An online auction system and method

Abstract

The application discloses an online auction system and method, comprising: an information publishing module for receiving auction information published by a user; an intelligent matching module for obtaining a client matching the auction information, and pushing the auction information to the matched client; an online auction module for generating quote auxiliary information according to a preset intelligent quote model, sending the quote auxiliary information to the client, and receiving quote information of the client during an auction period, and sorting and displaying the quote information; a transaction module for autonomous order transaction between the buyer and the seller after the auction period ends; and a query module for querying historical auction session information, auction results and bill information, and generating an auction statistical report. The system simplifies the transaction process, improves the transaction efficiency, and ensures convenient, safe, fast and efficient transactions, and has the advantages of convenience, fairness and transparency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet, in particular to an online auction system and method. BACKGROUND

[0002] With the enterprise production process, often due to overcapacity, market demand changes, equipment aging or obsolete factors, resulting in a large number of inventory, waste materials, unsold goods, long-term accumulation will occupy a large number of enterprise turnover funds, while increasing the cost of warehouse management. Therefore, the enterprise procurement personnel need to find sales channels to sell these goods.

[0003] Sales of enterprise waste materials and unsold goods often face many problems, for example: 1, the market demand is insufficient, and there is a lack of effective sales channels, leading to sales difficulties; 2, the enterprise cannot quickly find the buyer, and needs to invest a large number of procurement personnel to carry out offline marketing or promotion, offline sales, price negotiation, price negotiation, and the transaction process is slow and inefficient, on the other hand, it also increases the labor cost, time-consuming and laborious; 3, because the waste material has no market unified price, and the quality of the unsold goods cannot be trusted, the buyer's scattered price has no competitiveness, resulting in the need to reduce the price concession or the final transaction of part of the goods in the sales process, which cannot achieve the expected transaction.

[0004] On the other hand, with the development of society, enterprises need to continuously expand production scale, for example, in construction, machinery manufacturing, production and processing, new machines, new equipment, and a large amount of raw materials are usually purchased, and the purchase cost of bulk purchase of raw materials and new machines and equipment is high. Some enterprises will choose to purchase waste materials (such as edge cutting, head cutting, and steel materials in components) as raw materials for processing, and recycle second-hand machines and equipment to replace new equipment, thereby reducing the cost of enterprises. Enterprises often face many problems when purchasing materials: 1, due to limited purchase channels, they can only purchase in local second-hand trading markets or waste material traders, because enterprises need to configure procurement personnel to find sources in the market, increasing the procurement cost of enterprises; 2, procurement personnel need to spend a lot of time and effort to find suppliers, and carry out price negotiation, price negotiation, and price comparison offline, which is very inefficient; 3, the entire transaction process is not transparent, and each procurement personnel has different personal preferences and negotiation abilities, which makes it difficult for enterprises to manage uniformly, and is prone to gray transactions or dark box operations. SUMMARY

[0005] The embodiments of the present application provide an online auction system and method to at least solve the technical problems of low efficiency and time-consuming in offline transactions in the related art.

[0006] According to an aspect of an embodiment of the present application, an online auction system is provided, comprising:

[0007] An information publishing module is configured to receive auction information published by a user.

[0008] An intelligent matching module is configured to obtain a client matching the auction information and push the auction information to the matching client.

[0009] An online auction module is configured to generate a bidding auxiliary information according to a preset intelligent bidding model, send the bidding auxiliary information to the client, and receive bidding information of the client and sort and display the bidding information during an auction period.

[0010] A transaction module is configured to enable a buyer and a seller to independently place an order after the auction period ends.

[0011] A query module is configured to query information of a historical auction, an auction result, and a bill of lading, and generate an auction statistical report.

[0012] In an optional embodiment, the information publishing module comprises:

[0013] A selling item publishing unit is configured to publish one or more information of a category, a name, a model, a quantity, and a picture of a selling item.

[0014] A buying item publishing unit is configured to publish one or more information of a category, a name, a model, a quantity, and a picture of a buying item.

[0015] An auction rule publishing unit is configured to publish one or more auction rules of a starting price, a price increase step value, a delivery area, a delay mechanism, a payment mode, and a settlement mode.

[0016] In an optional embodiment, the intelligent matching module comprises:

[0017] An intelligent search matching unit is configured to set a multi-level search condition according to the auction information, obtain a client meeting the condition, and obtain a matching client and a push order according to benefit information, setting information, and feature information of the client meeting the condition.

[0018] A push unit is configured to push the auction information to the matching client according to the push order.

[0019] In an optional embodiment, the online auction module comprises:

[0020] An auxiliary bidding unit is configured to obtain historical bidding information of a user, market trend information, and historical transaction information of an item.

[0021] A bidding influence factor is extracted according to the historical bidding information, the market trend information, and the historical transaction information of the item.

[0022] The weight of each bid influencing factor is obtained, the bid influencing factor and the weight are input into the intelligent bid model, the intelligent bid model adopts a Kalman filtering algorithm, a bid coefficient is estimated according to the Kalman filtering algorithm, and auxiliary bid information calculated is obtained according to a product of the bid coefficient and a historical transaction price;

[0023] The bidding unit is configured to receive the bid information from the client in real time, and display the bid information in a sorted manner on the user terminals of the buyer and the seller and on a real-time broadcasting page.

[0024] In an optional embodiment, the transaction module comprises:

[0025] The auxiliary merchant selection unit is configured to obtain bid information, enterprise credit information, scale information, enterprise qualification information and historical transaction evaluation information of the client participating in the bidding;

[0026] The characteristic factors are extracted according to the bid information, the enterprise credit information, the scale information, the enterprise qualification information and the historical transaction evaluation information to form a characteristic matrix;

[0027] The characteristic matrix is input into a pre-trained intelligent merchant selection model to obtain a score of each bidding client;

[0028] The bidding clients are sorted according to the scores from large to small, and a preset number of clients in the front are recommended to the user.

[0029] In an optional embodiment, before the characteristic matrix is input into the pre-trained intelligent merchant selection model, the method further comprises:

[0030] The characteristic matrix of each client is formed according to the bid information, the enterprise credit information, the scale information, the enterprise qualification information and the historical transaction evaluation information of each client, and a score label is added to the characteristic matrix of each client;

[0031] The training data set is obtained according to the characteristic matrix of each client and the corresponding score label;

[0032] The BP neural network model is trained according to the training data set, the BP neural network model comprises an input layer, a hidden layer, an output layer and a Softmax layer connected in sequence, and a trained intelligent merchant selection model is obtained.

[0033] In an optional embodiment, the transaction module further comprises:

[0034] The online negotiation unit is configured to initiate online negotiation to all the clients participating in the bidding by one key after the bidding period ends, and synchronize negotiation information to each user terminal and a real-time broadcasting page;

[0035] A transaction unit is configured to enable the buyer and the seller to independently place orders after the bidding period ends.

[0036] In an optional embodiment, the system further comprises:

[0037] An interaction module is configured to generate a group chat tool, and the buyer and the seller communicate and ask questions according to the group chat tool.

[0038] In an optional embodiment, the query module comprises:

[0039] A query unit is configured to query information about a historical bidding session, a bidding result, and a bill of lading.

[0040] A statistics unit is configured to count information about a successful bidding enterprise, a bidding commodity, and a bidding price, and generate a bidding statistics report.

[0041] According to another aspect of the embodiments of the present application, an online bidding method is provided, which comprises:

[0042] Receiving bidding information published by a user;

[0043] Obtaining a client matching the bidding information, and pushing the bidding information to the matching client;

[0044] Generating quote assistance information according to a preset intelligent quote model, sending the quote assistance information to the client, and receiving quote information from the client during a bidding period, and displaying the quote information in a sorted manner;

[0045] Enabling the buyer and the seller to independently place orders after the bidding period ends;

[0046] Querying information about a historical bidding session, a bidding result, and a bill of lading, and generating a bidding statistics report.

[0047] The technical solutions provided by the embodiments of the present application can have the following beneficial effects:

[0048] The online bidding system provided by the present application enables the buyer and the seller to conduct transactions online, simplifies the transaction process, improves transaction efficiency, and ensures that transactions are convenient, safe, fast, and efficient, and has the advantages of convenience, fairness, and transparency.

[0049] Further, the online bidding system of the present application can update quote information in real time, automatically push bidding information, and intelligently assist in filling quotes and selecting merchants during bidding, thereby improving the accuracy and efficiency of user quotes, giving suggestions for selecting merchants based on multiple factors, and improving user experience and transaction rate. BRIEF DESCRIPTION OF DRAWINGS

[0050] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:

[0051] Figure 1 is a structural diagram of an online auction system according to an embodiment of the application;

[0052] Figure 2 is a flowchart of an online auction service according to an embodiment of the application;

[0053] Figure 3 is a flowchart of an auxiliary bidding method according to an embodiment of the application;

[0054] Figure 4 is a flowchart of an auxiliary merchant selection method according to an embodiment of the application;

[0055] Figure 5 is a schematic diagram of a BP neural network model according to an embodiment of the application;

[0056] Figure 6 is a technical architecture diagram of an interaction module according to an embodiment of the application;

[0057] Figure 7 is an auction rule publishing interface diagram provided according to an embodiment of the application;

[0058] Figure 8 is an auction result interface diagram provided according to an embodiment of the application;

[0059] Figure 9 is an auction Q&A interface diagram provided according to an embodiment of the application;

[0060] Figure 10 is a schematic diagram of a bidding coefficient acquisition method provided according to an embodiment of the application. DETAILED DESCRIPTION

[0061] In order to enable persons skilled in the art to better understand the application scheme, the technical solutions in the embodiments of the application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the application. Obviously, the described embodiments are only a part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by persons skilled in the art without creative work should fall within the protection scope of the application.

[0062] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and in the above drawings are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0063] The embodiments of the present application provide an online auction system, which can at least solve the technical problems that enterprises cannot quickly find a transaction channel in the transaction process of backlog inventory, waste materials, unsalable goods, second-hand equipment, etc., transaction efficiency is low, bidding is not competitive, and the transaction process is not transparent, can help supply and demand enterprises to quickly match demand, so that the supply and demand enterprises can quickly obtain buying and selling demand and first-hand bidding information; at the same time, it can also make multiple supply and demand enterprises online fully auction, ensure that the supply and demand parties conduct online auction in a fair, open and just environment, fair trade, so that the interests of both parties are not damaged, and the supply and demand parties can quickly and efficiently trade.

[0064] The embodiments of the present application will be described in detail below with reference to the accompanying drawings Figure 1 The online auction system of the embodiments of the present application will be described in detail as follows Figure 1 As shown in the figure, the system mainly includes:

[0065] The information publishing module is used for receiving the auction information published by the user.

[0066] In an optional embodiment, the information publishing module includes: a competitive selling commodity publishing unit, which is used for publishing one or more information of the category, name, model, quantity and picture of the commodity to be sold.

[0067] The auction system provided by the embodiments of the present application includes two auction modes, one is a competitive selling mode and the other is a competitive buying mode. In the competitive selling mode, the competitive selling commodity information can be published, and the competitive selling party publishes the competitive selling commodity information through the client or logs in the Internet platform, and the Internet platform publishes the competitive selling information to the designated interface after auditing the competitive selling information. Specifically, when the enterprise needs to sell waste materials, backlog inventory and unsalable goods, it can publish the request information of selling a certain commodity or multiple commodities in a package through the client or the Internet platform. The auction information here includes: commodity category, commodity name, model specification, quantity, picture and other information.

[0068] The auction item publishing unit is configured to publish one or more of the following information of the auction item: category, name, model, quantity, and picture.

[0069] In the auction mode, the bidder publishes the purchase request information through the client or the Internet platform, and the Internet platform publishes the real-time auction information to the designated interface after auditing. Specifically, when an enterprise needs to recycle waste materials (such as offcuts and cut heads in steel scrap), second-hand mobile phones or equipment, it can publish the request information for purchasing a certain type of goods or multiple goods in a package through the client or the Internet platform. The auction request information includes: category, name, model, quantity, picture and other information.

[0070] The auction rule publishing unit is configured to publish one or more of the following auction rules: starting price, price increase step value, delivery area, time delay mechanism, payment mode, and settlement method.

[0071] When publishing the auction item information, the auction rules are also set, such as the starting price of the whole order, the expected price of a single item (optional), the price increase step value, the delivery area, the time delay mechanism, the payment mode, and the settlement method.

[0072] When publishing the auction item information, the auction rules are also set, such as the highest acceptable bid, the delivery date, the price reduction gradient of each bid, the delivery area, the time delay mechanism, the payment mode, and the settlement method.

[0073] As shown in Figure 7 The auction information publishing interface includes auction publishing, auction rules, bid requirements, payment mode and other information. According to the business needs, the bidder selects the auction mode (price increase or price decrease), and fills in the auction material information (such as auction items, model specifications, quantity, material, measurement method, etc.), sets the auction rules, starting price, auction time, time delay mechanism, price step value, payment mode, settlement method, and other information, and publishes the auction after completing the filling.

[0074] The system includes an intelligent matching module for obtaining a client matching the auction information and pushing the auction information to the matched client.

[0075] The online auction system provided by the embodiments of the present application can intelligently identify the client matching the auction information after obtaining the published auction information, and intelligently push the auction information.

[0076] In an alternative embodiment, the intelligent matching module comprises: an intelligent search matching unit, configured to set multi-level search conditions according to the auction information, obtain the client meeting the conditions, and obtain the matched client and the push sequence according to the right information, setting information and feature information of the client meeting the conditions; and a push unit, configured to push the auction information to the matched client according to the push sequence.

[0077] Specifically, the search conditions are determined according to the auction information. First, the demand goods in the auction information are determined, and the target demander / supplier is matched according to the multi-level category corresponding to the demand goods.

[0078] For example, the demander / supplier publishes the auction information, which contains the goods (threaded steel HRB400E-25), and the background three-level category of the goods is (building steel- building materials-threaded steel). According to the background three-level category, the target demander / supplier meeting the demander's condition can be screened, including A, Hangxiao Steel Structure Co., Ltd., B, Henan X Steel Co., Ltd., C, Shandong Zhi X Trading Co., Ltd., and D, Tianjin Long X Building Material Co., Ltd.

[0079] Further, the demander / supplier right information in the auction information is determined, the target enterprise is matched according to the demander / supplier right information, and the inquiry request information is sent to the client of the target demander / supplier.

[0080] For example, the demander / supplier right information in the auction information is determined, the target enterprise is matched according to the demander / supplier right information, and the inquiry request information is sent to the client of the target demander / supplier.

[0081] Further, the demander / supplier right information in the auction information is determined, the target enterprise is matched according to the demander / supplier right information, and the inquiry request information is sent to the client of the target demander / supplier.

[0082] Further, the demander / supplier right information in the auction information is determined, the target enterprise is matched according to the demander / supplier right information, and the inquiry request information is sent to the client of the target demander / supplier.

[0083] Further, according to the setting conditions of the system platform and / or the demander and / or the supplier user, the target enterprise matching the bidding information is determined, for example, the B enterprise is set to the do-not-disturb mode by itself, at this time, the system platform cannot push the bidding information, and the bidding information is sent to the target enterprise to the client of the A, C and D enterprises; the C enterprise sets not to receive the inquiry information in the terminal, at this time, the system platform will not push to the target terminal; when the D enterprise sets the A enterprise as the non-cooperative purchaser in the target interruption, the D enterprise will also not receive the inquiry information.

[0084] Further, according to the characteristic information marked by the demander / supplier user in the system platform, the target demander / supplier matching the bidding information is determined, and the bidding information is sent to the client of the target demander / supplier.

[0085] For example, the platform combines the random forest algorithm according to the demander inquiry category and the receiving area, the Tianyancha risk, the refund rate, the historical transaction record, the platform activity degree (the platform access times, the commodity category inquiry registration rate, the bidding rate and the like) and the random feature selection calculation, and finally only the A company meets the pushing condition, the bidding information is sent to the client of the target supplier A company, so that the corresponding bidding information of the demander / supplier can be accurately pushed.

[0086] In the embodiment of the application, the server stores the inquiry sheets corresponding to all the bidding information to a preset delay queue according to the session time corresponding to each bidding information, wherein the delay queue is used to consume the inquiry sheets in the delay queue in time sequence and based on an optimistic lock mechanism. That is, the server determines the corresponding relationship between the current time and the above session, and when the session time reaches the corresponding moment, the demander user is pushed one or more bidding commodity information of the merchant.

[0087] In a single or multiple instruction requests, the server cannot normally open the session due to network interruption, server exception and the like in a preset commodity display time period, including that the session is not normally opened within the predetermined time, the system alarm is used to monitor the abnormal opening, the platform can separately open the session, and the timer can scan the abnormal session.

[0088] In an optional embodiment, the bidding registration module is further included, in the bidding mode, the bidder independently registers, and when the number of the demander registrations is less than the minimum registration number, the auction is automatically closed.

[0089] Specifically, the demander can view the real-time bidding information published by the supplier through the client or login the Internet platform, the demander registers after opening the right after the demander according to the enterprise self-purchasing demand, if it is set that at least 3 bidders participate in the current bidding, less than 3 before the specified registration deadline, the auction is automatically closed. Through this mechanism, the supplier obtains more reasonable and competitive bidding.

[0090] In the bidding mode, the bidding party independently registers, and the auction is automatically closed when the number of supplier registrations is less than the minimum number of registrations. Specifically, the bidding party can view real-time bidding information published by the demand party through a client or by logging in to an Internet platform, and the bidding party independently registers.

[0091] The online bidding module is also included, which is used to generate bidding assistance information according to a preset intelligent bidding model, send the bidding assistance information to the client, and receive the bidding information from the client and sort and display the bidding information during the bidding period.

[0092] After the bidding party or the bidding party successfully registers, they can jointly participate in the bidding during the bidding start and end time period.

[0093] The embodiment of the application provides an auxiliary bidding unit, which intelligently assists in filling the bid by analyzing user historical bidding behavior, current market information and commodity historical transaction information, and gives appropriate bidding suggestions. This can improve the accuracy and efficiency of user bidding.

[0094] Figure 3 The flowchart of an auxiliary bidding method according to an embodiment of the application is shown in FIG. 1. Figure 3 The auxiliary bidding method includes the following steps.

[0095] S301, obtaining user historical bidding information, market information and commodity historical transaction information;

[0096] S302, extracting bidding influence factors according to the historical bidding information, market information and commodity historical transaction information;

[0097] S303, constructing an intelligent bidding model;

[0098] S304, inputting the bidding influence factors into the intelligent bidding model to obtain calculated auxiliary bidding information;

[0099] S305, pushing the auxiliary bidding information to the client.

[0100] In an exemplary scenario, the commodity that the buyer needs to purchase is steel, and the seller needs to provide the selling price information. In order to improve the accuracy and efficiency of user bidding, the bidding suggestion can be given through the intelligent bidding model.

[0101] Specifically, the current market information of steel, historical quotation information of similar goods of the user and historical transaction information of the goods are pulled, and quotation influencing factors are extracted, the quotation influencing factors including the selling price of steel in the current market, historical quotation of similar goods and historical transaction price, weights of the quotation influencing factors are obtained, the quotation influencing factors and the weights are input into an intelligent quotation model, the intelligent quotation model adopts Kalman filtering algorithm, a quotation coefficient is estimated according to the Kalman filtering algorithm, auxiliary quotation information calculated according to the product of the quotation coefficient and the historical transaction price is obtained, and the weights can be preset according to actual conditions.

[0102] The Kalman filtering algorithm dynamically estimates the quotation coefficient according to the following formula

[0103] K k is an optimal gain matrix;

[0104] P k / k-1 is an estimation error matrix;

[0105]

[0106] Ф k+1,k is a state matrix, H k is an observation matrix, and Г k+1,k is a coefficient matrix of noise, and R and Q are hyperparameters, which are set according to actual experience.

[0107] The Kalman filtering algorithm receives historical price data as input measurement values, and obtains the accuracy of the measurement values. With the measurement values, the algorithm estimates the real output of the system according to the measurement values, and at the same time gives the accuracy range of the newly estimated quotation value. Then the Kalman filtering estimates the state of the system by using the model of the system and the observation. This process is continuously performed, that is, the system is continuously measured and estimated, and after a period of time, a very accurate quotation coefficient of the system can be estimated. As shown in Figure 10 , input quotation data and hyperparameters, continuously predict and correct to obtain an optimal quotation coefficient.

[0108] The auction unit is further included, which is used for receiving quotation information of the client in real time, and displaying the quotation information in order on the user terminals of the buyer and the seller and a real-time broadcast page.

[0109] In an exemplary scenario, the buyer can freely bid (enter the total bid amount of the entire order, the unit price information of each item, the note content, etc.) within the auction start and end time period. The buyer can bid multiple times within the number of bids set by the bidding party. Each bid amount must be higher than the previous bid amount, and each bid amount of the bidding party must be higher than the highest bid amount of all current bidding parties. Optionally, when the bidding party specifies the bid step amount, the bidding party must bid according to the bid step amount specified by the bidding party. The bid information of the buyer is displayed in descending order and is updated in real time.

[0110] After the bidding party submits a bid, the bidding party can view the current leading bid amount and the corresponding enterprise profile, as well as the bid amounts of other bidding parties and the corresponding rankings. All bidding parties are desensitized data information, i.e., the bidding party can only see the bid information of the bidding party and cannot see the enterprise name of the bidding party, to avoid the bidding party directly contacting the bidding party and engaging in unfair transactions. Optionally, when the bidding party sets the public bidding party bid, each bid amount of the bidding party is publicly visible to other bidding parties participating in the current competition. When the bidding party sets the bid not to be public, the bidding party can only view its own bid and the ranking corresponding to the bid amount of its own enterprise. Optionally, when the bidding party sets the delay rule, the bidding party can bid before the extended countdown ends, until no one continues to bid, the auction time no longer continues to be automatically delayed, and the countdown automatically ends.

[0111] In an exemplary scenario, the seller can freely bid (enter the total bid amount of the entire order, the unit price information of each item, the note content, etc.) within the auction start and end time period. The seller can bid multiple times within the number of bids set by the bidding party. Each bid amount must be lower than the previous bid amount, and each bid amount of the bidding party must be lower than the lowest bid amount of all current bidding parties. Optionally, when the bidding party specifies the bid step amount, the bidding party must bid according to the bid step amount specified by the bidding party. The bid information of the seller is displayed in descending order and is updated in real time.

[0112] After the bidding party submits a bid, the bidding party can view the current leading bid amount and the corresponding enterprise profile, as well as the bid amounts of other bidding parties and the corresponding rankings. All bidding parties are desensitized data information, i.e., the bidding party can only see the bid information of the bidding party and cannot see the enterprise name of the bidding party, to avoid the bidding party directly contacting the bidding party and engaging in unfair transactions. Optionally, when the bidding party sets the delay rule, the bidding party can bid before the extended countdown ends, until no one continues to bid, the auction time no longer continues to be automatically delayed, and the countdown automatically ends.

[0113] Figure 8It is a bidding interface schematic diagram, which can display bidding start and end time information, can display bidding commodity information, current bidding price information, bidding requirement information, etc. Enterprises can independently register according to demand, bidding materials, bidding requirements, etc.

[0114] After the bidding time starts, the platform will publicly disclose the bidding process. Optionally, other users of the platform can enter the Internet platform, client, mobile terminal to observe the whole bidding process, ensuring that the bidding process of the supply and demand parties is completely open and transparent.

[0115] It also includes a transaction module for the buyer and seller to independently place orders for transactions after the bidding period ends.

[0116] The transaction module includes an auxiliary merchant selection unit that can give the best merchant selection suggestion based on current market trends, user qualifications, performance capabilities, and other information.

[0117] Figure 4 It is a flowchart of an auxiliary merchant selection method according to an embodiment of the present application, which includes:

[0118] S401 obtains the bid information, enterprise credit information, scale information, enterprise qualification information, and historical transaction evaluation information of the client participating in the bidding;

[0119] S402 extracts feature factors from the bid information, enterprise credit information, scale information, enterprise qualification information, and historical transaction evaluation information to form a feature matrix;

[0120] S403 obtains a training data set according to the feature matrix and corresponding score labels of multiple clients;

[0121] S404 trains a BP neural network model according to the training data set to obtain a trained intelligent merchant selection model;

[0122] S405 inputs the feature matrix into the pre-trained intelligent merchant selection model to obtain the score of each bidding client;

[0123] S406 sorts the bidding clients according to the order of the scores from large to small and recommends the top pre-set number of clients to the user.

[0124] Specifically, first, the intelligent business selection model provided in the embodiments of the present application is trained. The bidding client's bid information, enterprise credit information, scale information, enterprise qualification information, and historical transaction evaluation information are obtained. According to the bid information, enterprise credit information, scale information, enterprise qualification information, and historical transaction evaluation information of each client, feature data is extracted, the enterprise credit, scale, evaluation, and the like are converted into different dimension score levels, a feature matrix of each client is formed, and a score label is added to the feature matrix of each client. According to the feature matrix of each client and the corresponding score label, a training data set is obtained.

[0125] According to the training data set, a BP neural network model is trained to obtain a trained intelligent business selection model. As shown in the formula (1), the BP neural network model used in the embodiments of the present application includes sequentially connected input layer, hidden layer, output layer, and Softmax layer. The BP network can learn and store a large number of input-output mode mapping relationships without needing to describe the mathematical equation of the mapping relationship in advance. Figure 5

[0126] In the model training process, the mean square error function can be used as the loss function, and the gradient descent algorithm is used on all parameters, so that the loss function of the neural network model on the training data set reaches a smaller value. The optimization process includes: first, define the neural network structure, calculate the predicted value by the forward propagation algorithm, define the loss function, calculate the difference between the predicted value and the true value, select the back propagation optimization algorithm, calculate the gradient of the loss function to each parameter, update each parameter using the gradient descent algorithm according to the gradient and the learning rate, repeatedly run the back propagation algorithm on the training data set, and train the neural network.

[0127] Further, the bid information, enterprise credit information, scale information, enterprise qualification information, and historical transaction evaluation information of the client participating in the bidding are obtained. According to the bid information, enterprise credit information, scale information, enterprise qualification information, and historical transaction evaluation information, feature factors are extracted to form a feature matrix. The feature matrix is input into the pre-trained intelligent business selection model to obtain the score of each bidding client. According to the order from large to small of the score, the bidding clients are sorted, and the top pre-set number of clients are recommended to the user.

[0128] In an exemplary scenario, the buyer user can select the seller of his heart according to the ranking of the seller recommended by the system to place an order transaction. Through intelligent auxiliary business selection, the user no longer needs to carefully check the enterprise bid, credit, qualification, and the like of each bidding party. The system will automatically assist in selecting the business.

[0129] ​In an alternative embodiment, the transaction module further comprises: an online negotiation unit configured to initiate online negotiation with all participating clients after the end of the auction period, and to synchronize negotiation information to each user terminal and real-time broadcast page.

[0130] Due to the complexity of B-end bulk commodity transactions, enterprise procurement personnel often use different strategies to repeatedly negotiate with suppliers after receiving the quotation information in order to reduce procurement costs. The negotiation is not a one-time purchase, and only a part of the suppliers may be negotiated during the negotiation process, resulting in some unfair situations in transactions. To effectively solve such problems, an online negotiation tool is provided.

[0131] As an alternative, after the end of the auction time, the system compares the prices of each commodity from each seller, helps the buyer to identify the lowest price of each commodity, and automatically calculates the lowest average price and initiates negotiation to all sellers. After receiving the negotiation notice, the sellers can choose to accept the negotiation, continue to reduce the price, or refuse the negotiation. This allows enterprise procurement personnel to purchase the required commodities at a low price.

[0132] For example, A enterprise has a total price of 1 million, ranking first, and B enterprise has a total price of 1.1 million, falling behind A enterprise's price ranking. However, compared to individual commodities, B enterprise's price for some commodities is significantly higher, affecting the final price ranking and falling behind A enterprise. However, B enterprise's other commodity prices are actually more advantageous than A enterprise's prices. By identifying the lowest price of each commodity, the sellers can fully and fairly bid, and the buyers can effectively negotiate, thereby reducing enterprise procurement costs and protecting the interests of both buyers and sellers.

[0133] The transaction unit is also included for self-ordering transactions between buyers and sellers after the end of the auction period.

[0134] In an exemplary scenario, the seller selects one intended buyer to place an order for the entire transaction. The system automatically calculates the transaction service fee based on the seller's membership level and the order amount.

[0135] For example, the user's membership level is obtained, the service fee payment ratio for that level is queried, and the transaction service fee is automatically calculated based on the corresponding order amount and payment ratio. A payment interface is generated, and the buyer receives an order notification after the seller's transaction service fee is fully paid. The order can be fulfilled online. Alternatively, when the seller publishes a real-time auction, the buyer can download the standard contract template provided by the platform to sign the order offline.

[0136] In an alternative embodiment, further comprising: an interactive module for generating a group chat tool, and the bidders and the sellers communicate according to the group chat tool.

[0137] During the entire auction process, the bidders and the sellers can consult or answer questions about the auction item in the interactive communication tool, and the communication tool is in a group chat mode. The answer record generated by each auction is real-time open to all users in the real-time auction, that is, the bidders can view the questions and answers at any time. The enterprise names of the bidders and the sellers are desensitized data information. The method adopted by the application can quickly match all the supply and demand transaction information on the platform, and the bidders can conduct centralized bidding, so as to realize an open, transparent, fair and just auction environment.

[0138] As shown in Figure 9 , in the auction process interface, the left side can display the auction item information, the bidder information and the auction price, and the right side can enter the interactive Q&A group chat, and the bidders can communicate in the group chat. Questions can be raised to the bidders through the interactive Q&A tool, and the bidders can answer the questions. The interactive Q&A is in a group chat mode, and all bidders can see it. The bidding unit information is desensitized information. When any party initiates information containing sensitive words or price information, the system will automatically shield the sensitive content.

[0139] Figure 6 is a technical architecture diagram of an interactive module according to an embodiment of the application, as shown in Figure 6 , the user can log in on a mobile terminal, a computer terminal or the like to conduct online auction, realize group chat through deployment of a gateway service cluster, a chat service cluster, a WebSocKet cluster and a message cluster, synchronize the group chat message to each user terminal, and store the chat data in a MySQL cluster.

[0140] Further comprising a query module for querying historical auction session information, auction results and bill information, and generating an auction statistical report. The query module comprises: a query unit for querying historical auction session information, auction results and bill information. A statistical unit for statistics of successful auction enterprise information, auction item information and auction price information, and generation of an auction statistical report.

[0141] The online auction system provided by the embodiments of the present application ensures the real-time nature of the bidding information based on a real-time pushing component, and transmits the changes in the bidding information to the supply-demand side user terminals and a real-time broadcasting page in real time according to the auction settings, thereby ensuring the symmetry and transparency of the information of the transaction parties and avoiding the possibility of dark box operation in the transaction process. It is mainly used for the real-time viewing of the changed data of the other party by the negotiation parties during the business operation process. A real-time computing engine is introduced, the original Apache Flink and websocket features are utilized, the duplex communication principle is combined, and the front-end and back-end long connections are established to realize real-time interaction of data. In this way, the real-time nature and accuracy of data transmission can be improved, and the optimal price in the transaction project is searched. The bidding auxiliary system based on artificial intelligence intelligently assists in filling the bid by analyzing the historical bidding behavior of the user, the current market situation and the user preference settings, and gives appropriate bidding suggestions. In this way, the accuracy and efficiency of the bidding of the supply side user can be improved. The intelligent computing tool can assist the demand side in quickly calculating the lowest average bid, and give the best merchant selection suggestion in combination with the current market trend, the user qualification and the performance capability. The hot data is entrusted to the cache manager, the hot and cold data is separated, the frequently accessed auction business data is entrusted to the cache manager for management, and the millisecond level data read-write performance is provided. In this way, the jitter rate and the packet loss rate of the system can be reduced, and the timeliness of the response to the user terminal request can be ensured. The cloud service is used throughout, the entire negotiation link system is deployed on the cloud platform, and cloud publishing and cloud deployment are realized. In this way, the stability and reliability and the operation efficiency of the system can be improved. Meanwhile, the cloud platform also has the elastic expansion function, the system resources can be dynamically adjusted according to the actual demand, the response capability and the throughput of the system are improved, and the stability of the negotiation link system is ensured. In addition, the front-end and the back-end use the encryption protocol communication, the data is transmitted and interacted through the encryption protocol, the data transmission safety is ensured; the pre-compiled or parameterized query is used, the statement parameter special character check and verification are performed on the database, the database execution statement is prevented from being tampered and injected, the database data is encrypted, the negotiation link is entirely processed by the program, no human intervention is performed, and the data safety is ensured; the sensitive data in the transaction process is encrypted by using the SHA-256 algorithm, the data confidentiality is effectively protected, and the data of the user is prevented from being stolen or leaked.

[0142] According to another aspect of the embodiments of the present application, an online auction method is provided, which comprises: receiving the auction information published by a user; acquiring the client matched with the auction information, and pushing the auction information to the matched client; generating bidding auxiliary information according to a preset intelligent bidding model, sending the bidding auxiliary information to the client, and receiving the bidding information of the client during the auction period, and sorting and displaying the bidding information; after the auction period ends, the buyer and the seller independently place orders for transaction; querying the session information, the auction result and the bill of lading information of the historical auction, and generating an auction statistical report.

[0143] Figure 2 is a flowchart of an online auction service according to an embodiment of the present application, as shown in Figure 2 The auction service flow includes:

[0144] First, a user publishes an auction request information through a PC client or a mobile terminal, including an auction category (waste materials, excess inventory), auction start and end time, starting price, and other requirements; the platform publishes the information to a specific interface after passing the audit.

[0145] Further, the user registers according to the auction requirements; optionally, if there is a registration benefit fee, the platform guarantee fund needs to be paid in full at the same time.

[0146] Further, after the user successfully registers, he or she can participate in the bidding (whole order bidding) within the auction start and end time period; the server sends the bid information of each supplier to the user interface or client of the publisher of the auction demand in real time, and the server automatically ranks according to the total bid amount. (Competitive selling mode: the higher the price, the higher the ranking; competitive buying mode: the lower the price, the higher the ranking).

[0147] Finally, after the bidding ends, the user can choose one of the bidding parties to place an order transaction, and the supplier can view the order and perform order fulfillment after paying the transaction service fee.

[0148] Compared with the traditional bargaining mode, the buyer and seller transact offline, and there is too much offline human intervention, which is not conducive to the fairness and transparency of the transaction link and the safety of the information. The project uses Internet front-end technology to reform the traditional bargaining process, simplify the transaction link, improve transaction efficiency, and ensure that the transaction is convenient, safe, fast, and efficient to promote the transaction between the buyer and seller, and solve the problems of long transaction cycle, complex process, and too many dark box operations in the traditional bargaining process.

[0149] The technical features of the above embodiments can be combined in any way. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0150] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the present application. It should be noted that for those skilled in the art, without departing from the concept of the present application, some modifications and improvements can be made, which are all within the scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. An online auction system, characterized by, Comprise: Information publishing module, for receiving user published bidding information; Intelligent matching module, for obtaining the client matching the bidding information, and pushing the bidding information to the matched client; comprising: determining the retrieval condition according to the bidding information, first determining the demand goods in the bidding information, and matching the target demander / supplier according to the multi-level category corresponding to the demand goods; determining the consignee address condition in the bidding information, matching the target enterprise according to the consignee address information, and sending the inquiry request information to the client of the target demander / supplier; determining the demander / supplier right information in the bidding information, screening the target enterprise according to the right information, and sending the bidding information to the client of the target enterprise; according to the characteristic information marked by the demander / supplier user in the system platform, determining the target demander / supplier matching the bidding information, and sending the bidding information to the client of the target demander / supplier; Online bidding module, for generating price auxiliary information according to the preset intelligent pricing model, sending the price auxiliary information to the client, and receiving the price information of the client within the bidding period, and sorting and displaying the price information; the online bidding module comprises: an auxiliary pricing unit for obtaining user historical pricing information, market quotation information and commodity historical transaction information; extracting pricing influence factors according to the historical pricing information, market quotation information and commodity historical transaction information; obtaining the weight of each pricing influence factor, inputting the pricing influence factor and weight into the intelligent pricing model, the intelligent pricing model adopts Kalman filtering algorithm, estimates the pricing coefficient according to the Kalman filtering algorithm, and obtains the calculated auxiliary pricing information according to the product of the pricing coefficient and the historical transaction price; a bidding unit for receiving the price information of the client in real time, sorting and displaying the price information on the user terminal of the buyer and seller and the real-time broadcast page; the Kalman filtering algorithm receives historical price data as input measurement value, obtains the accuracy of the measurement value; with this measurement value, the algorithm estimates the true output of the system according to the measurement value, and at the same time gives the accuracy range of the new estimated price value; then Kalman filtering uses the model and observation of the system to estimate the state of the system; this process is continuously carried out, that is, the system is continuously measured and estimated, and after a period of time, a pricing coefficient of the system can be estimated; Transaction module, for buying and selling parties to carry out self-ordering transaction after the bidding period ends; the transaction module comprises: an auxiliary selection unit for obtaining the price information, enterprise credit information, scale information, enterprise qualification information and historical transaction evaluation information of the client participating in the bidding; extracting feature factors according to the price information, enterprise credit information, scale information, enterprise qualification information and historical transaction evaluation information to form a feature matrix; inputting the feature matrix into the pre-trained intelligent selection model to obtain the score of each bidding client; according to the order from large to small of the score, the bidding clients are sorted, and the user is recommended to the first preset number of clients; the intelligent selection model is a BP neural network model; Before inputting the pre-trained intelligent merchant selection model, further comprising: according to the bid information, enterprise credit information, scale information, enterprise qualification information and historical transaction evaluation information of each client, the feature matrix of each client is formed, and a score label is added to the feature matrix of each client; according to the feature matrix of each client and the corresponding score label, a training data set is obtained; a BP neural network model is trained according to the training data set, the BP neural network model includes sequentially connected input layer, hidden layer, output layer and Softmax layer, and a trained intelligent merchant selection model is obtained; The query module is used for querying the session information, bidding result and bill of lading information of the historical bidding, and generating a bidding statistical report; The interactive module is used for generating a group chat tool, and the bidders and the sellers communicate according to the group chat tool; the bidders can communicate in the group chat, can initiate a question to the bidders through the interactive question answering tool, and the bidders can answer; the interactive question answering is in the group chat mode, and all bidders can see; the group chat is realized through the deployment of gateway service cluster, chat service cluster, WebSocKet cluster and message cluster, and the group chat messages are synchronized to each user terminal, and the chat data is stored through MySQL cluster.

2. The system of claim 1, wherein, The information publishing module comprises: The selling commodity publishing unit is used for publishing one or more information of the category, name, model, quantity and picture of the goods to be sold; The buying commodity publishing unit is used for publishing one or more information of the category, name, model, quantity and picture of the goods to be bought; The bidding rule publishing unit is used for publishing one or more bidding rules of the starting price, price increase step value, delivery area, delay mechanism, payment mode and settlement mode.

3. The system of claim 1, wherein, The intelligent matching module comprises: The intelligent search matching unit is used for setting multi-level search conditions according to the bidding information, obtaining the client meeting the conditions, and obtaining the matched client and the push sequence according to the interest information, setting information and feature information of the client meeting the conditions; The push unit is used for pushing the bidding information to the matched client according to the push sequence.

4. The system of claim 1, wherein, The transaction module further comprises: The online negotiation unit is used for initiating online negotiation to all the clients participating in the bidding after the bidding period ends, synchronizing the negotiation information to each user terminal and a real-time broadcast page, and broadcasting the negotiation information in real time; The transaction unit is used for automatically calculating the transaction service fee after the bidding period ends, enabling the user to pay the transaction service fee online, independently placing an order and storing the transaction data through SHA-256 encryption algorithm.

5. The system of claim 1, wherein, The query module comprises: The query unit is used for querying the session information, bidding result and bill of lading information of the historical bidding; The statistical unit is used for counting the information of the two parties of the successful bidding, the bidding commodity information and the bidding price information, and generating a bidding statistical report.

6. An online auction method, characterized by, It comprises: Receiving the bidding information published by the user; The method comprises the following steps: acquiring the client matching the auction information, and pushing the auction information to the matched client; determining the retrieval condition according to the auction information, first determining the demand commodity in the auction information, and matching the target demander / supplier according to the multi-level category corresponding to the demand commodity; determining the consignee address condition in the auction information, matching the target enterprise according to the consignee address information, and sending the inquiry request information to the client of the target demander / supplier; determining the demander / supplier right information in the auction information, screening the target enterprise according to the right information, and sending the auction information to the client of the target enterprise; determining the target demander / supplier matching the auction information according to the characteristic information marked by the demander / supplier user in the system platform, and sending the auction information to the client of the target demander / supplier; According to the preset intelligent bidding model, the bidding auxiliary information is generated, the bidding auxiliary information is sent to the client, and the bidding information of the client is received and sorted and displayed within the bidding period; the method comprises the following steps: acquiring the user historical bidding information, market quotation information and commodity historical transaction information; extracting the bidding influence factor according to the historical bidding information, market quotation information and commodity historical transaction information; acquiring the weight of each bidding influence factor, inputting the bidding influence factor and weight into the intelligent bidding model, the intelligent bidding model adopts the Kalman filtering algorithm, estimating the bidding coefficient according to the Kalman filtering algorithm, and obtaining the calculated auxiliary bidding information according to the product of the bidding coefficient and the historical transaction price; receiving the bidding information of the client in real time, and sorting and displaying the bidding information on the user terminal of the buyer and seller and the real-time broadcast page; the historical price data is received as the input measurement value by using the Kalman filtering algorithm, and the accuracy of the measurement value is obtained; with the measurement value, the algorithm estimates the real output of the system according to the measurement value, and at the same time gives the accuracy range of the new estimated bidding value; then the Kalman filtering uses the model and observation of the system to estimate the state of the system; this process is continuously carried out, that is, the system is continuously measured and estimated, and after a period of time, a bidding coefficient of the system can be estimated; Further comprising: acquiring the bidding information, enterprise credit information, scale information, enterprise qualification information and historical transaction evaluation information of the client participating in the auction; extracting the characteristic factor according to the bidding information, enterprise credit information, scale information, enterprise qualification information and historical transaction evaluation information, and forming a characteristic matrix; inputting the characteristic matrix into the pre-trained intelligent merchant selection model to obtain the score of each auction client; sorting the auction clients according to the score from large to small, and recommending the first preset number of clients to the user; the intelligent merchant selection model is a BP neural network model. Before inputting the pre-trained intelligent business selection model, further comprising: according to the quotation information, enterprise credit information, scale information, enterprise qualification information and historical transaction evaluation information of each client, the feature matrix of each client is formed, and a score label is added to the feature matrix of each client; according to the feature matrix and the corresponding score label of each client, a training data set is obtained; a BP neural network model is trained according to the training data set, the BP neural network model comprises an input layer, a hidden layer, an output layer and a Softmax layer connected in sequence, and a trained intelligent business selection model is obtained; After the bidding period ends, the buyer and the seller conduct self-ordering transactions; Inquiring the session information, bidding results and bill information of the historical bidding, and generating a bidding statistical report; An interactive module is used to generate a group chat tool, the bidders and the sellers communicate and answer questions according to the group chat tool, the bidders can communicate in the group chat, can initiate questions to the bidders through the interactive question answering tool, and the bidders can answer the questions, the interactive question answering is in the group chat mode, and all the bidders can see; the group chat is realized by deploying a gateway service cluster, a chat service cluster, a WebSocKet cluster and a message cluster, and the group chat messages are synchronized to each user terminal, and the chat data is stored in a MySQL cluster.

Images