbers of a plurality of bidding parties;
wherein, the calculation formula of the index number is as follows:
S bidder ＝ξ 0 *ER+ξ 1 *MR；
wherein S is bidder Expressed as the number of marks; ER represents the score for the expert scoring module; scoring of the machine scoring module by MR; zeta type toy 0 A degree of influence expressed as the expert scoring module; zeta type toy 1 A degree of influence expressed as the machine scoring module; and xi 0 +ξ 1 ＝1，ξ 0 ＞ξ 1 。
A bid review supervision system comprising:
the bid price information system comprises a bid price file issuing unit, a bidder information acquiring unit, a bid price storage unit, a bid screening unit, a bid evaluation unit and a score display unit;
the bid-bidding-target-file distribution unit is used for acquiring bid-target files issued by a bid-target party and extracting the advertising content of the bid-target-file;
the bidder information acquisition unit is used for recording information data related to the bidder; wherein the information data includes bidding party performance, project organization plans, project management capabilities, and bid offers;
wherein the bidder performance includes enterprise grade, reputation and credit, similar engineering experience, near five years of quality records and near three years of major security incidents; the project organization plan comprises an engineering project scheme, an address organization structure, a project network plan table and a quality insurance system; the project management capabilities include primary equipment and labor schemes, safety precautions, and delivered project quality; the bid offers include engineering offers, manual in bid price, machine rationality, and prime material cost rationality.
The tender book storage unit is used for safely storing tender books submitted by bidders, and the tender books are stored on a blockchain after being divided into a plurality of data blocks for encryption;
the bid screening unit is used for screening unqualified bidders and bid books with problems; the bidding screening unit comprises bidder qualification screening and tender book searching and rescreening screening;
the bidding party qualification screening sets preset conditions through the system, and screens personnel participating in bidding according to the data of the bidding party information acquisition unit;
the bid checking and re-screening of the bid is carried out by adopting a bid abnormality analysis model to analyze the bid; the specific process of the bidding anomaly analysis model is as follows:
extracting features of the first bidding information and the second bidding from the bidding set by using the bidding anomaly analysis model to obtain a first bidding feature vector and a second bidding feature vector;
calculating similarity values of the first bidding feature vector and the second bidding feature vector by adopting a bidding measurement method; the calculation formula of the similarity value is as follows:
wherein S () is represented as a function of the bid metric method; d, d 1 Expressed as the first bid feature vector; d, d 2 Represented as the second bid feature vector; d, d 1j Expressed as j features in the first bidding feature vector; d, d 2j A j-th feature represented as the second bidding feature vector; n (N) * () Expressed as a feature variability function of the bid; n (N) ∪ () Represented as a feature addition function of the bid;
wherein the N is ∪ () The calculation formula of (2) is as follows:
wherein the N is * () The calculation formula of (2) is as follows:
wherein λ is represented as a constant value; sigma (sigma) j A hyper-parameter denoted as the j-th feature; n (N) ∩ () Expressed as a feature phase function of the bid;multiplying the feature of the bid to obtain a function;
wherein the N is ∩ () The calculation formula of (2) is as follows:
wherein the saidThe calculation formula of (2) is as follows:
and sorting the bidding information in a descending order according to the similarity, grading, and eliminating the bidder with string bidding behavior.
The bid evaluation unit is used for bid winning evaluation of the bidder meeting the requirements; the bid evaluation unit comprises an expert scoring module and a machine scoring module;
the expert scoring module obtains the score of the optimal winning candidate by adjusting the comprehensive combination of the weighted evaluation, the optimal value and the advantage selection of multiple factors; the expert scoring module comprises the following specific processes:
Randomly extracting n experts from an expert database; wherein n is an integer greater than 5; the expert database consists of personnel of a bidding department, an discipline inspection and supervision department, a project charging department and a bidding agency;
selecting an optimal bidder according to the weighted evaluation, the optimal value and the advantage;
wherein the weighted items of the weighted evaluation include bidder performance, project organization plan and project management capability, respectively denoted as P r1 、P r2 And P r3 R is more than or equal to 1 and less than or equal to n, r is expressed as the number of bidders, and n is expressed as the number of bidders; the best value is expressed as a bid price, recorded as TP r The method comprises the steps of carrying out a first treatment on the surface of the The advantage selection includes the expert formulating a bid by varying the evaluation factors, noted as ITD k K is equal to or more than 1 and equal to or less than m, k is the number of the evaluation factors, and m is the number of the evaluation factors;
calculating a composite qualification score for the weighted evaluation of bidders; the calculation mode of the comprehensive qualification score is as follows:
wherein CQS r The composite qualification score expressed as bidder r; w (w) i A weight denoted as the i-th said weighted term;
the bidder is ordered in descending order according to the qualification comprehensive score, and a qualification ordering list is obtained;
The expert reviews the bid according to the evaluation factors and obtains a bid evaluation error value; the calculation formula of the average error value of the bid evaluation is as follows:
wherein ES r The tender review average error value expressed as the bidder r; mu (mu) k The influence coefficient of the bidder r on the kth evaluation factor is expressed; TDV (time domain reflectometer) r Representing the bidder rA review score for the bid; ITDV (International traffic light vector) k Represented as the ITD k A set score for (2); EP (EP) r Representing additional scores obtained by the bidder r according to the qualification ordering list;
and scoring the bidder according to the average error value of the bid evaluation, and outputting a bidder scoring list.
The machine scoring module inputs the candidate bidder set and the information set for scoring;
the specific process of the machine scoring module comprises the following steps:
initializing the weight of the machine scoring module;
the weight is a bidding factor set by a bidding party; the weight adopts w 0 ＝{w 1 ,...,w n -representation;
inputting information data of bidder after preprocessing, namely x i The method comprises the steps of carrying out a first treatment on the surface of the Wherein said x i For the information data, i is an arbitrary natural number of 1 to n;
calculating a value obtained by the information data through shallow neurons by the machine scoring module; wherein, the calculation formula of the value is:
Wherein Z is j A value calculated for the j-th said shallow neuron; f () is a calculation function of the machine scoring module;represented as w at initialization i Is the j-th weight value of (2);
after forward propagation for many times, an output layer outputs scores; the score calculation formula is:
wherein Y is represented as the score;a weight value representing n-1 layers; />Expressed as n layers of output values;
updating the weight value when each forward propagation occurs; the updating formula of the weight value is as follows:
wherein,representing the weight value updated by n times of calculation; η is expressed as a bidding factor gain factor of the bidder; sigma (sigma) ij Expressed as the difference between the bidder and the i-th bidder's expected value for j factors;
wherein the sigma ij The calculation formula of (2) is as follows:
wherein y is ij A score for the ith factor for the ith said bidder;expressed as a desired score by the bidding party for the ith bidding party with respect to the jth factor.
Comprehensively calculating the output of the expert scoring module and the output of the machine scoring module to obtain the bid marking numbers of a plurality of bidding parties;
wherein, the calculation formula of the index number is as follows:
S bidder ＝ξ 0 *ER+ξ 1 *MR；
wherein S is bidder Indicated as the winning bid A score; ER represents the score for the expert scoring module; scoring of the machine scoring module by MR; zeta type toy 0 A degree of influence expressed as the expert scoring module; zeta type toy 1 A degree of influence expressed as the machine scoring module; and xi 0 +ξ 1 ＝1，ξ 0 ＞ξ 1 。
And the scoring display unit is used for displaying scoring results of all the bidders and performing descending order arrangement.
Compared with the prior art, the invention has the beneficial effects that:
1. the invention provides a bidding anomaly analysis model which is used for analyzing bidding parties with string bidding behaviors; the analysis of the string bidding behavior mainly adopts a bid amount measurement method to measure the similarity of the collected bid amounts, and the method considers three conditions for the vectors of the two bid amounts: (1) the features considered appear in both of the markups; (2) the features considered appear in only one of the bid documents; (3) the features considered are not present in any document. For the first case, a lower bound is given and the similarity is reduced based on the difference between the feature values of the two documents. For the second case, a fixed value is given regardless of the magnitude of the characteristic value; the similarity of the tender books can be accurately compared by the method, and fairness in the tender process is improved.
2. According to the invention, the expert scoring module is provided on a traditional expert evaluating mechanism, and generates the evaluating scheme through weighting evaluation, optimal value and advantage selection and through combination of various factors, so that the phenomenon of string bidding with bidders in the evaluating process can be effectively avoided, bidders can be better selected through different winning schemes, and the evaluating scheme has fairness to the bidders.
3. The invention provides a machine scoring module which is used for assisting an expert in evaluating a bid. The machine scoring module performs evaluation and calculation according to the information data of the bidding party, weight parameters are given to the evaluation factors in the process of calculation, and in order to evaluate more weight parameters, the evaluation of the suitability of the bidding party can be better calculated by updating the bidding factor gain coefficient of the bidding party and the expected difference of the bidding party; in addition, the machine scoring module reduces errors in scoring caused by subjective factors of experts.
Drawings
FIG. 1 is a flow chart of a bid and ask review supervision method provided by an embodiment of the invention;
FIG. 2 is a schematic diagram of a scenario for bid review supervision provided by an embodiment of the present invention;
FIG. 3 is a schematic diagram of an index of a main index of an information set according to an embodiment of the present invention;
FIG. 4 is a diagram of a comparison experiment of a method for measuring a bid amount according to an embodiment of the present invention;
FIG. 5 is a graph showing a comparison of performance of different bidding schemes of an expert scoring module according to an embodiment of the present invention;
FIG. 6 is an analysis chart of the optimal winning bidder provided by the embodiment of the invention;
FIG. 7 is a diagram of a bid review supervision system provided by an embodiment of the invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
FIG. 1 is a flow chart of a bid review supervision method according to the present invention, and the following embodiments will be described in detail about the steps in the flow chart. First, referring to fig. 1, the specific steps include: s10, acquiring bidding party data; s20, screening the bidding parties; s30, evaluating the screened bidder; s40, obtaining an evaluation score list and determining a winning bidder. A specific description of the embodiments will be given below.
Example 1:
providing a scenario diagram of the present invention implemented in a specific bidding process in embodiment 1, referring to fig. 2, the fig. 2 shows an actual scenario in the bidding process;
in the bidding process, various information data of bidders need to be collected; in the step S10, as shown in the figure 1, bidder data are acquired;
the bidding party data comprises submitted tender books and information data related to the bidding party, and a bidding book set and an information set are respectively constructed for unified management;
in the tender book, the bidder needs to specify the contents such as own technical scheme, construction organization design, engineering quantity list, quotation and the like, and promises to fulfill a contract according to the contents of the tender book after winning a bid.
Further, in order to select a more appropriate bidder, additional information data needs to be added for auxiliary examination; the information set comprises four basic items, namely: bidder performance, project organization plans, project management capabilities, and bid quotes;
wherein the bidder performance includes enterprise grade, reputation and credit, similar engineering experience, near five years of quality records and near three years of major security incidents; the project organization plan comprises an engineering project scheme, an address organization structure, a project network plan table and a quality insurance system; the project management capabilities include primary equipment and labor schemes, safety precautions, and delivered project quality; the bid offers include engineering offers, manual unit price in bidding, rationality of construction machinery, and rationality of major material costs. The information-centric index case is expanded in fig. 3;
Further, carrying out standardization processing on the information data;
wherein, the standardized formula is:
wherein N () is the normalization formula; z is Z i,j A j-th information index expressed as an i-th bidder; z is Z j,B An optimal value expressed as a bidder jth information index; z is Z i,w Expressed as the worst value of the bidder jth information indicator.
Further, the bidding behavior among bidders is avoided, and similar comparison is required for the delivered tender books; screening the bidder at S20; the S20 adopts a bidding anomaly analysis model to screen the bidding book; the specific process of the bidding anomaly analysis model is as follows:
extracting features of the first bidding information and the second bidding from the bidding set by using the bidding anomaly analysis model to obtain a first bidding feature vector and a second bidding feature vector;
calculating similarity values of the first bidding feature vector and the second bidding feature vector by adopting a bidding measurement method; the calculation formula of the similarity value is as follows:
wherein S () is represented as a function of the bid metric method; d, d 1 Expressed as the first bid feature vector; d, d 2 Represented as the second bid feature vector; d, d 1j Expressed as j features in the first bidding feature vector; d, d 2j A j-th feature represented as the second bidding feature vector; n (N) * () Expressed as a feature variability function of the bid; n (N) ∪ () Represented as a feature addition function of the bid;
wherein the N is ∪ () The calculation formula of (2) is as follows:
wherein the N is * () The calculation formula of (2) is as follows:
wherein λ is represented as a constant value; sigma (sigma) j A hyper-parameter denoted as the j-th feature; n (N) ∩ () Expressed as a feature phase function of the bid;multiplying the feature of the bid to obtain a function;
wherein the N is ∩ () The calculation formula of (2) is as follows:
wherein the saidThe calculation formula of (2) is as follows:
sorting the bidding information in a descending order according to the similarity, and grading; wherein the rating comprises-1, 0, and 1; wherein, -1 is represented as the bidder string; 0 is represented as the bid Fang Yishi string; 1 indicates that the bidder is normal; and eliminating the bidder with the string bidding behavior to obtain a candidate bidding set.
In this embodiment, the bidding anomaly analysis model is experimentally described to demonstrate the applicability of the model to anomaly analysis. The predicted results of the behavior of strings considered to exist over the internet for the last year because of the high similarity of the tender books are given in table 1.
TABLE 1 analysis of historical bidding records for Bid anomaly analysis model
| Bidding party | Experimental data | Experimental results | Real data | True results | Comparison of results |
| Bidding party 1 | 82.5；0 | Promotion of | 89.00；1 | Promotion of | Anastomosis of |
| Bidding party 2 | 56.75；1 | Promotion of | 75.34；1 | Promotion of | Anastomosis of |
| Bidding party 3 | 93.58；-1 | Obsolete | 92.23；1 | Obsolete | Anastomosis of |
| Bidding party 4 | 42.36；1 | Promotion of | 46.78；1 | Promotion of | Anastomosis of |
| Bidding party 5 | 94.18；-1 | Obsolete | 93.64；1 | Obsolete | Anastomosis of |
The bidding conditions of the 5-bit bidder history are listed in table 1, and the results of the analysis of the bidding anomaly analysis model are consistent with the actual determination results, and the validity of the model can be seen from the table.
Further, the bid abnormality analysis model is mainly added with the bid measurement method to calculate the similarity degree of a bid; the method considers three cases for vectors of two bid amounts, the proposed metric: (1) The features considered appear in both of the markups, (2) the features considered appear in only one of the markups; (3) the features considered do not appear in any document; thus, the two parties of the string label can be prevented from escaping from the similarity investigation by modifying the content of the bid.
Experiments are presented in the step to illustrate that the effect of the bid amount measurement method is better than that of a comparison text similarity comparison method. Furthermore, referring to fig. 4, three alternatives are given as value choices for the constant parameter λ in the bid metric method, where λ is equal to m, 1 and AL, respectively; wherein m is expressed as a feature quantity; AL represents the average training length of the markup book; there is a clear effect in said fig. 4 compared to other text similarity measures.
In order to intuitively understand the implementation process of the bid amount measurement method, the embodiment also includes other description schemes; the calculation of the method is illustrated as follows. The calculation process comprises the following steps:
given bid book d 1 And bid book d 2 Respectively denoted as d 1 ＝＜3,2,0,1,3＞，d 2 = < 2,4,1,1,0 >, where λ=1, feature quantity m=5;
the similarity value of the bid d1 and the bid d2 is as follows:
/>
example 2:
the evaluation process of the bidder will be described in detail in the embodiments of the present application; evaluating the screened bidder corresponding to the S30; the bid evaluation model is adopted in the evaluation method; wherein the bid evaluation model comprises an expert scoring module and a machine scoring module; the expert scoring module obtains the score of the optimal winning candidate by adjusting the comprehensive combination of the weighted evaluation, the optimal value and the advantage selection of multiple factors; the expert scoring module comprises the following specific processes:
randomly extracting n experts from an expert database; wherein n is an integer greater than 5; the expert database consists of personnel of a bidding department, an discipline inspection and supervision department, a project charging department and a bidding agency;
Selecting an optimal bidder according to the weighted evaluation, the optimal value and the advantage;
wherein the weighted items of the weighted evaluation include bidder performance, project organization plan and project management capability, respectively denoted as P r1 、P r2 And P r3 R is more than or equal to 1 and less than or equal to n, r is expressed as the number of bidders, and n is expressed as the number of bidders; the best value is expressed as a bid price, recorded as TP r The method comprises the steps of carrying out a first treatment on the surface of the The advantage selection includes the expert formulating a bid by varying the evaluation factors, noted as ITD k K is equal to or more than 1 and equal to or less than m, k is expressed as the number of the evaluation factor, and m is expressed asThe number of said evaluation factors;
calculating a composite qualification score for the weighted evaluation of bidders; the calculation mode of the comprehensive qualification score is as follows:
wherein CQS r The composite qualification score expressed as bidder r; w (w) i A weight denoted as the i-th said weighted term;
the bidder is ordered in descending order according to the qualification comprehensive score, and a qualification ordering list is obtained;
the expert reviews the bid according to the evaluation factors and obtains a bid evaluation error value; the calculation formula of the average error value of the bid evaluation is as follows:
wherein ES r The tender review average error value expressed as the bidder r; mu (mu) k The influence coefficient of the bidder r on the kth evaluation factor is expressed; TDV (time domain reflectometer) r A review score representing the tender book of the bidder r; ITDV (International traffic light vector) k Represented as the ITD k A set score for (2); EP (EP) r Representing additional scores obtained by the bidder r according to the qualification ordering list;
and scoring the bidder according to the average error value of the bid evaluation, and outputting a bidder scoring list.
The specific process of the expert scoring module in the bid evaluation model is described in detail; the expert scoring module is designed to avoid the gambling behavior of the bidder, and to make up for the performance deficiency in the technical or other aspects by increasing the bid price, so that the decision on bidding is unreliable according to a single high-weight factor. The data is presented in FIG. 5 to illustrate the evaluation of bidders for the weighted evaluation, the optimal value, and the advantage selection;
in addition, in order to illustrate the role of the expert scoring module in the scoring process, the present embodiment is illustrated by specific data. Referring to the contents in Table 2, the comprehensive ranking conditions of 5 bidders obtained by the transformation of the main factors are given in the table; wherein the main factors are 15 indexes mentioned in example 1, respectively: enterprise grade, reputation and credit, similar engineering experience, recent five years of quality records, recent three years of major security incidents, engineering project plans, address organization structures, project network schedules, quality assurance systems, major equipment and labor plans, security precautions, delivered project quality, engineering quotes, manual unit price in bidding, rationality of construction machinery, and rationality of major material costs; the indicators are represented in the examples using x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13, x14 and x15, respectively.
TABLE 2 expert scoring Module experiment
| Bidding party 1 | Bidding party 2 | Bidding party 3 | Bidding party 4 | Bidding party 5 | |
| x1 | 0.85 | 0.87 | 0.90 | 0.80 | 0.83 |
| x2 | 0.84 | 0.83 | 0.85 | 0.82 | 0.83 |
| x3 | 0.88 | 0.88 | 0.87 | 0.81 | 0.82 |
| x4 | 0 | 1 | 0 | 1 | 2 |
| x5 | 0 | 0 | 0 | 0 | 0 |
| x6 | 0.88 | 0.90 | 0.89 | 0.82 | 0.83 |
| x7 | 0.87 | 0.85 | 0.88 | 0.81 | 0.85 |
| x8 | 0.84 | 0.82 | 0.83 | 0.79 | 0.81 |
| x9 | 0.87 | 0.86 | 0.91 | 0.80 | 0.82 |
| x10 | 0.89 | 0.88 | 0.88 | 0.83 | 0.84 |
| x11 | 0.86 | 0.87 | 0.89 | 0.81 | 0.83 |
| x12 | 0.91 | 0.89 | 0.93 | 0.82 | 0.81 |
| x13 | 0.90 | 0.90 | 0.92 | 0.85 | 0.89 |
| x14 | 0.91 | 0.92 | 0.90 | 0.84 | 0.86 |
| x15 | 0.85 | 0.84 | 0.89 | 0.80 | 0.79 |
| Ordering of | 2 | 3 | 1 | 5 | 4 |
Example 3:
the specific process of the machine scoring module comprises the following steps:
initializing the weight of the machine scoring module;
the weight is a bidding factor set by a bidding party; the weight adopts w 0 ＝{w 1 ,...,w n -representation; b= {0};
inputting information data of bidder after preprocessing, namely x i The method comprises the steps of carrying out a first treatment on the surface of the Wherein said x i For the information data, i is an arbitrary natural number of 1 to n;
calculating a value obtained by the information data through shallow neurons by the machine scoring module; wherein, the calculation formula of the value is:
wherein Z is j A value calculated for the j-th said shallow neuron; f () is a calculation function of the machine scoring module;represented as w at initialization i Is the j-th weight value of (2);
after forward propagation for many times, an output layer outputs scores; the score calculation formula is:
wherein Y is represented as the score;a weight value representing n-1 layers; />Expressed as n layers of output values;
updating the weight value when each forward propagation occurs; the updating formula of the weight value is as follows:
wherein,representing the weight value updated by n times of calculation; η is expressed as a bidding factor gain factor of the bidder; sigma (sigma) ij Expressed as the difference between the bidder and the i-th bidder's expected value for j factors;
wherein the sigma ij The calculation formula of (2) is as follows:
wherein y is ij A score for the ith factor for the ith said bidder;expressed as a desired score by the bidding party for the ith bidding party with respect to the jth factor.
In this embodiment, specific data is given to the machine scoring module to illustrate the ranking of the bidders for experiments, see table 3; table 3 gives the rank 5 as evaluated by the machine module for the bidder.
TABLE 3 machine scoring Module experiment
| Bidding party 1 | Bidding party 2 | Bidding party 3 | Bidding party 4 | Bidding party 5 | |
| x1 | 0.83 | 0.84 | 0.87 | 0.92 | 0.91 |
| x2 | 0.79 | 0.81 | 0.83 | 0.87 | 0.86 |
| x3 | 0.71 | 0.79 | 0.80 | 0.81 | 0.82 |
| x4 | 2 | 1 | 1 | 0 | 0 |
| x5 | 1 | 0 | 1 | 0 | 0 |
| x6 | 0.74 | 0.76 | 0.81 | 0.88 | 0.86 |
| x7 | 0.73 | 0.72 | 0.83 | 0.86 | 0.82 |
| x8 | 0.76 | 0.77 | 0.80 | 0.89 | 0.83 |
| x9 | 0.82 | 0.80 | 0.84 | 0.90 | 0.88 |
| x10 | 0.80 | 0.82 | 0.83 | 0.87 | 0.88 |
| x11 | 0.81 | 0.83 | 0.84 | 0.88 | 0.90 |
| x12 | 0.87 | 0.84 | 0.84 | 0.91 | 0.85 |
| x13 | 0.83 | 0.86 | 0.88 | 0.90 | 0.89 |
| x14 | 0.79 | 0.80 | 0.82 | 0.84 | 0.83 |
| x15 | 0.81 | 0.80 | 0.81 | 0.86 | 0.86 |
| Ordering of | 5 | 4 | 3 | 1 | 2 |
The expert scoring module and the machine scoring module give data in example 2 and example 3, respectively; and finally, comprehensively judging the optimal winner according to the expert scoring module and the machine scoring module.
Further, S40, an evaluation score list is obtained, and a winning bidder is determined.
Comprehensively calculating the output of the expert scoring module and the output of the machine scoring module to obtain the bid marking numbers of a plurality of bidding parties;
wherein, the calculation formula of the index number is as follows:
S bidder ＝ξ 0 *ER+ξ 1 *MR；
Wherein S is bidder Expressed as the number of marks; ER represents the score for the expert scoring module; scoring of the machine scoring module by MR; zeta type toy 0 A degree of influence expressed as the expert scoring module; zeta type toy 1 A degree of influence expressed as the machine scoring module; and xi 0 +ξ 1 ＝1，ξ 0 ＞ξ 1 。
According to the scoring ordering of example 2 and example 3, the two scoring modules are comprehensively scored below to obtain the final winner. The overall scoring trend for the 5 bidders is given with reference to fig. 6.
Example 4:
referring to fig. 7, a bid review supervision system, the technical scheme of the system includes:
the bid price information system comprises a bid price file issuing unit, a bidder information acquiring unit, a bid price storage unit, a bid screening unit, a bid evaluation unit and a score display unit;
the bid-bidding-target-file distribution unit is used for acquiring bid-target files issued by a bid-target party and extracting the advertising content of the bid-target-file;
the bidder information acquisition unit is used for recording information data related to the bidder; wherein the information data includes bidding party performance, project organization plans, project management capabilities, and bid offers;
wherein the bidder performance includes enterprise grade, reputation and credit, similar engineering experience, near five years of quality records and near three years of major security incidents; the project organization plan comprises an engineering project scheme, an address organization structure, a project network plan table and a quality insurance system; the project management capabilities include primary equipment and labor schemes, safety precautions, and delivered project quality; the bid offers include engineering offers, manual in bid price, machine rationality, and prime material cost rationality.
The tender book storage unit is used for safely storing tender books submitted by bidders, and the tender books are stored on a blockchain after being divided into a plurality of data blocks for encryption;
the bid screening unit is used for screening unqualified bidders and bid books with problems; the bidding screening unit comprises bidder qualification screening and tender book searching and rescreening screening;
the bidding party qualification screening sets preset conditions through the system, and screens personnel participating in bidding according to the data of the bidding party information acquisition unit;
the bid checking and re-screening of the bid is carried out by adopting a bid abnormality analysis model to analyze the bid; the specific process of the bidding anomaly analysis model is as follows:
extracting features of the first bidding information and the second bidding from the bidding set by using the bidding anomaly analysis model to obtain a first bidding feature vector and a second bidding feature vector;
calculating similarity values of the first bidding feature vector and the second bidding feature vector by adopting a bidding measurement method; the calculation formula of the similarity value is as follows:
wherein S () is represented as a function of the bid metric method; d, d 1 Expressed as the first bid feature vector; d, d 2 Represented as the second bid feature vector; d, d 1j Expressed as j features in the first bidding feature vector; d, d 2j A j-th feature represented as the second bidding feature vector; n (N) * () Expressed as a feature variability function of the bid; n (N) ∪ () Represented as a feature addition function of the bid;
wherein the N is ∪ () The calculation formula of (2) is as follows:
wherein the N is * () The calculation formula of (2) is as follows:
wherein λ is represented as a constant value; sigma (sigma) j A hyper-parameter denoted as the j-th feature; n (N) ∩ () Expressed as a feature phase function of the bid;multiplying the feature of the bid to obtain a function;
wherein the N is ∩ () The calculation formula of (2) is as follows:
wherein the saidThe calculation formula of (2) is as follows:
and sorting the bidding information in a descending order according to the similarity, grading, and eliminating the bidder with string bidding behavior.
The bid evaluation unit is used for bid winning evaluation of the bidder meeting the requirements; the bid evaluation unit comprises an expert scoring module and a machine scoring module;
the expert scoring module obtains the score of the optimal winning candidate by adjusting the comprehensive combination of the weighted evaluation, the optimal value and the advantage selection of multiple factors; the expert scoring module comprises the following specific processes:
Randomly extracting n experts from an expert database; wherein n is an integer greater than 5; the expert database consists of personnel of a bidding department, an discipline inspection and supervision department, a project charging department and a bidding agency;
selecting an optimal bidder according to the weighted evaluation, the optimal value and the advantage;
wherein the weighted items of the weighted evaluation include bidder performance, project organization plan and project management capability, respectively denoted as P r1 、P r2 And P r3 R is more than or equal to 1 and less than or equal to n, r is expressed as the number of bidders, and n is expressed as the number of bidders; the best value is expressed as a bid price, recorded as TP r The method comprises the steps of carrying out a first treatment on the surface of the The advantage selection includes the expert formulating a bid by varying the evaluation factors, noted as ITD k K is equal to or more than 1 and equal to or less than m, k is the number of the evaluation factors, and m is the number of the evaluation factors;
calculating a composite qualification score for the weighted evaluation of bidders; the calculation mode of the comprehensive qualification score is as follows:
wherein CQS r The composite qualification score expressed as bidder r; w (w) i A weight denoted as the i-th said weighted term;
the bidder is ordered in descending order according to the qualification comprehensive score, and a qualification ordering list is obtained;
The expert reviews the bid according to the evaluation factors and obtains a bid evaluation error value; the calculation formula of the average error value of the bid evaluation is as follows:
/>
wherein ES r The tender review average error value expressed as the bidder r; mu (mu) k The influence coefficient of the bidder r on the kth evaluation factor is expressed; TDV (time domain reflectometer) r A review score representing the tender book of the bidder r; ITDV (International traffic light vector) k Represented as the ITD k A set score for (2); EP (EP) r Representing additional scores obtained by the bidder r according to the qualification ordering list;
and scoring the bidder according to the average error value of the bid evaluation, and outputting a bidder scoring list.
The machine scoring module inputs the candidate bidder set and the information set for scoring;
the specific process of the machine scoring module comprises the following steps:
initializing the weight of the machine scoring module;
the weight is a bidding factor set by a bidding party; the weight adopts w 0 ＝{w 1 ,...,w n -representation;
inputting information data of bidder after preprocessing, namely x i The method comprises the steps of carrying out a first treatment on the surface of the Wherein said x i For the information data, i is an arbitrary natural number of 1 to n;
calculating a value obtained by the information data through shallow neurons by the machine scoring module; wherein, the calculation formula of the value is:
Wherein Z is j A value calculated for the j-th said shallow neuron; f () is a calculation function of the machine scoring module;represented as w at initialization i Is the j-th weight value of (2);
after forward propagation for many times, an output layer outputs scores; the score calculation formula is:
wherein Y is represented as the score;a weight value representing n-1 layers; />Expressed as n layers of output values;
updating the weight value when each forward propagation occurs; the updating formula of the weight value is as follows:
wherein,representing the weight value updated by n times of calculation; η is expressed as a bidding factor gain factor of the bidder; sigma (sigma) ij Expressed as the difference between the bidder and the i-th bidder's expected value for j factors;
wherein the sigma ij The calculation formula of (2) is as follows:
wherein y is ij A score for the ith factor for the ith said bidder;expressed as a desired score by the bidding party for the ith bidding party with respect to the jth factor.
Comprehensively calculating the output of the expert scoring module and the output of the machine scoring module to obtain the bid marking numbers of a plurality of bidding parties;
wherein, the calculation formula of the index number is as follows:
S bidder ＝ξ 0 *ER+ξ 1 *MR；
wherein S is bidder Expressed as the number of marks; ER represents the score for the expert scoring module; scoring of the machine scoring module by MR; zeta type toy 0 A degree of influence expressed as the expert scoring module; zeta type toy 1 A degree of influence expressed as the machine scoring module; and xi 0 +ξ 1 ＝1，ξ 0 ＞ξ 1 。
And the scoring display unit is used for displaying scoring results of all the bidders and performing descending order arrangement.
The foregoing describes specific embodiments of the present disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
The present disclosure is intended to embrace all such alternatives, modifications and variances which fall within the broad scope of the appended claims. Any omissions, modifications, equivalents, improvements, and the like, which are within the spirit and principles of the one or more embodiments of the disclosure, are therefore intended to be included within the scope of the disclosure.
Claims (10)
1. A bid review supervision method, the method comprising the steps of:
Acquiring a tender book and information data of a tender party, and respectively constructing a tender book set and an information set; constructing a bidding anomaly analysis model to screen the bidding book; the specific process of the bidding anomaly analysis model is as follows: extracting features of a first tender book and a second tender book of the tender book set by using the tender abnormality analysis model to obtain a first tender book feature vector and a second tender book feature vector; calculating similarity values of the first bidding feature vector and the second bidding feature vector by adopting a bidding measurement method; the calculation formula of the similarity value is as follows:
wherein S () is represented as a function of the bid metric method; d, d 1 Expressed as the first bid feature vector; d, d 2 Represented as the second bid feature vector; d, d 1j Expressed as j features in the first bidding feature vector; d, d 2j A j-th feature represented as the second bidding feature vector; n (N) * () Expressed as a feature variability function of the bid; n (N) ∪ () Represented as a feature addition function of the bid; wherein the N is ∪ () The calculation formula of (2) is as follows:
wherein the N is * () The calculation formula of (2) is as follows:
wherein λ is represented as a constant value; sigma (sigma) j Super-feature denoted as jth featureParameters; n (N) ∩ () Expressed as a feature phase function of the bid;multiplying the feature of the bid to obtain a function; sorting a plurality of bidding information in a descending order according to the similarity, grading, and eliminating the bidders with string bidding behaviors to obtain candidate bidding sets; inputting the candidate bid set and the information set into a bid evaluation model; and outputting the label scores of a plurality of bidding parties.
2. The bid review and supervision method of claim 1, wherein the set of information comprises: bidder performance, project organization plans, project management capabilities, and bid quotes; wherein the bidder performance includes enterprise grade, reputation and credit, similar engineering experience, near five years of quality records and near three years of major security incidents; the project organization plan comprises an engineering project scheme, an address organization structure, a project network plan table and a quality insurance system; the project management capabilities include primary equipment and labor schemes, safety precautions, and delivered project quality; the bid offers include engineering offers, manual unit price in bidding, rationality of construction machinery, and rationality of major material costs.
3. The bid review and supervision method of claim 1, further comprising preprocessing information prior to constructing the information set; wherein the preprocessing includes normalizing the information data; wherein, the standardized formula is:
wherein N () is the normalization formula; z is Z i,j A j-th information index expressed as an i-th bidder; z is Z j,B An optimal value expressed as a bidder jth information index; z is Z i,w Represented asWorst value of jth information index of bidding party.
4. The bid review supervision method of claim 1, wherein the ranking comprises-1, 0, and 1; wherein, -1 is represented as the bidder string; 0 is represented as the bid Fang Yishi string; 1 indicates that the bidder is normal.
5. The bid review supervision method of claim 1, wherein the bid evaluation model comprises an expert scoring module and a machine scoring module; comprehensively calculating the output of the expert scoring module and the output of the machine scoring module to obtain the bid marking numbers of a plurality of bidding parties; wherein, the calculation formula of the index number is as follows: s is S bidder ＝ξ 0 *ER+ξ 1 * MR; wherein S is bidder Expressed as the number of marks; ER represents the score for the expert scoring module; scoring of the machine scoring module by MR; zeta type toy 0 A degree of influence expressed as the expert scoring module; zeta type toy 1 A degree of influence expressed as the machine scoring module; and xi 0 +ξ 1 ＝1，ξ 0 ＞ξ 1 。
6. The bid review and supervision method according to claim 5, wherein the specific process of the expert scoring module comprises: randomly extracting n experts from an expert database; wherein n is an integer greater than 5; the expert database consists of personnel of a bidding department, an discipline inspection and supervision department, a project charging department and a bidding agency; selecting and obtaining an optimal bidding party according to the weighted evaluation, the optimal value and the advantages; wherein the weighted items of the weighted evaluation include bidder performance, project organization plan and project management capability, respectively denoted as P r1 、P r2 And P r3 R is more than or equal to 1 and less than or equal to n, r is expressed as the number of bidders, and n is expressed as the number of bidders; the best value is expressed as a bid price, recorded as TP r The method comprises the steps of carrying out a first treatment on the surface of the The advantage selection includes the expert formulating a bid by varying the evaluation factors, noted as ITD k K is equal to or more than 1 and equal to or less than m, k is the number of the evaluation factors, and m is the number of the evaluation factors; calculating a composite qualification score for the weighted evaluation of bidders; the calculation mode of the comprehensive qualification score is as follows: Wherein CQS r The composite qualification score expressed as bidder r; w (w) i A weight denoted as the i-th said weighted term; the bidder is ordered in descending order according to the qualification comprehensive score, and a qualification ordering list is obtained; the expert reviews the bid according to the evaluation factors and obtains a bid evaluation error value; the calculation formula of the average error value of the bid evaluation is as follows: />Wherein ES r The tender review average error value expressed as the bidder r; mu (mu) k An influence coefficient expressed as a kth evaluation factor of the bidder r; TDV (time domain reflectometer) r A review score representing the tender book of the bidder r; ITDV (International traffic light vector) k Represented as the ITD k A set score for (2); EP (EP) r Representing additional scores obtained by the bidder r according to the qualification ordering list; and scoring the bidder according to the average error value of the bid evaluation, and outputting a bidder scoring list.
7. A bid review supervision method according to claim 5, wherein the specific process of the machine scoring module comprises: initializing the weight of the machine scoring module; the weight is a bidding factor set by a bidding party; the weight adopts w 0 ＝{w 1 ,...,w n -representation; inputting information data of bidder after preprocessing, namely x i The method comprises the steps of carrying out a first treatment on the surface of the Wherein said x i For the information data, i is an arbitrary natural number of 1 to n; calculating a value obtained by the information data through shallow neurons by the machine scoring module; wherein,the calculation formula of the value is as follows:wherein Z is j A value calculated for the j-th said shallow neuron; f () is a calculation function of the machine scoring module; />Represented as w at initialization i Is the j-th weight value of (2); after forward propagation for many times, an output layer outputs scores; the score calculation formula is:wherein Y is represented as the score; />A weight value representing n-1 layers; />Expressed as n layers of output values; updating the weight value when each forward propagation occurs; the updating formula of the weight value is as follows: />Wherein (1)>Representing the weight value updated by n times of calculation; η is expressed as a bidding factor gain factor of the bidder; sigma (sigma) ij Expressed as the difference between the bidder and the i-th bidder's expected value for j factors; wherein the sigma ij The calculation formula of (2) is as follows: />Wherein y is ij A score for the ith factor for the ith said bidder; />Expressed as a desired score by the bidding party for the ith bidding party with respect to the jth factor.
8. A bid review supervision system, the system comprising: the bid price information system comprises a bid price file issuing unit, a bidder information acquiring unit, a bid price storage unit, a bid screening unit, a bid evaluation unit and a score display unit; the bid-bidding-target-file distribution unit is used for acquiring bid-target files issued by a bid-target party and extracting the advertising content of the bid-target-file; the bidder information acquisition unit is used for recording information data related to the bidder; the tender book storage unit is used for safely storing tender books submitted by bidders, and the tender books are stored in a blockchain after being divided into a plurality of data blocks for encryption; the bid screening unit is used for screening unqualified bidders and bid books with problems; the bid evaluation unit is used for bid winning evaluation of the bidder meeting the requirements; and the scoring display unit is used for displaying scoring results of all the bidders and performing descending order arrangement.
9. The bid review supervision system according to claim 8, wherein the bid screening unit includes bidder qualification screening and bid book review screening; the bidding party qualification screening sets preset conditions through the system, and screens personnel participating in bidding according to the data of the bidding party information acquisition unit; the bid checking and re-screening of the bid is carried out by adopting a bid abnormality analysis model to analyze the bid; the specific process of the bidding anomaly analysis model is as follows: extracting features of the first bidding information and the second bidding from the bidding set by using the bidding anomaly analysis model to obtain a first bidding feature vector and a second bidding feature vector; calculating similarity values of the first bidding feature vector and the second bidding feature vector by adopting a bidding measurement method; the calculation formula of the similarity value is as follows:
Wherein S () is represented as a function of the bid metric method; d, d 1 Expressed as the first bid feature vector; d, d 2 Represented as the second bid feature vector; d, d 1j Expressed as j features in the first bidding feature vector; d, d 2j A j-th feature represented as the second bidding feature vector; n (N) * () Expressed as a feature variability function of the bid; n (N) ∪ () Represented as a feature addition function of the bid; wherein the N is ∪ () The calculation formula of (2) is as follows:
wherein the N is * () The calculation formula of (2) is as follows:
wherein λ is represented as a constant value; sigma (sigma) j A hyper-parameter denoted as the j-th feature; n (N) ∩ () Expressed as a feature phase function of the bid;multiplying the feature of the bid to obtain a function; and sorting the bidding information in a descending order according to the similarity, grading, and eliminating the bidder with string bidding behavior.
10. The bid review supervision system of claim 8, wherein the bid evaluation unit comprises an expert scoring module and a machine scoring module; the expert scoring module obtains the score of the optimal winning candidate by adjusting the comprehensive combination of the weighted evaluation, the optimal value and the advantage selection of multiple factors; the special purpose is The specific process of the home scoring module comprises the following steps: randomly extracting n experts from an expert database; wherein n is an integer greater than 5; the expert database consists of personnel of a bidding department, an discipline inspection and supervision department, a project charging department and a bidding agency; selecting an optimal bidder according to the weighted evaluation, the optimal value and the advantage; wherein the weighted items of the weighted evaluation include bidder performance, project organization plan and project management capability, respectively denoted as P r1 、P r2 And P r3 R is more than or equal to 1 and less than or equal to n, r is expressed as the number of bidders, and n is expressed as the number of bidders; the best value is expressed as a bid price, recorded as TP r The method comprises the steps of carrying out a first treatment on the surface of the The advantage selection includes the expert formulating a bid by varying the evaluation factors, noted as ITD k K is equal to or more than 1 and equal to or less than m, k is the number of the evaluation factors, and m is the number of the evaluation factors; calculating a composite qualification score for the weighted evaluation of bidders; the calculation mode of the comprehensive qualification score is as follows:wherein CQS r The composite qualification score expressed as bidder r; w (w) i A weight denoted as the i-th said weighted term;
the bidder is ordered in descending order according to the qualification comprehensive score, and a qualification ordering list is obtained; the expert reviews the bid according to the evaluation factors and obtains a bid evaluation error value; the calculation formula of the average error value of the bid evaluation is as follows: Wherein ES r The tender review average error value expressed as the bidder r; mu (mu) k The influence coefficient of the bidder r on the kth evaluation factor is expressed; TDV (time domain reflectometer) r A review score representing the tender book of the bidder r; ITDV (International traffic light vector) k Represented as the ITD k A set score for (2); EP (EP) r Representing additional scores obtained by the bidder r according to the qualification ordering list; according toThe average error value of the tender book review scores the bidder, and a bidder scoring list is output; the specific process of the machine scoring module comprises the following steps: initializing the weight of the machine scoring module; the weight is a bidding factor set by a bidding party; the weight adopts w 0 ＝{w 1 ,...,w n -representation; inputting information data of bidder after preprocessing, namely x i The method comprises the steps of carrying out a first treatment on the surface of the Wherein said x i For the information data, i is an arbitrary natural number of 1 to n; calculating a value obtained by the information data through shallow neurons by the machine scoring module; wherein, the calculation formula of the value is: />Wherein Z is j A value calculated for the j-th said shallow neuron; f () is a calculation function of the machine scoring module; />Represented as w at initialization i Is the j-th weight value of (2); after forward propagation for many times, an output layer outputs scores; the score calculation formula is: / >Wherein Y is represented as the score; />A weight value representing n-1 layers; />Expressed as n layers of output values; updating the weight value when each forward propagation occurs; the updating formula of the weight value is as follows: />Wherein (1)>Representing the weight value updated by n times of calculation; η is expressed as a bidding factor gain factor of the bidder; sigma (sigma) ij Expressed as the difference between the bidder and the i-th bidder's expected value for j factors; wherein the sigma ij The calculation formula of (2) is as follows:wherein y is ij A score for the ith factor for the ith said bidder; />Representing a desired score for the bidder for the ith of the bidders with respect to the jth factor; comprehensively calculating the output of the expert scoring module and the output of the machine scoring module to obtain the bid marking numbers of a plurality of bidders; wherein, the calculation formula of the index number is as follows: s is S bidder ＝ξ 0 *ER+ξ 1 * MR; wherein S is bidder Expressed as the number of marks; ER represents the score for the expert scoring module; scoring of the machine scoring module by MR; zeta type toy 0 A degree of influence expressed as the expert scoring module; zeta type toy 1 A degree of influence expressed as the machine scoring module; and xi 0 +ξ 1 ＝1，ξ 0 ＞ξ 1 。/>
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311535768.4A CN117495298A (en) | 2023-11-17 | 2023-11-17 | Bid bid review supervision method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311535768.4A CN117495298A (en) | 2023-11-17 | 2023-11-17 | Bid bid review supervision method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117495298A true CN117495298A (en) | 2024-02-02 |
Family
ID=89674287
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311535768.4A Pending CN117495298A (en) | 2023-11-17 | 2023-11-17 | Bid bid review supervision method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117495298A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117853211A (en) * | 2024-03-07 | 2024-04-09 | 安徽博诺思信息科技有限公司 | Intelligent management monitoring system for bidding site |
-
2023
- 2023-11-17 CN CN202311535768.4A patent/CN117495298A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117853211A (en) * | 2024-03-07 | 2024-04-09 | 安徽博诺思信息科技有限公司 | Intelligent management monitoring system for bidding site |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108564286B (en) | Artificial intelligent financial wind-control credit assessment method and system based on big data credit investigation | |
| Cho et al. | Development of a new technology product evaluation model for assessing commercialization opportunities using Delphi method and fuzzy AHP approach | |
| CN107851097B (en) | Data analysis system, data analysis method, data analysis program, and storage medium | |
| JP5566422B2 (en) | High risk procurement analysis and scoring system | |
| CN107025596B (en) | Risk assessment method and system | |
| Huang et al. | Winner determination for risk aversion buyers in multi-attribute reverse auction | |
| CN109711955B (en) | Poor evaluation early warning method and system based on current order and blacklist base establishment method | |
| CN111507831A (en) | Credit risk automatic assessment method and device | |
| CN112102073A (en) | Credit risk control method and system, electronic device and readable storage medium | |
| CN117495298A (en) | Bid bid review supervision method and system | |
| CN110751355A (en) | Scientific and technological achievement assessment method and device | |
| CN113537807B (en) | Intelligent wind control method and equipment for enterprises | |
| Arena | Impossible certainty: Cost risk analysis for Air Force systems | |
| CN113919886A (en) | Data characteristic combination pricing method and system based on summer pril value and electronic equipment | |
| CN112559900A (en) | Product recommendation method and device, computer equipment and storage medium | |
| CN114139065B (en) | Talent screening and recommending method and system based on big data and readable storage medium | |
| CN116468273A (en) | Customer risk identification method and device | |
| CN113011788B (en) | Emergency decision method for offshore traffic accident, terminal equipment and storage medium | |
| CN111191889A (en) | Scoring card development method based on logistic regression and voting type model integration | |
| Rahmanita et al. | Model multi criteria decision making with fuzzy ANP method for performance measurement small medium enterprise (SME) | |
| CN110750572A (en) | Adaptive method and device for heuristic evaluation of scientific and technological achievements | |
| CN117689209A (en) | Bridge security risk assessment method, equipment and medium based on entropy weight fuzzy synthesis | |
| CN117114812A (en) | Financial product recommendation method and device for enterprises | |
| Kozioł-Kaczorek | The use of combined multicriteria method for the valuation of real estate | |
| CN107220778A (en) | A kind of method, device and the electronic equipment of employee's credit appraisal and application |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |