KR102360384B1 - System for providing bigdata based reservation price probability distribution validation service for procurement auction - Google Patents

Abstract

Provided is a system for providing a bigdata based probability distribution verification service for bidding in a public procurement market, which comprises: a user terminal which selects a notice number for which a bid price is to be calculated from a notice list, extracts and outputs at least one reservation price which can be won when participating in a public bidding of the selected notice number through big data-based data analysis, and outputs the probability distribution verification result of at least one reservation price based on at least one condition; a collection unit which collects at least one public procurement bidding data using a web crawler; a loading unit for loading structured data, semi-structured and unstructured data in the collected at least one public procurement bidding data; a preprocessing unit including purification and standardization procedures corresponding any one or at least one combination of modification, correction, and deletion by finding an error in which at least the one public procurement bidding data is inconsistent with government bidding enforcement standards or government contract enforcement standards; an analysis unit which visualizes the at least one reservation price of a public bid selected by the user terminal as a probability distribution and delivers the visualization result to the user terminal; and a verification unit which verifies the probability distribution based on at least one condition and transmits the verification result to the user terminal.

Description

SYSTEM FOR PROVIDING BIGDATA BASED RESERVATION PRICE PROBABILITY DISTRIBUTION VALIDATION SERVICE FOR PROCUREMENT AUCTION}

The present invention relates to a system for providing a big data-based probability distribution verification service for bidding in a public procurement market, and provides a platform capable of verifying and evaluating the accuracy of the predicted price probability distribution based on big data, inverse probability, and data mining do.

Currently, in Korea, an electronic bidding system called the Nara Marketplace (G2B system) was introduced to simplify the bidding method and prevent bidding fraud, and electronic bidding is a system that selects successful bidders by electronically processing bidding. As for the method, the winning bid method through multiple preliminary lottery is used. It is a method used by several e-bidding companies. It is a prediction method that statistically uses the collection of bid information notified from the public and private sectors and the section of accumulated bid results. There are a selection method, a method using the distribution map of the bid amount predicted by an expert, and a method using the section probability of the estimated price assessment rate. It predicts the bid amount by applying a plurality of statistical analysis methods such as the distribution of the winning bid amount and the frequency from the bid result.

At this time, a method for predicting the successful bid price of the ordering organization and verifying the prediction model was researched and developed. In No. 2021-0047158 (published on April 29, 2021), the successful bid prediction server receives from the ordering organization server successful bid information including at least one of the ordering organization's bid notice, assessment rate, open bid, basic amount, and preliminary price, A successful bid prediction database is built using the successful bid information, and when data including the region, ordering organization, order item and period entered by the user are input, the prediction point corresponding to the predictor price and assessment rate is determined by executing the analysis and statistics. In order to provide a configuration that provides a result output screen including, and to validate and verify the reliability of the finite element model through statistical analysis, statistically verify the validity of the analysis model and improve the prediction result by introducing and verifying the statistical model and a configuration for performing correction are respectively disclosed.

However, in the former case, it is difficult to accurately guess the winning bid and predict the exact bid because there is not enough information about the random number system, which is the winning bid lottery method used by the ordering party. However, it is difficult to apply because it is a configuration that verifies and corrects the reliability of the mechanical properties of the finite element model designed by simulating a tensile specimen with a statistical model. Therefore, research and development of a platform that verifies the probability distribution of the expected price under different conditions that are different for each ordering institution, range of reserve price, and industry, and the prediction result before and after opening of public bidding is required.

An embodiment of the present invention calculates a probability distribution using big data, inverse probability, and data mining-based models for at least one reservation price that can be successfully bid in a public bidding, and verifies the accuracy of the probability distribution Big data-based probability distribution for bidding in the public procurement market, which can verify the probability distribution of the expected price under different conditions that are different for each ordering institution, price range, and industry, and verify the prediction results before and after the opening of public bidding. A verification service providing system may be provided. However, the technical task to be achieved by the present embodiment is not limited to the technical task as described above, and other technical tasks may exist.

As a technical means for achieving the above-described technical problem, an embodiment of the present invention selects a public notice number for which a bid price is to be calculated from the notice list, and participates in the public bidding of the selected notice number, at least one A user terminal that extracts and outputs a reservation price through big data-based data analysis, and outputs a probability distribution verification result of at least one reserved price based on at least one condition, and at least one public procurement bidding data A collection unit that collects using a crawler, a loading unit that loads structured data, semi-structured and unstructured data in the collected at least one public procurement bidding data, at least one public procurement bidding A pre-processing unit that performs pre-processing including purification and standardization procedures corresponding to any one or at least one combination of correction, correction, and deletion, in which data is found for errors that deviate from government bid enforcement standards or government contract enforcement standards, users; A verification service providing server comprising: an analysis unit that visualizes at least one expected price of a public bid selected in a terminal as a probability distribution and delivers it to a user terminal; and a verification unit that verifies the probability distribution based on at least one condition and transmits it to the user terminal includes

According to any one of the above-described problem solving means of the present invention, a probability distribution is calculated using big data, inverse probability, and data mining-based models for at least one reservation price that can be successful in a public bidding, In order to verify the accuracy of the probability distribution, it is possible to verify the probability distribution of the expected price under different conditions that are different for each ordering institution, the price range, and each industry, and to verify the prediction results before and after the opening of a public bid.

1 is a view for explaining a system for providing a big data-based probability distribution verification service for bidding in a public procurement market according to an embodiment of the present invention.
FIG. 2 is a block diagram illustrating a verification service providing server included in the system of FIG. 1 .
3 and 4 are diagrams for explaining an embodiment in which a big data-based probability distribution verification service for bidding in a public procurement market according to an embodiment of the present invention is implemented.
5 is an operation flowchart illustrating a method of providing a big data-based probability distribution verification service for bidding in a public procurement market according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement them. However, the present invention may be embodied in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part is "connected" with another part, this includes not only the case of being "directly connected" but also the case of being "electrically connected" with another element interposed therebetween. . In addition, when a part "includes" a certain component, it means that other components may be further included, rather than excluding other components, unless otherwise stated, and one or more other features However, it is to be understood that the existence or addition of numbers, steps, operations, components, parts, or combinations thereof is not precluded in advance.

The terms "about", "substantially", etc. to the extent used throughout the specification are used in or close to the numerical value when manufacturing and material tolerances inherent in the stated meaning are presented, and are intended to enhance the understanding of the present invention. To help, precise or absolute figures are used to prevent unfair use by unconscionable infringers of the stated disclosure. As used throughout the specification of the present invention, the term "step of (to)" or "step of" does not mean "step for".

In this specification, a "part" includes a unit realized by hardware, a unit realized by software, and a unit realized using both. In addition, one unit may be implemented using two or more hardware, and two or more units may be implemented by one hardware. Meanwhile, '~ unit' is not limited to software or hardware, and '~ unit' may be configured to be in an addressable storage medium or to reproduce one or more processors. Thus, as an example, '~' denotes components such as software components, object-oriented software components, class components, and task components, and processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables. The functions provided in the components and '~ units' may be combined into a smaller number of components and '~ units' or further separated into additional components and '~ units'. In addition, components and '~ units' may be implemented to play one or more CPUs in a device or secure multimedia card.

Some of the operations or functions described as being performed by the terminal, apparatus, or device in the present specification may be performed instead of by a server connected to the terminal, apparatus, or device. Similarly, some of the operations or functions described as being performed by the server may also be performed in a terminal, apparatus, or device connected to the server.

In this specification, some of the operations or functions described as mapping or matching with the terminal means mapping or matching the terminal's unique number or personal identification information, which is the identification data of the terminal. can be interpreted as

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

1 is a view for explaining a system for providing a big data-based probability distribution verification service for bidding in a public procurement market according to an embodiment of the present invention. Referring to FIG. 1 , the big data-based probability distribution verification service providing system 1 for bidding in the public procurement market includes at least one user terminal 100 , a verification service providing server 300 , and at least one manager terminal ( 400) may be included. However, since the big data-based probability distribution verification service providing system 1 for bidding in the public procurement market of FIG. 1 is only an embodiment of the present invention, the present invention is not limitedly interpreted through FIG. 1 .

At this time, each component of FIG. 1 is generally connected through a network (Network, 200). For example, as shown in FIG. 1 , at least one user terminal 100 may be connected to the verification service providing server 300 through the network 200 . In addition, the verification service providing server 300 may be connected to at least one user terminal 100 and at least one manager terminal 400 through the network 200 . In addition, at least one manager terminal 400 may be connected to the verification service providing server 300 through the network 200 .

Here, the network refers to a connection structure in which information exchange is possible between each node, such as a plurality of terminals and servers, and an example of such a network includes a local area network (LAN), a wide area network (WAN: Wide Area Network), the Internet (WWW: World Wide Web), wired and wireless data communication networks, telephone networks, wired and wireless television networks, and the like. Examples of wireless data communication networks include 3G, 4G, 5G, 3rd Generation Partnership Project (3GPP), 5th Generation Partnership Project (5GPP), Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi (Wi-Fi) , Internet, LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), RF (Radio Frequency), Bluetooth (Bluetooth) network, NFC ( Near-Field Communication) networks, satellite broadcast networks, analog broadcast networks, Digital Multimedia Broadcasting (DMB) networks, and the like are included, but are not limited thereto.

In the following, the term at least one is defined as a term including the singular and the plural, and even if the at least one term does not exist, each element may exist in the singular or plural, and may mean the singular or plural. it will be self-evident In addition, that each component is provided in singular or plural may be changed according to embodiments.

The at least one user terminal 100 may be a terminal of individuals and companies who want to participate in a public bidding using a web page, an app page, a program or an application related to a big data-based probability distribution verification service for bidding in the public procurement market. have. In this case, the user terminal 100 may be a terminal that selects an announcement number to participate in the bidding and receives the expected price predicted by the verification service providing server 300 . In addition, the user terminal 100 verifies the probability distribution of the expected price in text and graph before and after the ticket opening by verifying the accuracy of the probability distribution provided under different conditions, such as by ordering organization, by preliminary price range, by industry, and by participant, in text and graphs in the future. It may be a terminal that wants to prevent subjective errors from occurring in an upcoming public bidding.

Here, the at least one user terminal 100 may be implemented as a computer that can access a remote server or terminal through a network. Here, the computer may include, for example, navigation, a laptop equipped with a web browser, a desktop, and a laptop. In this case, the at least one user terminal 100 may be implemented as a terminal capable of accessing a remote server or terminal through a network. At least one user terminal 100 is, for example, as a wireless communication device that guarantees portability and mobility, navigation, PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS(Personal Handyphone System), PDA(Personal Digital Assistant), IMT(International Mobile Telecommunication)-2000, CDMA(Code Division Multiple Access)-2000, W-CDMA(W-Code Division Multiple Access), Wibro(Wireless Broadband Internet) ) terminal, a smart phone, a smart pad, a tablet PC, etc. may include all kinds of handheld-based wireless communication devices.

The verification service providing server 300 may be a server that provides a big data-based probability distribution verification service web page, an app page, a program, or an application for bidding in the public procurement market. In addition, the verification service providing server 300 may be a server that collects data published by at least one institution using a web crawler or a bot. In addition, the verification service providing server 300 classifies data in order to construct big data, removes errors by performing pre-processing, and standardizes, then any one or at least one of big data analysis, data mining, and inverse probability model It may be a server that extracts a predetermined price using a combination of , and transmits it to the user terminal 100 . In addition, the verification service providing server 300 analyzes the probability distribution of the expected price before and after the ticket wicket to check the accuracy of the probability distribution provided under different conditions, such as by ordering organization, by preliminary price range, by industry, by participant, in text and graph for each model. It may be a server that performs probability distribution verification.

Here, the verification service providing server 300 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, navigation, a laptop equipped with a web browser, a desktop, and a laptop.

At least one manager terminal 400 checks whether there is an error in data when constructing big data using a web page, an app page, a program or an application related to a big data-based probability distribution verification service for bidding in the public procurement market. It may be an administrator's terminal that confirms, corrects, deletes, corrects, and standardizes the form of data. In addition, the manager terminal 400, when verifying the probability distribution output from the big data, inverse probability, and data mining model in the verification service providing server 300, before and after the ticket opening by ordering organization, by reserve price range, by industry, It may be a terminal that sets different conditions, such as for each participant, or checks the accuracy once again when the accuracy is automatically verified.

Here, the at least one manager terminal 400 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer may include, for example, navigation, a laptop equipped with a web browser, a desktop, and a laptop. In this case, the at least one manager terminal 400 may be implemented as a terminal capable of accessing a remote server or terminal through a network. At least one manager terminal 400, for example, as a wireless communication device that guarantees portability and mobility, navigation, PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS(Personal Handyphone System), PDA(Personal Digital Assistant), IMT(International Mobile Telecommunication)-2000, CDMA(Code Division Multiple Access)-2000, W-CDMA(W-Code Division Multiple Access), Wibro(Wireless Broadband Internet) ) terminal, a smart phone, a smart pad, a tablet PC, etc. may include all kinds of handheld-based wireless communication devices.

2 is a block diagram illustrating a verification service providing server included in the system of FIG. 1, and FIG. 3 is a big data-based probability distribution verification service for bidding in the public procurement market according to an embodiment of the present invention. It is a diagram for explaining an embodiment.

Referring to FIG. 2 , the verification service providing server 300 includes a collection unit 310 , a loading unit 320 , a preprocessing unit 330 , an analysis unit 340 , a verification unit 350 , and an upload unit 360 . and an artificial intelligence unit 370 .

The verification service providing server 300 according to an embodiment of the present invention or another server (not shown) operating in conjunction with at least one user terminal 100 and at least one manager terminal 400 is used for bidding in the public procurement market. When transmitting a big data-based probability distribution verification service application, program, app page, web page, etc. for Big data-based probability distribution verification service applications, programs, app pages, web pages, etc. can be installed or opened. In addition, the service program may be driven in at least one user terminal 100 and at least one manager terminal 400 by using a script executed in a web browser. Here, the web browser is a program that enables the use of a web (WWW: World Wide Web) service, and refers to a program that receives and displays hypertext written in HTML (Hyper Text Mark-up Language), for example, Netscape. , Explorer, Chrome, and the like. In addition, the application means an application on the terminal, for example, includes an app (App) executed in a mobile terminal (smartphone).

Before describing FIG. 2, a basic concept for understanding the service according to an embodiment of the present invention will be described first. The basic concept described below will not be repeatedly described with reference to FIG. 2 .

In Korea, the method of selecting successful bidders through competitive bidding is stipulated in Article 42 Paragraph 1 of the Enforcement Decree of the Act on Contracts to which the State is a Party, which is a related statute. In other words, prior to the conclusion of a bid or contract, the qualification screening system that determines the successful bidder by examining the contract performance ability, etc. are doing When a public institution procures goods or implements a public project, the bidder with the lowest price is selected as the successful bidder through competitive bidding. Various types of procurement auction methods are being used.

Korea's qualification screening system requires companies that have achieved a certain score or higher to be selected as successful bidders in order to prevent a decrease in the quality of delivered goods, construction, and services that may occur due to excessive price competition. If the non-price factor score is judged to be the same as that of other companies, in order to become a successful bidder, the bid price must be presented at a level that can achieve a certain score or higher. Accordingly, if the non-price factor scores such as product production capacity satisfy certain conditions in the qualification screening system, the bid price offered by the participating companies plays a very important role in determining the successful bidder.

Currently, in the procurement market, it is a principle that the expected price is not disclosed until the opening of the bid, and in order to prevent corruption problems due to leakage, etc. In this case, 15 preliminary prices are randomly selected within a certain range of the construction cost of the project investigated by the ordering party, and the estimated price is determined by arithmetic average of the 4 drawn preliminary prices. In this case, neither the bidder nor the ordering party (organization) will know which expected price will be selected until the bid is opened. Therefore, due to the nature of the qualification examination system in Korea, which judges whether a bid is successful in the order of bidding with the lowest price below the expected price, the disclosure of the expected price is very important in determining at what price the bidder will participate in the bidding. .

In this regard, domestic and international studies have been conducted on the effect of the disclosure of the expected price on the bid price and bid result suggested by bidders. These studies present somewhat contradictory analysis results depending on the research hypothesis. Elyakime, Laffont, Loisel, and Vuong analyzed the First-Price Sealed Bid Auction used by the French government in a Standing Timber auction, where a single seller is expected in a competitive bidding. A model was set up for the case of not disclosing the reservation price to bidders and a Bayesian equilibrium strategy was obtained. Here, they have shown theoretically and empirically that the strategy of disclosing the expected price is better for the seller than the strategy of disclosing it.

Tan analyzed the risk-neutral buyer (government) signing procurement contracts with many suppliers in a model with independent probability distribution and private values. He showed that the optimal procurement mechanism is to indicate their demand information by announcing the expected price to the suppliers in advance when the buyer proposes a contract, such as Elyakime, Laffont, Loisel, and Vuong, and to use the highest-price sealed bid. In this case, the private value refers to a case in which the value obtained by the bidder differs for each individual participating in the bid, and each participant knows the value of himself/herself but does not know the value of the other participant exactly. Therefore, under the assumption of private value, the bidder's profit depends only on the information he knows, whether he will win the bid, and the amount he will pay. On the other hand, under the assumption of joint value, where all bidders get the same value and each bidder knows its own value as well as its value to the other bidders, the bidder's profit is determined by the information, tastes and preferences of not only itself but also other bidders. It also depends on the preferences of those who did not participate in it.

On the other hand, it was also shown that, if the supplier is risk averse, the lowest-price sealed bid with the disclosed expected price may not be the optimal mechanism. It was analyzed that this was because they were induced to bid more aggressively. On the other hand, in the case of Vicent, when bidders have a common value auction, a seller with a probabilistic reservation value can obtain a higher expected utility than when the expected price is disclosed when it is not disclosed. showed that It is analyzed that this is because, by making the expected price undisclosed, those wishing to purchase below the expected price, who would not have participated when it was disclosed, participate in the bidding.

In the case of Rosar, he analyzed the case of using a plurality of expected prices with uncertainty in bidding. He paid attention to the period in which the auction rules were established and announced before the actual auction when the auctioneer sells goods as a seller. During the period during which the auction rules are announced, potential buyers prepare bids and sellers can obtain more information about the value of their use of the auctioned items. When you sell, you can get higher utility. Accordingly, the seller will be tempted to do so if the seller can set the expected price after obtaining such information rather than using the previously disclosed estimated price before obtaining information on his/her auction item. Rosar paid attention to the incentives of the auctioneers and showed that it is optimal for the risk-averse seller to announce the auction rules in advance and to have the right to set the expected price at a later time in the sealed highest price bid. It has also been shown to be optimal when buyers are prohibited from bidding at mid-range prices and are forced to choose extreme bids.

On the other hand, as a theoretical study related to the determination of successful bidders in Korea's qualification system, there is a study by Kim Bong-ju. The concept of Bayesian Equilibrium was used to analyze the price strategies of bidders and the characteristics of the balance. He showed that in the private value auction model with an independent probability distribution, in terms of the buyer's expected cost, like Elyakime, Laffont, Loisel, and Vuong, revealing the expected price can have better policy performance than when the price is not disclosed. However, Bong-Ju Kim's research has limitations in drawing general conclusions because it depends on specific values of parameters in calculating the expected benefits of bidders. The analysis of cases also has limitations that are not rigorous. In addition to this, basic concepts are described in detail in the National Public Procurement Service's Nara Marketplace, such as the calculation principle of multiple reserve prices and scheduled prices, so please refer to them.

Accordingly, an embodiment of the present invention collects and converts procurement data provided by national institutions, local governments and public institutions using a web crawler, and transmits the collected and converted semi-structured or unstructured data to a real-time stream (Real-time). Stream) is processed and loaded into the RDB (Relational Database), and the quantitative data and qualitative data in the RDB are automatically corrected for errors that deviate from the execution standards, and preprocessing is performed through purification and standardization, and based on big data By modeling a model that predicts the probability distribution of the reservation price according to a random lottery, and verifying it under various conditions and before and after the ticket opening, it is decided to provide a platform that can obtain the accuracy of each model and correct it in real time.

Referring to FIG. 2 , the collection unit 310 may collect at least one piece of public procurement bidding data using a web crawler. The collection unit 310 may use a macro bot or a web crawler. The macro program is software that automatically programs and processes simple/repetitive tasks, and keyboard and mouse input values and inputs that are normally performed manually. It is used to increase the efficiency of simple tasks using a computer by inputting tasks such as sequence in advance and executing them automatically and repeatedly. Automating simple/repetitive tasks using a macro program is necessary for efficient business processing, and the macro program can be configured in a way that automates permitted commands.

Alternatively, the collection unit 310 may use a web crawler, which is a computer program that searches the World Wide Web in an organized and automated manner. The work of a web crawler is called web crawling or spidering, and it is a type of bot or software agent. There are two main types of web crawlers: general web crawlers and distributed web crawlers. When explaining the basic operation of a web crawler, it starts by fetching a URL from the URL Frontier module and fetching a web page of the URL using the HTTP protocol. Then, the Fetch module stores the web page in temporary storage, the Parser module extracts the text and links, and the text is sent to the Indexer. In the case of a link, whether it should be added to the URL frontier is determined by going through the Content Seen, URL Filter, and Duplication URL Element modules. At this time, since it is virtually impossible to crawl all web documents with a general web crawler, a distributed web crawler can be further used.

Distributed web crawler is largely divided into two types, one of which is centralized and the other is P2P (or Fully-Distributed). A centralized distributed web crawler is a structure in which the URL manager acts like a server and the crawler acts as a client. After downloading the document from the crawler, extracting the OutLink URL and passing it to the URL manager, the URL manager checks whether the URL of the downloaded document is the URL and removes the URL duplication. In other words, the URL manager takes over the URL duplication and URL management in the general web crawler. On the other hand, in the P2P method, each crawler has a completely independent structure. In the P2P method, each crawler operates like a general web crawler. Each crawler downloads documents, extracts OutLink URLs, and deduplicates URLs. Each crawler operates independently. To do this, the list of downloaded URLs managed by each crawler must be mutually exclusive. Otherwise, the phenomenon of downloading the same document from different crawlers will occur. As a way to solve this, each crawler can manage the download URL domains by dividing them exclusively. In other words, it is a method of managing only what belongs to the download domain and passing the rest of the URLs to other crawlers. If this method is used, each crawler can operate independently.

Next, web content needs to be extracted. Web content extraction technology provides a function to automatically extract content to be used for information analysis from a web document. The web content extraction system is a device that extracts only content by automatically producing rules for extracting content, a rule generator that automatically creates content extraction rules, and a navigation content eliminator that removes navigation content from a given web document. , it may be composed of a content extractor (Core Context Extractor) that extracts content through comparison of keyword similarity with content extraction rules. Of course, it is possible to collect real-time data by various methods other than the above-described crawling method, and is not limited to the listed ones, and is not excluded for reasons not listed.

The loading unit 320 may load structured data, semi-structured data, and unstructured data within the collected at least one public procurement bidding data. The loading unit 320 stores structured data in RDBMS (Relational DataBase Management System) database storage method in MSSQL (Microsoft SQL), semi-structured data stores in RDB (Relational Database), and unstructured data in a distributed file system It can be stored in the Distributed File System (DFS). The loading unit 320 may process and load at least one public procurement bid data in a real time stream. At this time, the structured data is data stored in a fixed field because the data standard and form are determined, the semi-structured data is data that is not in a fixed form, but includes a kind of schema, and the unstructured data is data that is irregular in form. Since the types and specifications of data are different, they are stored in different databases, and the data stored in this way is called raw data. It will be built with big data, which is the basis for this.

The loading unit 320, for example, through a robot that collects, categorizes and stores web data on the Internet through an analysis step of the HTML source and structure of the procurement organization for collection of data for analysis, national institutions, autonomous After collecting and transforming public procurement data provided by organizations and public institutions, the collected and transformed data of various patterns (semi-structured or unstructured) is processed as a real-time stream and loaded into RDB (Relational Database) go through Data provided in XML and JSON format from public data portals such as Nara Market is collected and loaded into RDB. load up The method of loading unstructured data is to collect data from words, Hangul documents, images, etc. or change it to a file format and load it into the RDB. Non-structured data is loaded into the RDB in HTML format using a Java library.

The pre-processing unit 330, at least one public procurement bidding data finds errors that deviate from government bid execution standards or government contract execution standards, and refines corresponding to any one or at least one combination of correction, correction and deletion Pretreatment including standardization procedures may be performed. The type of the at least one public procurement bid data including an error may include a bid rate that is a ratio of a successful bid amount to a predetermined price range, a predetermined price, a lowest price, and the predetermined price. At this time, at least one manager terminal 400 may manually find it and align the deviated data, but if the data is different by automatically programming it, the confirmed data, which is the data after the actual ticket opening, is used as the reference data. Therefore, errors can be eliminated by correcting the other data.

The analysis unit 340 may visualize at least one predetermined price of the public bid selected by the user terminal 100 as a probability distribution and transmit it to the user terminal 100 . The user terminal 100 selects a notice number for which a bid price is to be calculated from the notice list, and when participating in the public bidding of the selected notice number, at least one reserved price (Reservation Price) that can be successful in big data-based data analysis It can be extracted and printed. At this time, the analysis unit 340 may perform analysis by using all three methods or a combination of at least one of the three methods. The first is big data-based, the second is data mining-based, and the third is inverse probability-based. . In addition to this, it will be obvious that prediction values can be derived in various ways.

It provides a probability distribution information service for expected prices related to public bidding. It provides bidding information, qualification information, and comprehensive examination information related to public bidding, and uses analysis technology of big data, inverse probability, and data mining based on quantitative data. Thus, business performance is achieved by providing information on the probability distribution of the expected price of public bidding to decision makers. First, big data-based predictive price probability distribution analysis analyzes descriptive statistics between independent variables, analyzes correlations based on analysis data, and analyzes the validity of statistical hypotheses based on correlation analysis. , it provides information by predicting the probability distribution of the uncertain expected price based on the verification data.

Second, inverse probability-based predicted price probability distribution analysis analyzes descriptive statistics between independent variables, analyzes correlations based on analysis data, and uses analysis technology of inverse probability-based modeling based on analysis data. It provides information by predicting the expected price probability distribution using Third, in the data mining-based probability distribution analysis stage, descriptive statistics between independent variables are analyzed, correlation is analyzed based on the analysis data, and statistical modeling such as a data mining model is used based on the analysis data. Predict the price probability distribution. In addition, information on the probability distribution of the expected price of the decision maker (user) can be utilized. It provides various visual graphs according to probability distribution conditions, and can support final decision making by synthesizing the above-mentioned data.

The verification unit 350 may verify the probability distribution based on at least one condition and transmit it to the user terminal 100 . The at least one condition may include at least one ordering organization, a preliminary price range, and different conditions for each industry. The user terminal 100 may extract and output a reservation price through big data-based data analysis, and output a result of verifying a probability distribution of at least one reserved price based on at least one condition. When the user terminal 100 selects and bids any one of at least one predetermined price for the public bidding, the verification unit 350 may verify a case in which the accuracy of prediction results before and after the opening of the public bidding is different. Since the expected price is the result of the bid price, it is very important to find a probability distribution that can adequately represent the data extracted from the sample with high prediction rate in order to calculate the exact expected price.

This method of determining how appropriate the probability distribution of the predicted predicted price for each model is called the goodness-of-fit test. There are methods such as Von-Mises test and Probability Plot Correlation Coefficient (PPCC) test. Among these methods, PPCC verification is a verification method to check the normality of data in a normal distribution by Filliben, and it is a verification statistic using PPCC verification for various distribution types such as Gumbel, GEV, and LP3 (Log-Pearson Type III) distribution. Values can be presented, and the PPCC verification statistic values for various distribution types can be expressed as a regression equation.

The upload unit 360, at least one manager terminal 400, at least one ordering institution, a preliminary price range, and a different condition for each industry in any one or a combination of at least one probability distribution based on the results before and after the opening You can upload the verified results.

When the prediction results before and after the public bidding are different, the artificial intelligence unit 370 identifies different parts and causes, updates the identified different parts and causes in big data, and then patterns for different parts and causes. Deep learning using artificial neural networks can be performed to extract and predict the error rate through classification, and the data derived from deep learning can be reflected in big data. In this case, a box plot can be used, and the reason for using it is that when it is difficult to visually check a large amount of data, the range and median of the data set can be quickly checked using pictures, and outliers can be identified. have. The box plot is the smallest value among the values in the first quartile minus 1.5 IQR (Inter Quatile Range), the first quartile (Q1) is 25% of the position, the second quartile (Q2) is the median value at the 50% position, the third quartile (Q3) is the 75% position, the maximum value is the largest value among the values in the third quartile plus 1.5 IQR, the outlier is the minimum value and the maximum value Displays the position value that exceeds the value. Values that exceed the minimum and maximum values are called outliers. In public bidding, the values of the companies that bid after the opening of the bid, the bid price, or the qualification requirements of the bidders vary. Rather than verifying using only the existing data, it is more accurate to verify the data at the moment the user inputs it using a larger amount of data that has already been loaded into the data.

<Correlation Analysis>

Among the feature selection techniques, data correlation analysis is a method of reducing the data that needs to be managed while satisfying the significance level of the quality policy by analyzing the data correlation. The correlation coefficient used for correlation analysis indicates what kind of linear relationship (direction and magnitude) there is between two variables. The correlation coefficient indicates the directionality by dividing the covariance indicating the absolute magnitude of the linear relationship between the two variables by the standard deviation of the two variables. When two variables are x and y, the Pearson correlation coefficient is expressed as an equation as shown in Equation 1 below.

In the correlation analysis step, correlation coefficients of independent variables are derived through Pearson correlation analysis, and effective persons with high data correlation can be selected to remove data having duplicate characteristics. And, in the SVM (Support Vector Machine) parameter optimization step, it is possible to find the optimal value of the parameter while changing the SVM parameter value based on cross validation. In the SVM test and accuracy analysis stage, the kernel with the highest accuracy can be selected by comparing and analyzing the performance according to the kernel. In addition, the kernel and SVM parameters with the highest accuracy can be applied to the system. In this case, a Fast Correlation Based Filter Approach algorithm for finding an optimal feature selection using a correlation coefficient may be used, but is not limited thereto. In addition, the closer the Pearson correlation coefficient is to 1, the more positive it is, and the closer it is to -1, the more negative it is. If the Pearson correlation coefficient is close to 0, it means that the correlation is sparse. When the significance probability (P-value) of the correlation coefficient is less than or equal to the set Significant Level, the correlation coefficient can be said to be significant.

Such correlation analysis can be used for big data modeling, data mining modeling, and inverse probability modeling. In addition, various methods other than Pearson's method for correlation analysis may be used.

<MLP & ANFIS>

Meanwhile, the electronic bidding system using the Nara Marketplace (G2B system) has been used since 2002 in the bidding by the Public Procurement Service. Currently, the winning bid method through multiple preliminary price lottery is mainly used in e-bidding, but it is not easy to accurately predict the winning bid price due to the nature of the winning bid method (random number system). In this case, in addition to the three modeling methods described above, a deep learning algorithm may be used for accurate data analysis and successful bid price prediction as a successful bid data analysis method in which an artificial intelligence technique is applied to e-bidding. A Multi-Layer Perceptron (MLP) or an Adaptive Neural Fuzzy Inference System (ANFIS) may be used.

Currently, the bidding method mainly used in electronic bidding is a successful bidding method through multiple preliminary price lottery. Many companies mainly use the high-frequency successful bid section selection method, the low-competition section selection method, and the high-frequency successful bid and low-competition section selection method as existing analysis methods for winning bids. It is a method of predicting the bid amount with a high probability of successful bid using a statistical method using the distribution of the frequency or amount of successful bid sections from the accumulated bidding results in the past. However, it is difficult to estimate the exact successful bid price due to the lack of information on the multiple example price (preliminary price) selection criteria (random number system) of the ordering party and the relationship between past and current bids. Recently, similar research using machine learning and deep learning algorithms has been used for public IT project successful bid prediction, real estate auction successful bid prediction, and foreclosed automobile successful bid prediction. Research using artificial neural networks is generally insufficient.

Therefore, in one embodiment of the present invention, the bid price closest to the lower limit of successful bid can be predicted by predicting the bid price through data analysis of electronic bidding rather than time series data using MLP and ANFIS, which are deep learning techniques rather than statistical methods. There are more methods available. A multi-layer perceptron consists of one or more hidden layers between an input layer and an output layer, as multiple layers of perceptrons are pasted together. It consists of one input layer, one or more hidden layers, and one output layer. The ANFIS algorithm is a model that combines the IF-THEN concept of FIS (Fuzzy Inference-System) and ANN learning ability. It learns fuzzy rules and membership functions from the data given and has the same network as the functionally fuzzy inference system.

After collecting and preprocessing 10 years of successful bid data as a construction item, you can test by inputting the basic price, estimated price, sample price range, and lower bid price as input and outputting the winning bid amount. In order to reduce the amount of deviation, it can be tested by dividing it based on the lower limit of the winning bid rate and normalizing and standardizing it. When the predicted value and the actual value are compared with the regression indicators of MSE, RMSE, MAE, and MAPE, it is possible to check the error between the predicted value and the actual value. If the performance index values are compared and the performance index of each model is closest to 0 in Min-Max, the winning bid price prediction model is constructed using the algorithm of the model to further increase the probability of winning the bid.

<k-nearest neighbor algorithm>

The k-Nearest Neighbors algorithm is a method of using the result information about the k nearest data that knows the result when estimating the result regarding the new input in a given situation with the training data composed of input and output. Estimate the result from k neighbors. In order to apply the k-nearest neighbor algorithm, there must be a way to calculate the distance between the query and the training data. When the data attribute is numeric, the Euclidean distance is used to measure the order and distance from the training data. When two points X(x1, x2, ... , xN) and Y(y1, y2, ..., yN) are given, it can be expressed as in Equation (2).

In a regression problem in which the output is a numerical value, the output value is the average value of k-nearest neighbors, and the value is averaged by applying a weight that is inversely proportional to the distance between the query and the nearest neighbor. Given k-nearest neighbor KNN = [(x1, y1), (x2, y2),...,(xk, yk)], the output y for the input X is the mean value of Equation 3 and the weighted average value of Equation 4 Calculated.

<KNeighborsRegressor analysis modeling>

In order to analyze the successful bid data of e-bidding, four input characteristics (basic amount, estimated price, sample price range, lower limit rate of successful bid) and 1 output characteristic (amount of successful bid) As a result, it can be analyzed by dividing it into 70% of the training dataset and 30% of the test dataset.

Prior to data pre-processing, the variable list of successful bid status data obtained from BidPro can be extracted. It is a measure commonly used when dealing with, and is the standard deviation of the residual. It is a measure of how dispersed the residuals are in the regression data points and is suitable for expressing precision. RMSE can be calculated by Equation (5).

n is the number of data points, yi is the actual data value, and (hat) yi is the predicted value. Here, it can be used to determine whether the model is well trained, and the smaller the RMSE value, the higher the precision. Modeling results can be identified by understanding the accuracy of training and testing according to changes. If the accuracy of the training dataset and the test dataset obtained using k-nearest neighbor regression is extracted, and the accuracy is calculated at the optimal k value estimated using RMSE, the prediction accuracy can be identified. The parameters required to design the model are called Hyper Parameters, and it is not easy to select them. This is because, depending on the Hyper Parameters selected value, low performance results may be obtained.

For this reason, as a way to solve this problem, hyper parameters can be tuned using cross-validation. To evaluate the model, the dataset is divided into a training dataset and a validation dataset, and the training dataset is trained. When the model fits very well on the training dataset but does not fit on the validation dataset, it is called overfilling. To prevent this problem, regularization is performed, and cross-validation, which is one of the tools to prevent overfitting, can be used.

For example, k-fold cross-validation and randomized cross-validation can be used. The k-fold cross-validation is a cross-validation that is performed by making k folds, and it is an evaluation method that has better performance than the method using the train_test_split function of Scikit-learn. For example, divide the dataset evenly into k folds, k-1 into the training dataset, and 1 for the test dataset, repeating k times in turn so that the data enters the test dataset exactly once. becomes more accurate. Accuracy can be improved for datasets with a small number of total data.

Due to the nature of the electronic bidding winning bid method (random number system), even if the accuracy in analyzing and predicting existing data is high, it may not match the actual lottery result. This is because even if the error is small, a difference may occur because the construction amount itself is large. In addition, in order to increase the probability of winning a bid, input variable addition and additional machine learning techniques may be used in combination, or additional machine learning such as TensorFlow MLP and RNN may be further used.

Hereinafter, an operation process according to the configuration of the verification service providing server of FIG. 2 will be described in detail with reference to FIG. 3 as an example. However, it will be apparent that the embodiment is only one of various embodiments of the present invention and is not limited thereto.

Referring to FIG. 3 , a service (bid pro) according to an embodiment of the present invention may proceed with a service according to an embodiment of the present invention in the same flow as in FIG. 3 . In this case, the bid pro is the verification service providing server 300 according to an embodiment of the present invention, and the decision maker is a user of the user terminal 100 . The verification service providing server 300 according to an embodiment of the present invention collects and inspects public procurement data, and then performs quantitative purification and loading. At this time, the error data can be manually or automatically inspected as described above, and for the refined and standardized data, the DB is set and loaded according to the structured, semi-structured, and unstructured data according to the type of data. In addition, the probability distribution is predicted by any one or at least one combination of three models, namely, a big data model, an inverse probability model, and a data mining model. When the range (expected price range) is automatically calculated, a probability distribution according to a random distribution or a normal distribution is provided. In addition, depending on the case where the multiple reserve price application classification is open or undisclosed, the final expected price is provided by performing public prediction and non-public prediction again.

<Click the notice number for which you want to calculate the bid price from the notice list>

Referring to Figure 4a, the user terminal 100, for example, 2021.02.16. It is possible to calculate the bid price for three cases with management checked among the closing announcements. When the user terminal 100 clicks [Bid Information] after logging in, the bidding information system is executed. If you click the notice number for which you want to calculate the bid price from the notice list, the information for each item is displayed (Management check notice number: 20201242334, 20210126177, 20210126191). ① Shows bid notice information such as qualifications for bidding participation, ② Shows information related to bid price calculation (basic amount, bid rate, net construction cost, etc.), and ③ shows qualification or comprehensive evaluation score. In this case, (-X) is ineligible (meaning that the performance score is insufficient), omission is omission of qualification examination (meaning that the final successful bidder is determined only by the bid price), joint contract, joint contract (summation of joint contractors is performed) Computing ability score), Pass means that you can participate in the bidding alone because you have met all the qualifications to participate in the bidding.

<Mac + click the chart, run the manual to predict the winning bid point>

Referring to FIG. 4B , the user terminal 100 clicks the [Mac + Chart] button to execute the predicted price prediction system. The predicted price prediction system is a tool that suggests one winning bid point by predicting the optimal number that can be combined with the winning bid section among multiple reserve price combinations. The user terminal 100 predicts a successful bid point by executing the manual, and each function is shown in Table 1 below.

division location Manual run Manual execution result rule One Click the section with the highest probability among 4 sections Present the highest probability winning bid section 2 Click on Mac Present 2-3 successful bid points 3 click on filter Present 1 or 2 winning bid points 4 Among the dots (↓↑), the dots marked in green Present one winning bid point final result 6 Execution of notice number 20201242334 99.2% prediction (successful bid point)

<Click the winning bid point, run the manual to calculate the final bid price>

Referring to FIG. 4C , the user terminal 100 clicks a successful bid point (99.2%) to execute the prediction system for multiple examples. The multiple example price prediction system is a tool that predicts an optimal number that can be combined with a successful bid point among 100 bid prices corresponding to the successful bid point and suggests one final bid price. The user terminal 100 calculates the final bid price by executing the manual, as shown in Table 2 below.

division location Manual run Manual execution result Predict 2 decimal places One Click on the highest probability number among the lows and highs Present 2 highest probability numbers 2 Predict 3 decimal places - Same as 2 decimal place analysis Present 3rd digit number 8 final result 3 Execution of notice number 20201242334 ￦ 25,335,897,451 won (99.2789%)

<Click to save the final bid price>

Referring to FIG. 4D , the user terminal 100 returns to the announcement list when all manual execution is completed, and transmits the final determined bid price to the bidding list. In order not to be excluded from the selection of successful bidders, 98% of the net construction cost and statutory insurance premiums (value A) are applied to the final bid price. ① Click the final bid price to show the basis for calculating the bid price, ② click the [Save] button to temporarily save the bid price, ③ check [Send] to save the bid price, and ④ If you click on the bid list, you can check the saved bid price. ⑤ If your company's qualification score is not enough in the bidding of the qualification screening system, enter the insufficient score in the input box of the lack of score supplementation item and click 'Price Compensation Before'.

The matters not described for the method of providing a big data-based probability distribution verification service for bidding in the public procurement market for bidding in the public procurement market of FIGS. The following description will be omitted because it is the same as the description of the method of providing a big data-based probability distribution verification service for bidding in the public procurement market or because it can be easily inferred from the described content.

As for the matters not described in the method of providing a big data-based probability distribution verification service for bidding in the public procurement market of FIGS. 2 and 3, the big data-based probability distribution verification service for bidding in the public procurement market through FIG. 1 above. Since it is the same as the content described with respect to the providing method or can be easily inferred from the described content, the following description will be omitted.

4 is a diagram illustrating a process in which data is transmitted/received between components included in the system for providing a big data-based probability distribution verification service for bidding in the public procurement market of FIG. 1 according to an embodiment of the present invention. Hereinafter, an example of a process in which data is transmitted/received between respective components will be described with reference to FIG. 4 , but the present application is not limited to such an embodiment, and the example shown in FIG. 4 according to the various embodiments described above will be described. It is apparent to those skilled in the art that the data transmission/reception process may be changed.

Referring to FIG. 4 , the verification service providing server 300 receives a request for medical advice from the user terminal 100 ( S4100 ), but transmits it to the manager terminal 400 , but operates the SBC in the manager terminal 400 and time By issuing and authenticating a stamp, personal information protection and tampering are prevented. Accordingly, the transmitted medical advice data is transmitted to the verification service providing server 300 (S4300), and the verification service providing server 300 collects the data accumulated as the medical history log and the information providing server 500 in this way. The process of building big data is performed using the data such as the guidelines (S4400) (S4500, S4600).

On the other hand, when a new medical advice arrives from the user terminal 100 (S4600), the verification service providing server 300 extracts an identifier by parsing the medical record data (S4700), and generates a query using the identifier (S4800) ), extracts similar medical advice data as an answer to the query (S4810), and generates medical advice data to correspond to the requester's record (S4820).

Then, the verification service providing server 300 requests authentication to the manager terminal 600 (S4830), and when receiving feedback corresponding to the request (S4840), if there is no error, it is transmitted to the user terminal 100 and (S4850), and when an error occurs, it is accumulated as data to identify the pattern in which the error occurred, and the pattern is extracted by artificial neural network deep learning (S4860, S4870), and then when a new request is received, the error is make sure it doesn't happen.

When the processing is completed, the processed medical record data and advisory data are mapped and stored (S4910), and the big data is updated by learning (S4920).

The order between the above-described steps (S4100 to S4920) is merely an example, and is not limited thereto. That is, the order between the above-described steps ( S4100 to S4920 ) may be mutually changed, and some of these steps may be simultaneously executed or deleted.

The matters that are not described above for the method of providing a big data-based probability distribution verification service for bidding in the public procurement market of FIG. 4 are big data-based probability distribution verification services for bidding in the public procurement market through FIGS. Since it is the same as the content described with respect to the providing method or it can be easily inferred from the described content, the following description will be omitted.

5 is an operation flowchart illustrating a method of providing a big data-based probability distribution verification service for bidding in a public procurement market according to an embodiment of the present invention. Referring to FIG. 5 , the verification service providing server receives the patient's medical record data transmitted from the medical advice request terminal to calculate medical expenses ( S5100 ).

In addition, the verification service providing server parses the received medical record data to extract a diagnosis name, and classifies the medical diagnosis by deriving and classifying the medical diagnosis, the medical diagnosis site, and the identifier of the detailed disease name from the diagnosis name (S5200).

In addition, the verification service providing server extracts at least one image identification information included in the received medical record data (S5300), and based on the classified identifier and the extracted image identification information, preset from the established big data Reference data having a similarity is extracted (S5400).

Then, the verification service providing server generates advisory data for the patient's medical record data by outputting the stored advisory data mapped to the extracted reference data (S5500), and transmits the generated advisory data to the medical advisory request terminal (S5600).

Matters that have not been explained about the method of providing a big data-based probability distribution verification service for bidding in the public procurement market of FIG. Since it is the same as the content described with respect to the providing method or it can be easily inferred from the described content, the following description will be omitted.

The method of providing a big data-based probability distribution verification service for bidding in the public procurement market according to an embodiment described through FIG. 5 is a record including instructions executable by a computer, such as an application or program module executed by a computer. It can also be implemented in the form of a medium. Computer-readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. Also, computer-readable media may include all computer storage media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data.

The method for providing a big data-based probability distribution verification service for bidding in the public procurement market according to an embodiment of the present invention described above includes an application basically installed in a terminal (which is a program included in a platform or operating system basically installed in the terminal) may be included), and may be executed by an application (ie, a program) installed directly on the master terminal by a user through an application providing server such as an application store server, an application, or a web server related to the corresponding service. In this sense, the method for providing a big data-based probability distribution verification service for bidding in the public procurement market according to an embodiment of the present invention described above is implemented as an application (that is, a program) installed basically in a terminal or directly installed by a user and may be recorded on a computer-readable recording medium such as a terminal.

The description of the present invention described above is for illustration, and those of ordinary skill in the art to which the present invention pertains can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a dispersed form, and likewise components described as distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalents should be interpreted as being included in the scope of the present invention. do.

Claims (5)

When a public notice number for which a bid price is to be calculated is selected from the notice list, and when participating in the public bidding of the selected notice number, at least one reservation price that can be successful is extracted and output through big data-based data analysis, and at least one a user terminal for outputting a probability distribution verification result of the at least one predetermined price based on a condition of ;
Big data-based probability distribution verification service for bidding in the public procurement market Provides any one of web page, app page, program and application, and collects data disclosed by at least one institution using a web crawler or bot In order to construct big data, data is classified and pre-processed to eliminate errors and standardize, and then use any one or a combination of big data analysis, data mining, and inverse probability model to determine the expected price. It is extracted and transmitted to the user terminal, and the probability distribution of the expected price is verified by text and graph to verify the accuracy of the probability distribution provided under different conditions, such as by ordering organization, by preliminary price range, by industry, and by participant, before and after the opening of the bid. Verification service providing server that performs; and
When constructing big data using any one of the web page, app page, program and application related to the big data-based probability distribution verification service for bidding in the public procurement market, check if there are any errors in the data, correct the data, Deletion and correction, standardization of data format, and verification of big data, inverse probability, and probability distribution output from data mining model in the verification service providing server When different conditions are set or the accuracy is automatically verified, the manager terminal checks it once again; including, but
The verification service providing server,
a collection unit for collecting at least one public procurement bidding data using a web crawler;
a loading unit for loading structured data, semi-structured and unstructured data within the collected at least one public procurement bidding data;
Pre-processing including a purification and standardization procedure corresponding to any one or at least one combination of correction, correction, and deletion in which the at least one public procurement bidding data finds errors that are inconsistent with the government bid execution standards or government contract execution standards a preprocessor that performs
an analysis unit that visualizes at least one predetermined price of a public bid selected by the user terminal as a probability distribution and transmits it to the user terminal; and
A verification unit that verifies the probability distribution based on at least one condition and transmits it to the user terminal;
The loading unit,
The structured data is stored in MSSQL (Microsoft SQL) as a relational data base management system (RDBMS) database storage method, the semi-structured data is stored in a relational database (RDB), and the unstructured data is distributed file system DFS (Distributed File System) File System), but
The at least one public procurement bid data is processed and loaded as a real-time stream,
The analysis unit,
It performs analysis based on big data, data mining and inverse probability,
The verification unit,
When the user terminal selects and bids any one of the at least one predetermined price for the public bidding, it is verified that the accuracy of the prediction results before and after the opening of the public bidding is different,
to provide a probability distribution extracted as a sample with a high prediction rate to calculate the at least one expected price,
Determination of the fitness of the probability distribution, x2-test, KS (Kolmogorov-Smirnov) test, CVM (Cramer Von-Mises) test, and Probability Plot Correlation Coefficient (PPCC) test, characterized in that made through public Big data-based probability distribution verification service provision system for bidding in the procurement market.
The method of claim 1,
The at least one condition is
Big data-based probability distribution verification service providing system for bidding in the public procurement market, characterized in that it includes at least one ordering institution, a preliminary price range, and different conditions for each industry.
delete The method of claim 1,
The verification service providing server,
An upload unit for uploading a result of verifying the probability distribution based on the result before and after the ticket wicket in at least one manager terminal, based on any one or at least one combination of at least one ordering organization, a preliminary price range, and different conditions for each industry; Big data-based probability distribution verification service provision system for bidding in the public procurement market, characterized in that it further comprises a.
The method of claim 1,
The verification service providing server,
If the prediction results before and after the open bidding of the public bidding are different, the different parts and causes are identified, the identified different parts and causes are updated in big data, and then patterns for the different parts and causes are extracted and classified through An artificial intelligence unit that performs deep learning using an artificial neural network to predict the error rate, and reflects the data derived by the deep learning to the big data;
The artificial intelligence unit uses a picture to check the range and median of the data set, and a box plot is used to identify outliers,
The deep learning uses MLP (Multi-Layer Perceptron) and ANFIS (Adaptive Neuro Fuzzy Inference System) to predict the bid price closest to the lowest successful bid price by predicting the bid price through data analysis of electronic bidding rather than time series data. Big data-based probability distribution verification service provision system for bidding in the public procurement market.

Images