A waste resource trading brokerage platform system using ai-based demand forecasting and market price information

Abstract

The present invention relates to a waste resource trading brokerage platform system utilizing AI-based demand forecasting and market price information, which can improve the stability and efficiency of the waste resource trading market by connecting waste resource generators and collectors in a mobile environment and providing standard trading prices and collection matching through AI-based demand forecasting and market price calculation. The present invention provides market price information for waste resources and new resources, stores transaction data for waste resources and new resources, analyzes market price information and transaction data using AI to forecast the demand for waste resources and new resources, and calculates the market prices of waste resources and new resources based on the forecast results to broker transactions.

Description

Technology Field

[0001] The present invention relates to a waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information, and more specifically, to a waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information that can improve the stability and efficiency of the waste resource trading market by connecting waste resource generators and collectors in a mobile environment and providing a standard price for trading and collection matching through artificial intelligence-based demand forecasting and market price calculation. Background Technology

[0002] Recently, the demand for bioenergy raw materials has been rapidly increasing due to the commercialization of bio-marine fuel, the consequent expansion of bio-heavy fuel usage, and the introduction of policies mandating the use of sustainable aviation fuel. In particular, waste cooking oil is recognized as a strategic resource as it can be utilized as a key raw material for biodiesel and bio-aviation fuel; however, a system for securing a stable supply of these materials has not yet been sufficiently established.

[0003] The existing waste cooking oil collection and distribution structure has operated using analog methods, primarily relying on telephone calls, visits, and manual ledger management. This has resulted in persistent issues such as price opacity, inefficient collection schedules, and information asymmetry between generators and collectors.

[0004] In particular, waste cooking oil generators find it difficult to check market price information in real time, and collectors also find it difficult to systematically predict demand fluctuations by region and time of day, leading to problems where transaction prices are unilaterally determined or trading opportunities are lost.

[0005] Due to these structural problems, the waste cooking oil trading market was characterized by high price volatility and low transaction stability, which limited the establishment of a long-term raw material supply chain and negatively impacted the overall growth of the bioenergy industry.

[0006] Meanwhile, although advancements in artificial intelligence and mobile platform technologies have enabled demand forecasting, price analysis, and logistics optimization based on large-scale transaction data, instances of their systematic application to the waste resource trading market have been limited.

[0007] Furthermore, as logistics movements, transaction history, and price information generated during the waste resource trading process were not digitized, there was a lack of a foundation to maintain the overall positive functions of the market and support fair price formation. Prior art literature

[0008] Korean Registered Patent No. 10-2199852 (Registered on January 4, 2021) The problem to be solved

[0009] Therefore, the present invention is derived to solve the aforementioned problems, and its purpose is to provide a waste resource trading brokerage platform system utilizing AI-based demand forecasting and market price information that can improve the stability and efficiency of the waste resource trading market by connecting waste resource generators and collectors in a mobile environment and providing standard trading prices and collection matching through AI-based demand forecasting and market price calculation.

[0010] In addition, another objective of the present invention is to provide a waste resource trading brokerage platform system utilizing AI-based demand forecasting and market price information, which can increase the efficiency of collection routes and shorten transaction processing times by integrating transaction matching and logistics management functions into the platform, thereby enabling the revitalization of the waste resource trading market and the continuous generation of revenue and profit.

[0011] Other objects of the present invention will become more apparent through the embodiments described below. means of solving the problem

[0012] A waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information according to one embodiment of the present invention may include: a seller terminal that registers waste resource sales information by inputting the type, quantity, and disposal location of waste resources and registers new resource purchase information by inputting the item and quantity of new resources; a collector terminal that registers waste resource collection information for collecting waste resources based on waste resource sales information and registers new resource delivery information for delivering new resources; and a waste resource trading brokerage server that provides waste resource market price information and new resource market price information, stores waste resource trading data and new resource trading data, predicts the demand for waste resources and new resources by analyzing waste resource market price information, new resource market price information, waste resource trading data, and new resource trading data using artificial intelligence, calculates the market price of waste resources and new resources based on the prediction results, and brokers the transaction.

[0013] In addition, the waste resource trading brokerage server collects waste resource trading data including waste resource sales information and waste resource collection information, and new resource trading data including new resource purchase information and new resource delivery information, and includes a data collection unit that collects and stores waste resource market price information and new resource market price information; a demand forecasting and price calculation unit that predicts the demand for waste resources and new resources by analyzing past trading patterns and current market prices based on waste resource trading data, new resource trading data, waste resource market price information, and new resource market price information, and calculates predicted waste resource market prices and predicted new resource market prices that serve as the basis for trading based on the prediction results; a trading brokerage unit that automatically matches the optimal collector or allows the seller to select a collector by considering the location, rating, and transaction history of the seller and the collector based on the predicted waste resource market prices and predicted new resource market prices, and mediates the trading and settlement of waste resources and new resources; a logistics management unit that groups and recommends sales routes, purchase routes, collection routes, and delivery routes based on at least one of waste resource trading data, new resource trading data, hub logistics bases, and transport vehicles, and waste resource transaction completion It may include a settlement and donation linkage unit that settles brokerage fees incurred at the time, donates at least a portion of the brokerage fees to a pre-registered donation organization, and manages the donation history and transaction history in a linked manner.

[0014] In addition, the demand forecasting and price calculation unit may be characterized by synthesizing the waste resource trading volume, weights based on the timing of the transaction, and the new resource price volatility coefficient, and calculating a two-way raw material-linked demand pressure index that indicates how much pressure there is on resource supply and demand according to the trading activity in the waste resource market and purchasing pressure in the new resource market by considering the new resource purchase volume and new resource trading reliability, and calculating the new resource trading reliability by reflecting the transaction success rate indicator, delivery performance stability indicator, transaction repeatability indicator, transaction dispute occurrence indicator, and transaction evaluation score indicator.

[0015] In addition, the demand forecasting and price calculation unit calculates a Predicted Price Confidence Collapse Sensitivity Index, which indicates how easily the predicted market price can collapse, by reflecting the duration of predicted price maintenance, the range of predicted price fluctuation, the time interval of change, and the market response sensitivity coefficient. The duration of predicted price maintenance is calculated by setting the time when the waste resource predicted market price or new resource predicted market price was calculated as the reference point, establishing a certain observation period after the reference point, and calculating the ratio or duration of actual transaction prices that occurred during that period that fall within the allowable error range relative to the predicted market price, and normalizing the ratio or duration of transactions. The range of predicted price fluctuation is calculated by setting the waste resource predicted market price or new resource predicted market price at the reference point as the reference predicted price, setting the maximum or minimum value among subsequently updated waste resource predicted market prices or new resource predicted market prices as the updated predicted price, and calculating based on the difference between the reference predicted price and the updated predicted price. The time interval of change records the initial time of calculation when the waste resource predicted market price or new resource predicted market price was calculated, and when the predicted market price is significant It may be characterized by recording the time of the changed market price change and calculating based on the time difference between the initial calculation time and the time of the market price change, and for the market reaction sensitivity coefficient, setting the period before the price change and the period after the price change based on the point in time when a change occurred in the waste resource prediction market price or the new resource prediction market price, and calculating the rate of increase / decrease in trading volume, the rate of increase / decrease in transaction cancellation rate, and user reaction indicators by comparing the period before the price change and the period after the price change, and calculating the market reaction level by combining the calculated rate of increase / decrease in trading volume, the rate of increase / decrease in transaction cancellation rate, and user reaction indicators according to pre-set weights, and normalizing the market reaction level to a reference range.

[0016] In addition, the demand forecasting and price calculation unit calculates a market price signal locking coefficient that indicates how strongly actual completed transactions depend on waste resource market prices or new resource market prices by reflecting the number of platform price-referenced transactions, external price-referenced transactions, and individual negotiation transactions; the number of platform price-referenced transactions is calculated by aggregating the number of transactions completed without the trading participant modifying the presented transaction price after the waste resource prediction market price, new resource prediction market price, waste resource standard market price, or new resource standard market price was presented as the transaction price; the number of external price-referenced transactions is calculated by identifying transactions where the trading participant indicated whether to reference an external price or entered external price information during the process of modifying the transaction price, and aggregating the number of transactions when the transaction was finally concluded; and the number of individual negotiation transactions is calculated by identifying transactions where the transaction price was modified multiple times without information on waste resource prediction market prices, new resource prediction market prices, waste resource standard market prices, new resource standard market prices, or external price reference prices, and where there is a record of direct negotiation between trading participants, and aggregating the number of transactions when the transaction was finally concluded. It can be characterized as a feature.

[0017] In addition, the settlement and donation linkage unit may be characterized by calculating a social return sustainability resonance index that indicates the degree to which social return activities do not remain a one-time occurrence but are linked to transaction activities, sustaining over time and resonating repeatedly, by reflecting the period-specific donation amount linked to transaction completion, transaction linkage strength, and time duration weighting based on time; the period-specific donation amount is calculated by extracting the amount for which actual execution was completed among the donation records that occurred during the donation calculation period; the transaction linkage strength is calculated by classifying donation records according to whether they are linked to a transaction, assigning linkage weighting based on whether the donation is automatically generated upon transaction completion or is unrelated to the transaction, and synthesizing the ratio and strength of transaction-linked donations; and the time duration weighting is calculated by setting a reference point and determining whether the reference point is a period during which donations occurred continuously or a period during which there was a gap in donations. Effects of the invention

[0018] A waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information according to an embodiment of the present invention provides the following effects.

[0019] The present invention has the effect of improving the stability and efficiency of the waste resource trading market by connecting waste resource generators and collectors in a mobile environment and providing standard transaction prices and collection matching through AI-based demand forecasting and market price calculation.

[0020] By integrating transaction matching and logistics management functions into a platform, the present invention can increase the efficiency of collection routes and shorten transaction processing times, thereby enabling the revitalization of the waste resource trading market and the continuous generation of revenue and profit.

[0021] The effects of the present invention are not limited to those mentioned above, and other unmentioned effects will be clearly understood by those skilled in the art from the description below. Brief explanation of the drawing

[0022] FIG. 1 is a schematic diagram illustrating a waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information according to one embodiment of the present invention. FIG. 2 is a block diagram illustrating a waste resource trading brokerage server of a waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information according to an embodiment of the present invention. Specific details for implementing the invention

[0023] Hereinafter, various embodiments are described in more detail with reference to the attached drawings. The embodiments described in this specification may be modified in various ways. Specific embodiments may be depicted in the drawings and described in detail in the detailed description. However, specific embodiments disclosed in the attached drawings are intended only to facilitate understanding of various embodiments. Accordingly, the technical concept is not limited by specific embodiments disclosed in the attached drawings, and it should be understood that it includes all equivalents or substitutions that fall within the spirit and scope of the invention.

[0024] Terms including ordinal numbers, such as first, second, etc., may be used to describe various components, but these components are not limited by the aforementioned terms. The aforementioned terms are used solely for the purpose of distinguishing one component from another.

[0025] Functions related to artificial intelligence according to the present disclosure are operated through a processor and memory. The processor may be composed of one or more processors. In this case, the one or more processors may be general-purpose processors such as CPUs, APs, and DSPs (Digital Signal Processors), graphics-dedicated processors such as GPUs and VPUs (Vision Processing Units), or artificial intelligence-dedicated processors such as NPUs. The one or more processors control the processing of input data according to predefined operation rules or artificial intelligence models stored in memory. Alternatively, if the one or more processors are artificial intelligence-dedicated processors, the artificial intelligence-dedicated processors may be designed with a hardware structure specialized for processing a specific artificial intelligence model.

[0026] The predefined rules of operation or artificial intelligence models are characterized by being created through learning. Here, being created through learning means that a basic artificial intelligence model is trained using a number of training data by a learning algorithm, thereby creating predefined rules of operation or artificial intelligence models configured to perform desired characteristics (or objectives). Such learning may be performed on the device itself where the artificial intelligence according to the present disclosure is executed, or it may be performed through a separate server and / or system. Examples of learning algorithms include supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning, but are not limited to the examples described above.

[0027] An artificial intelligence model may be composed of multiple neural network layers. Each of the multiple neural network layers has multiple nodes and weight values, and performs neural network operations through operations between the results of previous layers and the multiple weights. The multiple weights possessed by the multiple neural network layers can be optimized based on the learning results of the artificial intelligence model. For example, multiple weights may be updated so that the loss value or cost value obtained by the artificial intelligence model during the learning process is reduced or minimized. Additionally, to minimize the loss value or cost value, multiple weights may be updated in a direction that minimizes the gradient associated with the loss value or cost value. Examples of artificial neural networks include, but are not limited to, CNNs (Convolutional Neural Networks), DNNs (Deep Neural Networks), RNNs (Recurrent Neural Networks), RBMs (Restricted Boltzmann Machines), DBNs (Deep Belief Networks), BRDNNs (Bidirectional Recurrent Deep Neural Networks), or Deep Q-Networks.

[0028] A network is a network that serves as a transmission path for web pages; it may be a closed network such as a LAN (Local Area Network) or WAN (Wide Area Network), but it is desirable for it to be an open network such as the Internet. The Internet refers to a global open computer network structure that provides the TCP / IP protocol and various services existing at its upper layers, namely HTTP (Hyper Text Transfer Protocol), Telnet, FTP (File Transfer Protocol), DNS (Domain Name System), SMTP (Simple Mail Transfer Protocol), SNMP (Simple Network Management Protocol), NFS (Network File Service), and NIS (Network Information Service).

[0029] The terminal can be implemented in various forms. For example, the terminal described in this specification may be a mobile terminal such as a smartphone, tablet PC, PDA (Personal Digital Assistant), PMP (Portable Multimedia Player), or MP3 player, as well as a fixed terminal such as a smart TV or desktop computer.

[0030] In this specification, terms such as “comprising” or “having” are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof. When a component is described as being “connected” or “connected” to another component, it should be understood that it may be directly connected to or connected to that other component, or that there may be other components in between. On the other hand, when a component is described as being “directly connected” or “directly connected” to another component, it should be understood that there are no other components in between.

[0031] Meanwhile, a "module" or "part" for a component as used in this specification performs at least one function or operation. Furthermore, a "module" or "part" may perform a function or operation by hardware, software, or a combination of hardware and software. Additionally, a plurality of "modules" or a plurality of "parts," excluding a "module" or "part" that must be performed on specific hardware or on at least one processor, may be integrated into at least one module. A singular expression includes a plural expression unless the context clearly indicates otherwise.

[0032] In addition, power, power transmission, and control therefor for the following assembly configurations and embodiments, including "by control," follow conventional technology including terminals, applications, hardware control modules, etc., so they are omitted to avoid redundancy.

[0033] In addition, the operation embodiments and configurations described in a general manner without being explained in detail below follow the prior art and are omitted in order to focus on describing the purpose of the present invention and the resulting effects.

[0034] Furthermore, in describing the present invention, if it is determined that a detailed description of related known functions or configurations may unnecessarily obscure the essence of the invention, such detailed description is abbreviated or omitted.

[0035] A waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information according to embodiments of the present invention will be described in detail below with reference to FIGS. 1 and 2.

[0036] FIG. 1 is a schematic diagram illustrating a waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information according to one embodiment of the present invention.

[0037] Referring to FIG. 1, a waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information may include a seller terminal (100), a collector terminal (200), and a waste resource trading brokerage server (300).

[0038] A seller terminal (100) is a user terminal used by an individual, business operator, or institution that generates or possesses waste resources to participate in a waste resource sales transaction, and can be implemented in various forms such as a smartphone, tablet, PC, or industrial terminal. The seller terminal (100) includes an input interface, an information verification interface, and a transaction progress interface for the transaction of waste resources.

[0039] A seller terminal (100) receives and registers waste resource sales information including the type, quantity, and discharge location of the waste resource. At this time, the type of waste resource can be subdivided into waste oil including waste cooking oil, biodiesel, bio-heavy oil, sustainable aviation fuel (SAF), and hydro-treated vegetable oil (HVO), metal, plastic, battery, electronic scrap, construction waste, etc., and the quantity is configured to be entered based on at least one of weight, volume, and number. The discharge location can be entered using address information, coordinate information, or a map-based selection method.

[0040] Additionally, the seller terminal (100) is configured to register new resource purchase information by inputting the item and quantity of new resources in conjunction with the sale of waste resources. Accordingly, the seller can plan to purchase new resources simultaneously with the sale of waste resources, and the waste resource transaction brokerage server (300) utilizes this linked information for transaction optimization and market price calculation.

[0041] The seller terminal (100) receives and displays waste resource market price information, predicted price information, collector information, and expected profit information provided by the waste resource trading brokerage server (300) in real time or periodically, and through this, the seller can actively determine the timing of the transaction, whether to select a collector, and transaction conditions. In addition, the seller terminal (100) is configured to allow viewing of transaction history, settlement details, and donation linkage details, enabling transparent transaction management.

[0042] The collector terminal (200) is a user terminal used by a collector, transportation business operator, or logistics business operator that collects waste resources or delivers new resources, and can be implemented in various forms such as a smartphone, tablet, PC, or industrial terminal. The collector terminal (200) performs the function of registering waste resource collection information for the collection of waste resources and new resource delivery information for the delivery of new resources, checking transaction matching results, and managing the progress status of collection and delivery.

[0043] The collector terminal (200) inputs and registers waste resource collection information, including whether waste resources can be collected, collection schedule, and collection quantity, based on waste resource sales information registered from the seller terminal (100). In this process, the collector may also register information such as the type of vehicle owned, loading capacity, working hours, and types of waste resources that can be processed.

[0044] In addition, the collector terminal (200) can register new resource delivery information including the delivery items, quantity, delivery schedule, and delivery location of the new resources, thereby forming a transaction structure that combines waste resource collection and new resource delivery.

[0045] The collector terminal (200) can determine whether to participate in a transaction by receiving predicted market prices, recommended transactions, logistics routes, and expected profit information provided by the transaction brokerage server (300), and is configured to allow for rating assignment, transaction history management, and settlement result verification after the transaction is completed.

[0046] The waste resource trading brokerage server (300) is a central processing system that communicates with the seller terminal (100) and the collector terminal (200) to broker the overall trading of waste resources and new resources. The waste resource trading brokerage server (300) collects and stores waste resource market price information and new resource market price information from external market data, public data, affiliated institution data, etc., and systematically manages all transaction data generated during the trading process. In addition, by performing demand forecasting and price calculation through artificial intelligence-based analysis, it aims to go beyond simple brokerage to promote market stabilization and transaction efficiency.

[0047] That is, the waste resource trading brokerage server (300) provides waste resource market price information and new resource market price information, stores waste resource trading data and new resource trading data, analyzes waste resource market price information, new resource market price information, waste resource trading data and new resource trading data using artificial intelligence to predict the demand for waste resources and new resources, and calculates the market price of waste resources and new resources based on the prediction results to broker the transaction.

[0048] FIG. 2 is a block diagram illustrating a waste resource trading brokerage server of a waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information according to an embodiment of the present invention.

[0049] Referring to FIG. 2, the waste resource trading brokerage server (300) may include a data collection unit (310), a demand forecasting and market price calculation unit (320), a trading brokerage unit (330), a logistics management unit (340), and a settlement and donation linkage unit (350).

[0050] The data collection unit (310) can systematically collect, store, and manage various information generated throughout the entire process of trading waste resources and new resources. The data collection unit (310) is responsible for providing reliable and structured data so that the artificial intelligence analysis of the demand forecasting and market price calculation unit (320), which will be described later, can be performed accurately.

[0051] The data collection unit (310) can collect waste resource sales information input from the seller terminal (100) and waste resource collection information input from the collector terminal (200). The waste resource sales information may include the type, grade, quantity, disposal location, disposal time, desired transaction conditions, etc. of the waste resource. The waste resource collection information may include collectible items, collectible quantity, collection time, transportation means information, etc. The waste resource sales information and waste resource collection information may be classified and stored as waste resource transaction data.

[0052] The data collection unit (310) can collect new resource purchase information input from the seller terminal (100) and new resource delivery information input from the collector terminal (200). The new resource purchase information and new resource delivery information can be classified and stored as new resource transaction data. The new resource transaction data includes the item, specifications, quantity, delivery location, desired delivery schedule, and transaction conditions of the new resource to be purchased or delivered, and is stored in a structure that can be linked with waste resource transaction data. Through this, waste resource transactions and new resource transactions can be analyzed as a single transaction flow or a linked transaction flow.

[0053] The data collection unit (310) periodically or in real time collects and stores information on market prices of waste resources and new resources from external systems, public databases, affiliated market information providers, etc. Market price information for waste resources and new resources is managed by classifying it by region, item, and time period, and is configured to enable trend analysis by linking past price data with current price data. Accordingly, a data foundation is established that can reflect not only short-term volatility but also medium-to-long-term market trends.

[0054] The demand forecasting and market price calculation unit (320) performs artificial intelligence analysis based on waste resource trading data, new resource trading data, waste resource market price information, and new resource market price information stored in the data collection unit (310), predicts the demand for waste resources and new resources, and calculates the standard trading price based on this.

[0055] The demand forecasting and market price calculation unit (320) converts waste resource trading data, new resource trading data, waste resource market price information, and new resource market price information into an analyzable form by normalizing them according to temporal and spatial criteria. At this time, the data is compartmentalized according to the time of transaction occurrence, transaction region, and item classification criteria to prevent the analysis of transaction data under different conditions from being mixed.

[0056] The demand forecasting and market price calculation unit (320) analyzes the frequency of transactions and changes in transaction volume by item and region based on the history of transactions of waste resources and new resources over a certain period in the past. The demand forecasting and market price calculation unit (320) can derive a potential trend of increasing or decreasing demand for a specific item by determining whether the frequency of transactions of a specific item is continuously increasing or decreasing, or whether the fluctuation range of the transaction volume is expanding, during the analysis process.

[0057] The demand forecasting and market price calculation unit (320) analyzes the phenomenon of concentrated trading in a specific region, the shortage of collectible resources, and the possibility of supply-demand imbalance by taking into account that the degree of trading concentration varies by region even for the same item. For example, if the demand forecasting and market price calculation unit (320) repeatedly receives requests for the trading of waste resources of the same item in a specific region but the actual trading success rate is low, this can be used as an indicator that demand in that region exceeds supply.

[0058] The demand forecasting and market price calculation unit (320) may consider the success rate of past transactions and the history of transaction delays. If the ratio of actual transactions to transaction requests is low or the time required to complete a transaction is prolonged, the demand forecasting and market price calculation unit (320) determines that there is high demand pressure for the item or region or that logistics and collection conditions are unfavorable, and reflects this in a direction of adjusting the demand level upward.

[0059] Additionally, the demand forecasting and market price calculation unit (320) may consider the current market price fluctuation trends, seasonal factors, and transaction characteristics that occur repeatedly during specific periods. For example, the demand forecasting and market price calculation unit (320) reflects the possibility of a short-term oversupply during periods when the amount of specific waste resources generated increases rapidly due to increased seasonal production activities, and conversely, considers the possibility that the demand for waste resource recovery will also increase in tandem when the demand for new resources increases rapidly during specific periods. These factors are combined in a weighted manner to determine the level of demand for waste resources and new resources at specific times and in specific regions.

[0060] The demand forecasting and price calculation unit (320) utilizes the demand levels of waste resources and new resources as key input elements for calculating the standard market price for trading. At this time, the demand forecasting results are not used alone, but are applied in a manner that compares and corrects with the current market price and past price trends.

[0061] Specifically, the demand forecasting and price calculation unit (320) sets the average market price over a past period and the real-time or recent price at the current time as reference values, and then adjusts the reference values ​​upward or downward according to the demand forecasting results. For example, the demand forecasting and price calculation unit (320) is set to adjust the reference price upward within a certain range if it is determined that there is a high probability that demand for a specific item will exceed supply.

[0062] The demand forecasting and price calculation unit (320) can adjust the price by taking into account the logistics costs incurred for collection and delivery. That is, even for the same item, in the case of areas where the collection distance, accessibility, and difficulty of deploying transport vehicles are high, an adjusted price reflecting the logistics burden can be applied to induce the completion of transactions.

[0063] In addition, the demand forecasting and price calculation unit (320) can reflect the success rate of past transactions and the possibility of delays in the price calculation. For items or regions where transaction delays or cancellations occurred frequently in the past, the demand forecasting and price calculation unit (320) may conservatively adjust the price to ensure transaction stability or reset the price in a direction to increase the possibility of transaction completion.

[0064] These multiple factors are not simply applied as a single batch average, but are categorized into elements for determining supply and demand balance, logistics burden, and transaction stability, and are reflected in stages. The resulting predicted market prices for waste resources and new resources are utilized as standard market prices provided to trading participants, and actual transactions can be brokered to take place within a reasonable range centered on these standard prices.

[0065] In this way, the demand forecasting and price calculation unit (320) can calculate demand-based predicted prices by analyzing and combining transaction data, market price information, logistics conditions, and transaction stability factors step by step, moving away from a simple price provision method. Accordingly, transaction participants can mitigate the risk associated with rapid price fluctuations, and the waste resource trading brokerage server (300) can provide a stable and predictable trading environment that reflects regional and time-specific characteristics.

[0066] The demand forecasting and price calculation unit (320) can calculate a bidirectional raw material-linked demand pressure index by combining the waste resource trading volume, weights based on the trading time, and the new resource price volatility coefficient, and by considering the new resource purchase volume and the reliability of the new resource trading. The bidirectional raw material-linked demand pressure index is an indicator for quantitatively determining how much pressure there is on resource supply and demand for a specific time or a specific resource group by simultaneously considering the trading activity in the waste resource market and the purchasing pressure in the new resource market. More specifically, the bidirectional raw material-linked demand pressure index is configured to be calculated by integrating the recent trading volume and its temporal importance in the waste resource trading market, the actual purchasing volume and the reliability of the transaction in the new resource market, and external market instability factors due to the volatility of the new resource price into a single value.

[0067] Conventional methods for determining resource supply and demand were mostly based solely on the trading volume of waste resources or analyzed only the price or purchase volume of new resources in a fragmentary manner. As a result, there was a problem where an irrational situation occurred in which the purchase of new resources surged despite waste resources being sufficiently traded in the actual market, or where an excessive judgment of supply and demand crisis occurred due to a temporary decrease in trading.

[0068] Accordingly, the present invention enables a more precise assessment of the actual resource supply and demand pressure status, which could not be identified by a single indicator alone, by simultaneously reflecting both the supply-side flow of waste resources and the demand-side flow of new resources in both directions. Through this, technical effects such as the automatic induction of waste resource priority utilization policies, adjustment of new resource purchase timing, and the enhancement of price prediction and transaction recommendation logic can be provided.

[0069] The bidirectional raw material linkage demand pressure index is [ It can be calculated based on the formula.

[0070] The transaction volume of waste resources refers to the physical quantity or weight of specific waste resources actually traded through the platform, using a value determined upon transaction completion. In this context, the transaction volume applies only to waste resources for which a transaction has been established, collection completed, and settlement processing has been processed, rather than simply the registered quantity.

[0071] The demand forecasting and market price calculation unit (320) can determine the waste resource trading volume by calculating the waste resource trading volume based on the quantity of waste resources in the waste resource sales information included in the waste resource trading data, or by calculating the actual processing volume of waste resource trading based on the waste resource trading data and the new resource trading data.

[0072] More specifically, when the demand forecasting and market price calculation unit (320) calculates the volume of waste resource transactions based on waste resource transaction data and new resource transaction data, once a waste resource transaction is completed, it first sets the theoretical processing range for the quantity of waste resources in the transaction based on the waste resource transaction data corresponding to the transaction. At the same time, the demand forecasting and market price calculation unit (320) can refer to the purchase quantity and actual delivery quantity of new resources in the transaction based on the new resource transaction data generated in conjunction with the waste resource transaction. By comparing and linking the waste resource transaction data and the new resource transaction data, the demand forecasting and market price calculation unit (320) can calculate the scale of waste resources actually processed and input into the resource circulation process and determine this as the actual volume of waste resource transactions in the transaction. For example, the volume of waste resource transactions can be adjusted by reflecting the ratio of raw materials secured through waste resource transactions during a specific period that were converted into new resource deliveries.

[0073] The transaction time weight is a time-based weighting factor designed to give greater influence to recent transactions by reflecting how close a waste resource transaction is to the current time. The transaction time weight can be calculated by normalizing the time difference between the time of collection completion of each waste resource transaction and the reference time for index calculation, assigning a larger weight value as the time difference decreases. The weight for older transactions gradually decreases, and transactions exceeding a certain reference period can be set to have very low influence in the index calculation. For example, a transaction within 1 day can be set to a transaction time weight of 1.0, a transaction past 7 days to 0.7, and a transaction past 30 days to 0.3.

[0074] The new resource price volatility coefficient is a coefficient that reflects the impact of market instability on demand pressure by reflecting how significantly new resource prices fluctuate within a short period. The new resource price volatility coefficient can be calculated by analyzing the fluctuation range or rate of change of new resource prices collected during a base period, such that demand pressure expands as volatility increases. For example, the new resource price volatility coefficient can be set to 0.05 in a price stability range where the fluctuation range or rate of change of new resource prices is small, 0.15 in a range where the fluctuation range or rate of change of new resource prices is medium, and 0.3 in a range where the fluctuation range or rate of change of new resource prices is rapid.

[0075] The new resource purchase quantity refers to the physical quantity or weight of the new resource for which the actual purchase has been confirmed. The confirmed delivery quantity may be calculated as the new resource purchase quantity only when a new resource purchase contract has been concluded and the completion of delivery has been confirmed.

[0076] New resource transaction reliability is an indicator designed to quantitatively evaluate whether a specific new resource purchase transaction is a one-time occurrence or a continuous and stable repeatable transaction. Rather than being calculated based on a single factor, reliability is derived by extracting multiple reliability indicators derived from transaction history and evaluating them comprehensively. This approach enables the calculation of reliability that reflects objective transaction behaviors accumulated over a long period, without relying on the counterparty's temporary performance or subjective evaluations.

[0077] The demand forecasting and market price calculation unit (320) can calculate the reliability of new resource transactions by reflecting the transaction success rate indicator, delivery performance stability indicator, transaction repeatability indicator, transaction dispute occurrence indicator and transaction evaluation score indicator.

[0078] Specifically, the demand forecasting and market price calculation unit (320) can calculate a transaction success rate indicator. The transaction success rate indicator refers to the ratio of new resource purchase requests registered in the past by a specific new resource buyer or trading counterparty that have actually been confirmed as completed transactions, and this quantitatively reflects the likelihood that a contract will actually be executed after a transaction request. The higher the transaction success rate, the higher the transaction reliability of the corresponding trading counterparty is judged to be. Additionally, the demand forecasting and market price calculation unit (320) can calculate a delivery performance stability indicator. The delivery performance stability indicator is an indicator that evaluates whether delivery has been successfully completed according to agreed delivery conditions after a new resource purchase contract has been concluded, and is analyzed by including the frequency of delivery delays, delivery cancellations, or changes in conditions. The delivery performance stability indicator serves as a key factor reflecting contract compliance ability and operational stability.

[0079] Additionally, the demand forecasting and market price calculation unit (320) can calculate a transaction repeatability indicator. The transaction repeatability indicator is calculated based on whether new resource transactions with the same trading counterparty have occurred repeatedly over a certain period, and the cumulative number of transactions, time intervals between transactions, and the ratio of repeated transactions may be considered. If the transaction repeatability is high, it is determined that a long-term trust relationship has been established compared to a one-time transaction. The demand forecasting and market price calculation unit (320) can calculate a transaction dispute occurrence indicator. The transaction dispute occurrence indicator is calculated by analyzing the history of disputes, claims, settlement delays, or contract terminations that occurred during the new resource transaction process, and the frequency of dispute occurrence and the time required to resolve disputes are reflected. The transaction dispute occurrence indicator can be used as a deduction factor that directly lowers reliability. Finally, the demand forecasting and market price calculation unit (320) can calculate a transaction evaluation score indicator. The transaction evaluation score indicator is calculated based on ratings or feedback information given by the trading counterparty after the transaction is completed, and time weighting may be applied so that the evaluation results for recent transactions have a greater influence. This ensures that the latest transaction quality is properly reflected in the reliability calculation.

[0080] The demand forecasting and market price calculation unit (320) can finally calculate the reliability of a new resource transaction for a specific new resource transaction or a transaction counterparty by normalizing multiple calculated indicators so that they can be compared with each other, and then combining them comprehensively by assigning weights according to the importance of each indicator.

[0081] For example, if a specific new resource buyer has a total of 50 new resource purchase requests registered over the past 12 months, and 45 of these have been confirmed as completed transactions, the transaction success rate indicator can be evaluated as "High" because the ratio of completed transactions to total purchase requests is very high. Additionally, if, among the 45 completed transactions, 42 were delivered normally within the agreed delivery deadline, 3 experienced minor delivery delays, and there is no history of delivery cancellation or breach of contract, the delivery performance stability indicator can be evaluated as "Above Medium" or "High" because it is determined that most contract conditions were stably fulfilled. Furthermore, if there are three or more instances where a new resource buyer has repeatedly transacted with the same counterparty five or more times during the same analysis period, and repeated transactions account for more than half of the total transactions, the transaction repeatability indicator can be assessed as "Above Medium" because the continuity of the trading relationship is confirmed. If the total number of transaction disputes or claims that occurred during the analysis period is only two, and all such disputes were resolved amicably within a short period, the risk in the transaction process is judged to be relatively low, and the transaction dispute occurrence indicator may be evaluated as "low" or "minor." In addition, if the average transaction evaluation score given by the counterparty after the completion of the transaction is 4.6 out of 5 points, and the average evaluation score of the last 10 transactions is 4.8 points, showing an upward trend, the satisfaction with the transaction quality is judged to be high, and the transaction evaluation score indicator may be calculated as "high."

[0082] In this way, when the transaction success rate indicator is evaluated as “high,” the delivery performance stability indicator as “medium-high or higher,” the transaction repeatability indicator as “medium-high,” the transaction dispute occurrence indicator as “low,” and the transaction evaluation score indicator as “high,” the demand forecasting and market price calculation unit (320) can comprehensively reflect the indicators to finally calculate the reliability of the new resource purchase transaction as “high reliability level.” If the reliability of the new resource transaction is high, it is converted into a score of 0.8 or higher and 1 or lower; if the reliability of the new resource transaction is average, it is converted into a score of 0.5 or higher and less than 0.8; and if the reliability of the new resource transaction is low, it is converted into a score of 0 or higher and less than 0.5.

[0083] The division prevention constant is a minimum correction value to ensure that the exponent calculation is performed stably even when the amount of new resource purchases is temporarily non-existent or extremely small. The division prevention constant can be set to a predetermined fixed value to prevent the denominator from becoming zero even when the sum of the new resource purchases is zero or below a threshold value. For example, the division prevention constant can be set to 0.01 or 0.001.

[0084] Various examples of calculating the bidirectional raw material linkage demand pressure index are as follows.

[0085] (Division prevention constant 0.01)

[0086] Examples Waste resource trading volume Transaction timing weight Coefficient of variability New resource purchase volume Reliability of new resource trading Two-way raw material linkage demand pressure index Example 1 1,000 kg 1.0 0.95 900 kg 85 78 Example 2 1,500 kg 1.2 1.10 1,200 kg 82 92 Example 3 800 kg 0.9 1.25 1,100 kg 75 88 Example 4 1,800 kg 1.3 0.90 1,100 kg 90 95 Example 5 600 kg 0.8 1.40 700 kg 60 65

[0087] In Example 1, when the volume of waste resource trading and the volume of new resource purchases are at similar levels, and when market price volatility is low and transaction reliability is high, the bidirectional raw material linkage demand pressure index is calculated to be at a medium-to-high level. This indicates normal transaction pressure in a state of supply and demand equilibrium. In Example 2, when both the volume of waste resource trading and the volume of new resource purchases increase, and a high weighting is applied at the time of transaction, the bidirectional raw material linkage demand pressure index rises to 92. This reflects a state where recycled raw material turnover is fast and market liquidity is high, suggesting that it is a period where the need for price increases or transaction priority matching is high.

[0088] In Example 3, when the volume of waste resource trading is low but the volume of new resource purchases is relatively high and market price volatility is high, the bidirectional raw material linkage demand pressure index is calculated to be 88. This quantitatively reflects a state of excess demand relative to supply shortage and indicates a situation where policies to induce the securing of waste resources or price adjustments are necessary.

[0089] In Example 4, when the volume of waste resource trading is very high and both the trading time weight and the reliability of new resource trading are high, the bidirectional raw material linkage demand pressure index reaches its highest level at 95. This serves as a key trading segment in the market and can be utilized as a basis for platform-level priority trading allocation, automatic contracting, and the application of price premiums.

[0090] In Example 5, when both the volume of waste resource trading and the volume of new resource purchases are low, and while market price volatility is high, transaction reliability is low, the bidirectional raw material linkage demand pressure index is limited to 65. This reflects a state of high market uncertainty and low transaction stability, and signifies a range where transaction restrictions or additional verification procedures are required.

[0091] The Two-Way Raw Material-Linked Demand Pressure Index is a valid indicator based on the structure of actual transaction data, rather than an arbitrarily generated formula. It calculates the trading volume of waste resources based on actual transaction records, determines the purchase volume of new resources based on actual order and contract data, calculates transaction timing weights by reflecting time-based market volatility, uses price volatility coefficients derived from period-specific price change rates, and calculates transaction reliability based on the accumulated results of past transaction history.

[0092] The bidirectional raw material-linked demand pressure index can overcome the limitations of unilateral demand forecasting by simultaneously reflecting data from both the supply side of waste resources and the demand side of new resources. By calculating this index, real-time demand pressure in the waste and new resource markets can be quantified; it can provide a basis for automating price calculation, transaction prioritization, and supply inducement policies; it enables the implementation of a market interpretation model based on linked transactions rather than simple statistics; and it secures a self-validation structure in which the reliability of the index is strengthened as data accumulates.

[0093] The selection of individual variables for calculating the bidirectional raw material linkage demand pressure index described above has the following specific background and reasons, which supports the fact that the present invention is a systematic approach rather than a simple combination of data.

[0094] Waste Resource Trading Volume: In the waste resource market, it is difficult to accurately reflect the actual scale of resources entering the market based solely on the number of transactions or registered quantities. Even with the same number of transactions, the actual amount of waste resources processed can vary significantly, which directly affects market supply pressure. Therefore, this invention aims to quantitatively reflect the actual scale of waste resources entering the market by using the trading volume of waste resources for which transactions were actually completed as a variable. Since the waste resource trading volume is calculated based on actual transaction completion data rather than mere expressions of intent, it offers high reliability from the supply side. This allows for the accurate reflection of the intensity of waste resource supply and the degree of market pressure, and provides the effect of preventing market distortion compared to existing systems that treat small and large transactions the same.

[0095] Transaction Time Weighting: The supply and demand structures of waste and new resource markets fluctuate rapidly over time, and historical transaction data has limitations in explaining current market conditions. Accordingly, this invention applies weighting based on the transaction time, configured so that recent transactions have a greater influence on the calculation of market demand pressure. This allows for the reflection of temporal proximity to the current time point, even for the same transaction volume. Through transaction time weighting, the index is prevented from becoming excessively dependent on historical data, and a dynamic demand pressure index that sensitively reflects current market conditions can be provided.

[0096] New Resource Price Volatility Coefficient: The waste resource market is strongly linked to the new resource market, and when new resource prices fluctuate rapidly, the demand for replacement waste resources tends to increase or decrease sharply. Accordingly, this invention aims to quantitatively express the phenomenon where demand pressure for waste resources intensifies as new resource prices become more unstable by reflecting the degree of fluctuation in new resource prices in the form of a coefficient. Through this, unlike simple trading volume-based indices, it is possible to reflect the impact of market instability on the demand for waste resources, enabling more conservative and stable price forecasting during periods of rapid price changes.

[0097] New Resource Purchase Volume: Supply pressure on waste resources takes on completely different meanings depending on whether there is actual demand that can be absorbed in the new resource market. It is difficult to determine the actual capacity to absorb demand based solely on simple demand intentions or registration information. Accordingly, this invention uses the transaction volume in which actual new resource purchases occurred as a variable to reflect the actual scale of demand capable of absorbing the supply of waste resources in the market. Through this, it is possible to more accurately determine whether there is an oversupply or shortage of waste resources, and to interpret waste resource trading and new resource trading as an interconnected resource circulation structure rather than as disconnected markets.

[0098] New Resource Transaction Reliability: Even with the same purchase volume of new resources, whether the transaction is a stable one that is repeatedly executed or an unstable one prone to frequent cancellations and disputes has a significant impact on market analysis results. Accordingly, this invention uses new resource transaction reliability, calculated by synthesizing the transaction success rate, delivery stability, transaction repeatability, occurrence of disputes, and transaction evaluation scores, as a variable. This allows for the reflection of both the quantity and quality of demand. It prevents errors where low-reliability demand excessively alleviates market pressure, and simultaneously improves prediction accuracy and transaction stability by calculating the index based solely on substantially stable demand.

[0099] The demand forecasting and price calculation unit (320) can calculate a prediction price confidence collapse sensitivity index that indicates how easily the predicted market price can collapse by reflecting the duration of the prediction price maintenance, the range of the prediction price change, the time interval of change, and the market reaction sensitivity coefficient. The prediction price confidence collapse sensitivity index is an indicator for quantitatively evaluating how quickly and significantly the calculated prediction market price loses confidence in the market. More specifically, the prediction price confidence collapse sensitivity index is calculated by comprehensively considering whether the prediction market price was maintained for a certain period, how wide the prediction price fluctuated within a short period, how quickly the fluctuation occurred, and how sensitively the market reacted to price changes. In other words, the prediction price confidence collapse sensitivity index is an indicator that indicates the sensitivity regarding the possibility of collapse and the speed of collapse of the prediction price.

[0100] Existing price prediction systems have been evaluated based solely on prediction accuracy or average error; however, in actual trading environments, problems exist such as predicted prices fluctuating rapidly after only a short period, predictions quickly losing the confidence of market participants, and market reactions being excessively amplified even by the same price movement. These situations are difficult to explain using simple prediction error indicators alone.

[0101] Accordingly, the present invention introduces a prediction price confidence collapse sensitivity index as an indicator capable of explaining when, to what extent, how quickly, and how sensitively prediction market prices collapse. Through this, it becomes possible to determine the stability, risk zones, and warning timing of prediction market prices. Here, prediction market prices may include waste resource prediction market prices and new resource prediction market prices.

[0102] The predictive price confidence collapse sensitivity index can be calculated based on the formula [(1 - duration of predictive price maintenance) * (predictive price change range / change time interval) * market reaction sensitivity coefficient].

[0103] The predicted price maintenance rate is an indicator representing the ratio or degree to which the calculated predicted market price for waste resources or new resources is maintained based on actual market transactions. In other words, it reflects how long the actual transaction price or settlement price has been maintained within a certain allowable margin of error since the predicted market price was calculated.

[0104] The demand forecasting and price calculation unit (320) sets the time when the predicted market price is calculated as the reference time, sets a certain observation period after the reference time, calculates the ratio or duration of actual transaction prices that occurred during the certain observation period that are within the allowable error range compared to the predicted market price, and normalizes the ratio or duration of the transaction to calculate the predicted price duration.

[0105] For example, if the predicted market price is 1,000 won, the margin of error is ±5%, and the observation period is 10 days, and the actual trading price maintained the range of 950 to 1,050 won for 8 days out of the 10 days, the predicted price maintenance rate can be calculated as 0.8.

[0106] The predicted price fluctuation range is a value indicating how significantly the predicted market price has changed within a certain period. This reflects the absolute or relative magnitude of the fluctuation in the predicted market price.

[0107] The demand forecasting and price calculation unit (320) sets the price of the waste resource forecast market or the price of the new resource forecast market at the reference point as the reference forecast price, and then sets the maximum or minimum value among the updated price of the waste resource forecast market or the new resource forecast market as the updated forecast price, and calculates the difference between the reference forecast price and the updated forecast price to determine the range of change in the forecast price.

[0108] For example, if the base predicted price is 1,000 won and the updated predicted price after 3 days is 1,150 won, the predicted price change range can be calculated as 150 won.

[0109] The change time interval represents the length of time required for a change in predicted market prices to occur. In other words, even with the same magnitude of change, the shorter the time it takes to occur, the greater the risk of confidence collapse is considered to be.

[0110] The demand forecasting and price calculation unit (320) records the initial time of calculation when the waste resource forecast market price or new resource forecast market price is calculated, records the time of price change when the forecast market price is significantly changed, calculates the time difference between the initial time of calculation and the time of price change, and determines the time difference as the change time interval.

[0111] For example, if the initial calculation time is 10:00 on January 1 and the price change time is 10:00 on January 2, the change time interval can be calculated as 24 hours.

[0112] The market reaction sensitivity coefficient is a measure indicating how sensitively market participants react to changes in market prices. It comprehensively reflects factors such as changes in trading volume, trade cancellations, and increases in inquiries in the event of price fluctuations.

[0113] The demand forecasting and market price calculation unit (320) sets the period before the market price change and the period after the market price change based on the point in time when a change occurs in the waste resource forecasting market price or the new resource forecasting market price, and calculates the rate of increase / decrease in transaction volume, the rate of increase / decrease in transaction cancellation rate and user response indicators by comparing the period before the market price change and the period after the market price change, and calculates the market response level by combining the calculated rate of increase / decrease in transaction volume, the rate of increase / decrease in transaction cancellation rate and user response indicators according to pre-set weights, and calculates the market response level by normalizing it to a standard range.

[0114] Here, the user response indicator is an indicator that indicates what behavioral changes were actually triggered by the change in the predicted market price to the trading parties and the users of the waste resource trading brokerage server (300). That is, it can quantify active and passive behavioral changes of users, such as whether they intend to continue the transaction, cancel or delay the transaction, request reconfirmation, or change the conditions, rather than simply whether a price change occurred. The user response indicator can be calculated by including behavior-based data such as the number of transaction cancellation requests that occurred after the change in the predicted market price, the number of re-quote requests or price re-inquiries that occurred after the change in the predicted market price, whether the waiting time increased while the transaction was in progress, the frequency of requests to change transaction conditions (quantity adjustment, schedule change, etc.), and the number of price-related inquiry messages or contacts to the customer center.

[0115] The demand forecasting and market price calculation unit (320) sets a reference period prior to the market price change and a reaction observation period after the market price change based on the time when a change in the predicted market price occurs, collects user behavior data that occurred during the reaction observation period after the market price change, and classifies the collected user behavior data by category. At this time, the classification category includes at least one of transaction cancellation, re-quote request, transaction delay, transaction condition change, and inquiry regarding the market price. For each classification category, the demand forecasting and market price calculation unit (320) calculates the increase ratio or change in frequency of occurrence compared to the period prior to the market price change, converts the amount of change for each category into a value representing the user reaction intensity, and can calculate a comprehensive user reaction index according to the market price change by weighting and combining multiple user reaction intensity values.

[0116] At this time, the user response indicator does not treat all actions equally. For example, a transaction cancellation is reflected with the highest response intensity because it directly leads to the actual failure of the transaction, a request for a re-quote is reflected with a medium level of response intensity because it is judged that there is an intention to maintain the transaction but confidence in the price is shaken, and a simple re-check of market prices or inquiry is reflected with a relatively low response intensity as an information search action. Accordingly, the demand forecasting and market price calculation unit (320) can calculate the user response indicator by applying a pre-set weight for each type of action.

[0117] For example, when there are 10 transaction cancellations (+100% compared to before the change), 15 re-quote requests (+50% compared to before the change), and 30 price inquiries (+20% compared to before the change) over 3 days following a price change, with weights of 0.5 for transaction cancellations, 0.3 for re-quote requests, and 0.2 for price inquiries, the overall user response metric can be calculated as 0.69 (0.50+0.15+0.04) by applying weights to the growth rates of each item. This value can be interpreted as a figure indicating a state where user behavior reacted relatively strongly to the price change.

[0118] This method of calculating user response indicators makes it possible to distinguish whether price fluctuations have actually affected trading behavior, to analyze simple trading volume reductions separately from psychological anxiety reactions, and to incorporate behavior-based empirical data into market response sensitivity coefficients. As a result, the present invention provides a technical effect of determining market sensitivity based on whether users have reacted, rather than whether prices have changed.

[0119] Various examples of calculating the confidence collapse sensitivity index for predicted market prices are as follows.

[0120] Examples Sustainability of maintaining predicted price Predicted price fluctuation range Change time interval Market Response Sensitivity Coefficient Collapse of confidence in predicted market prices, sensitivity index Example 1 0.85 5 10 0.6 0.045 Example 2 0.60 10 5 0.8 0.64 Example 3 0.40 15 3 1.2 3.60 Example 4 0.20 20 2 1.5 12.0 Example 5 0.05 25 1 2.0 47.5

[0121] Example 1 represents a stable predicted market price range, where the duration of maintaining the predicted price is very high, price changes are gradual, and market reaction sensitivity is low, resulting in a very low index of confidence breakdown, thereby quantitatively proving that the predicted market price is in a reliable state. Example 2 represents an initial confidence wavering range, where the duration of maintaining the predicted price is shortened, the range of price changes expands within a short period, and market reaction increases visibly, causing the index of confidence breakdown to rise sharply. Accordingly, the range where a crack in confidence regarding the predicted market price is detected can be captured early.

[0122] Example 3 is an accelerated confidence collapse phase, in which the sustainability of maintaining the predicted price decreased sharply, price fluctuations were concentrated in a short period, and user reactions, transaction cancellations, and requests for re-quotes surged, causing the predicted price confidence collapse sensitivity index to exceed 1 and increase non-linearly. This indicates that the market confidence collapse is a structural transition rather than a simple fluctuation.

[0123] Example 4 is a confidence collapse threshold section where the predicted price is hardly maintained, and the prediction price confidence collapse sensitivity index surged to double digits due to short-term rapid changes and high-sensitivity reactions. This clearly warns that a decision-making model based on predicted market prices is no longer valid.

[0124] Example 5 is a complete collapse and panic reaction phase, in which the sustainability of maintaining the predicted price is virtually zero, and the prediction price confidence collapse sensitivity index explodes due to extreme price fluctuations and the maximum market reaction sensitivity coefficient. This numerically demonstrates the need for market panic, price realignment, and a change in trading strategy.

[0125] The Predicted Price Confidence Collapse Sensitivity Index is based on the flow of actual market observation data, such as three predicted values, logs of actual trading times, measured time data of the prediction maintenance period, trading volume before and after price changes, cancellation rates, inquiry growth rates, and user behavior-based response data; therefore, it accurately reflects the temporal and behavioral sequence of actual market phenomena—prediction, collapse, and reaction—rather than an arbitrary mathematical combination.

[0126] While existing simple volatility indicators only consider price, the Prediction Confidence Collapse Sensitivity Index simultaneously reflects whether prediction confidence is maintained, the speed of change, and the intensity of market reaction, making it possible to clearly distinguish between cases where market confidence is maintained and cases where it has collapsed, even with the same price change.

[0127] Furthermore, the Predicted Price Confidence Collapse Sensitivity Index is an interpretable figure and a controllable indicator, as it enables automatic invalidation of predicted prices, readjustment of trading conditions, upward adjustment of the safety factor, and switching of user notification levels when a certain threshold is exceeded.

[0128] Accordingly, the predicted price confidence collapse sensitivity index according to the present invention is calculated by comprehensively reflecting the duration of the predicted price maintenance, the speed of the predicted price change, and the degree of the market's actual reaction thereto; thereby, it provides the effect of quantitatively determining whether market confidence in predicted market prices has collapsed, going beyond simple price fluctuations. Since this is calculated based on actual transaction data and user behavior data, it can be proven to be an empirical indicator rather than an arbitrarily set formula.

[0129] The selection of individual variables for calculating the predicted price confidence collapse sensitivity index described above has the following specific background and reasons, which supports the fact that the present invention is a systematic approach rather than a simple combination of data.

[0130] Predicted price maintenance duration: In a waste resource trading brokerage server (300), a more important factor than the predicted value itself is how long the predicted price remains valid in the actual trading environment. Even with the same predicted price, the level of trust in the predicted price is fundamentally different depending on whether it is maintained for a long period and actual trading is carried out normally, or if it is modified, withdrawn, or ignored by the market within a short period. Accordingly, in the present invention, the predicted price maintenance duration is calculated based on the duration from the time the predicted market price is first presented until the actual trading price deviates from the predicted range or the predicted price is recalculated, and this duration is used as a variable. This is intended to reflect the temporal characteristics of trust, such as how long the prediction has been accepted in the market, rather than simple price accuracy. By reflecting the predicted price maintenance duration, it is possible to clearly distinguish between a stable period where the predicted market price is maintained for a long period and an unstable period where trust collapses within a short period, and it is possible to prevent the problem of misjudging whether trust has collapsed even if price fluctuations occur.

[0131] Range of Predicted Price Fluctuation: The collapse of confidence in predicted market prices does not occur merely because the actual price differs from the prediction. While minor price adjustments are accepted as normal fluctuations in the market, rapid price changes exceeding a certain level drastically undermine confidence in the prediction itself. Accordingly, this invention defines the difference between the predicted price and the actual trading price, or the relative degree of fluctuation before and after a change in the predicted price, as the range of predicted price fluctuation and uses it as a variable. This is intended to quantitatively reflect the intensity of the confidence collapse. By reflecting the range of predicted price fluctuation, it is possible to quantitatively distinguish between normal fluctuations with minimal prediction errors and situations of rapid change where the validity of the prediction model itself is questionable, and to ensure that the sensitivity index for the collapse of confidence in predicted prices responds rationally in proportion to the magnitude of the change.

[0132] Change Time Interval: Even with the same magnitude of price fluctuation, changes occurring gradually over a long period differ significantly from those occurring intensively over a short period in terms of their impact on market confidence. Rapid short-term fluctuations are highly likely to be perceived by users as prediction failures, whereas long-term gradual changes tend to be perceived as adjustments within the prediction range. Accordingly, this invention indirectly reflects the concept of the speed of change by introducing the time interval required for a predicted price change to occur as a variable. By considering the change time interval, it is possible to sensitively detect the collapse of prediction confidence caused by short-term rapid fluctuations and prevent the problem of generating excessive risk signals regarding long-term gradual fluctuations. In other words, the prediction confidence collapse sensitivity index provides the effect of reacting differentially according to the speed of price change.

[0133] Market Reaction Sensitivity Coefficient: Whether a price change actually leads to a collapse of trust is most clearly revealed by changes in the behavior of market participants rather than the price change itself. User reactions following a change in predicted prices—such as increased transaction cancellations, requests for re-quotes, price inquiries, and transaction delays—are the tangible results of the collapse of market trust. Accordingly, this invention calculates the level of market reaction by synthesizing the rate of increase / decrease in trading volume, changes in cancellation rates, and user reaction indicators before and after a price change, and normalizes this to use as the market reaction sensitivity coefficient. This is a key variable designed to complement the limitations of price-centric judgment. By including the market reaction sensitivity coefficient, it is possible to clearly distinguish between cases where the market reacts significantly and cases where it reacts barely, even with the same price change. Through this, the index of sensitivity to the collapse of confidence in predicted prices functions as an empirical indicator linked to actual trading behavior, and serves as a behavior-based indicator for judging the collapse of trust, rather than a simple mathematical indicator.

[0134] The demand forecasting and price calculation unit (320) can calculate a market price signal locking coefficient by reflecting the number of platform price reference transactions, the number of external price reference transactions, and the number of individual negotiation transactions. The market price signal locking coefficient is an indicator that shows how strongly transactions actually completed within the trading platform provided by the waste resource trading brokerage server (300) depend on the price information provided by the trading platform. Specifically, the market price signal locking coefficient is a value quantified by comparing the proportion of transactions concluded based on the trading platform price with the proportion of transactions concluded by referencing external prices or through individual negotiation.

[0135] When the price information provided by a trading platform functions as a reasonable standard in the market, trading participants tend to accept the platform's prices as they are and execute trades without searching for external prices or engaging in individual negotiations. Conversely, if the reliability of the platform's prices is low or they fail to adequately reflect market conditions, trading participants tend to refer to external prices or attempt to adjust prices through individual negotiations.

[0136] Accordingly, to evaluate how strongly trading platform prices influence the determination of actual trading prices, a market price signal locking factor is calculated by comparing the number of trading platform price-referenced trades with the relative ratio of external price-referenced trades and individual negotiated trades.

[0137] Through the market price signal locking factor, it is possible to quantitatively determine whether trading platform quotes are merely reference information or function as a central signal for market price formation, and to objectively evaluate the reliability and market influence of the quote calculation algorithm.

[0138] The market price signal locking factor can be calculated based on the formula [Number of platform quote-referenced trades / (Number of external quote-referenced trades + Number of individually negotiated trades)].

[0139] The number of platform price reference transactions refers to the number of transactions concluded without separate price adjustments by accepting the waste resource prediction market price, new resource prediction market price, waste resource standard market price, or new resource standard market price provided by the trading platform as is.

[0140] The demand forecasting and price calculation unit (320) can identify transactions in which a transaction participant accepts the presented transaction price without modification and the transaction is completed after the waste resource prediction market price, new resource prediction market price, waste resource standard market price, or new resource standard market price provided by the waste resource trading brokerage server (300) is presented as a transaction price, and calculate the number of platform price reference transactions by aggregating the number of identified transactions within the same analysis period. For example, if a total of 200 transactions occurred during a specific period, and among them, 120 transactions were concluded by accepting the price presented by the trading platform without modification, the number of platform price reference transactions can be calculated as 120.

[0141] The number of externally referenced trades refers to the number of trades executed after a trading participant adjusts or resets the trading price by referring to external market information, quotes from other exchanges, public price data, etc., in addition to the market quotes presented on the trading platform.

[0142] The demand forecasting and market price calculation unit (320) can identify transactions in which a transaction participant indicates whether to refer to an external market price or inputs external market price information during the process of modifying the transaction price, and calculate the number of external market price reference transactions by aggregating the number of transactions only when the transaction is finally concluded. For example, if there are 50 transactions concluded after modifying the transaction price by referring to an external market price during the same period, the number of external market price reference transactions can be calculated as 50.

[0143] The number of individually negotiated transactions refers to the number of transactions where the price is determined through an individual negotiation process between the trading parties, without directly referencing trading platform market prices or external market prices.

[0144] The demand forecasting and price calculation unit (320) can identify transactions where the transaction price has been modified multiple times without reference information on the trading platform price or external price and where there is a record of direct negotiation between the trading participants, and calculate the number of individual negotiation transactions by aggregating the number of transactions when the transaction is finally concluded. For example, if there are 30 transactions where the price was determined through individual negotiation between the trading parties during a specific period, the number of individual negotiation transactions can be calculated as 30.

[0145] Various examples of calculating market price signal locking factors are as follows.

[0146] Examples Platform price reference transaction count External market reference transaction volume Number of individual negotiated transactions Market price signal locking factor Example 1 150 20 10 5 Example 2 90 60 30 1 Example 3 40 80 60 0.29 Example 4 20 30 100 0.15

[0147] Example 1 is a stable market where trading platform prices are strongly trusted, and most transactions accept trading platform prices as they are, and the need for external price searching and negotiation is low, so trading platform prices function as the market price reference signal. Example 2 is a transitional market where trading platform prices and external prices are mixed, and trading platform prices and external prices have similar influence, and trading participants compare, verify, and select prices, so the reliability of trading platform prices is at an intermediate level.

[0148] Example 3 describes a state of collapsed confidence in trading platform prices, where few transactions follow the trading platform prices exactly, and external price searching and individual negotiation are the main methods of price determination. This implies that trading platform prices do not function as market signals.

[0149] Example 4 is an extremely negotiation-centered market where price determination relies almost entirely on individual negotiations, and trading platform quotes have lost their reference value. This means that the technical limitations of the trading platform quote provision function are clearly exposed.

[0150] It can be seen that as the market price signal lock-up factor decreases, the influence of trading platform prices weakens, while the cost of searching for market prices and transaction complexity increase.

[0151] The market price signal lock-in coefficient is a valid indicator based on actual transaction data, as it determines whether to reference platform quotes based on price acceptance logs, whether to reference external quotes based on price modification reasons and re-quote logs, and whether to conduct individual negotiations based on negotiation messages and price change history.

[0152] In other words, the market price signal locking coefficient according to the present invention is calculated based on which price determination path a trader selected during an actual trading process, thereby allowing for the quantitative evaluation of the extent to which the trading platform price functions as a reference signal for market price formation. Accordingly, the market price signal locking coefficient is not an arbitrarily set formula, but a technical indicator that can be repeatedly verified by actual transaction data, and can be utilized for evaluating the reliability of price calculation algorithms and controlling platform pricing policies.

[0153] The selection of individual variables for calculating the market price signal locking factor described above has the following specific background and reasons, which supports the fact that the present invention is a systematic approach rather than a simple combination of data.

[0154] Number of Platform Quote-Referenced Transactions: The number of platform quote-referenced transactions refers to the cumulative number of trades where traders directly accepted the quotes provided by the platform as the pricing standard without any modifications or comparisons. This is the most direct indicator reflecting whether the trading platform quotes function as a reference point for price reliability and whether they trigger additional price searches. A high number of platform quote-referenced transactions implies that the platform quotes are sufficiently consistent with the market average or expected price, and that price perceptions among traders have converged. In other words, this indicates that price signals within the trading platform are close to being locked. By reflecting this variable, the effectiveness of platform quotes, the reliability of pricing algorithms, and the technical completeness of the quote provision function can be quantitatively evaluated based on trading results.

[0155] Number of External Quote-Referenced Transactions: The number of external quote-referenced transactions refers to the number of trades where a trader determines the price after verifying external market information (other platforms, public quotes, third-party price data, etc.) rather than accepting the trading platform's quotes as they are. This variable is designed to clearly identify cases where the trading platform's quotes fail to function as the sole standard for price determination. An increase in the number of external quote-referenced transactions implies the existence of a discrepancy between the platform's quotes and external market prices, or that traders have doubts regarding the accuracy or fairness of the platform's quotes. In other words, this serves as an indicator of the openness or dispersion of the trading platform's price signals. By including this variable, it is possible to detect the relative competitiveness of the platform's quotes, price alignment with the external market, and the timing of price signal deviations at an early stage, while also providing the technical basis for automatically adjusting quote corrections or the ratio of external data reflection.

[0156] Number of Individually Negotiated Transactions: The number of individually negotiated transactions refers to the number of trades where the price is determined by individual negotiation between trading parties, without relying on either the trading platform's market price or external market prices. This is a key variable for capturing a state where market price signals fail to function structurally and trading becomes individualized and non-standardized. An increase in the number of individually negotiated transactions is often accompanied by a collapse of market confidence, increased price volatility, and deepening information asymmetry among traders. In other words, it quantitatively represents the unlocking or breakdown stage of price signal locking. By reflecting this variable, one can clearly identify the degree of market irregularity, the level of weakening of platform price control, and the limitations of price prediction models, and determine the critical state at which platform operators must intervene.

[0157] By using platform quote-referenced trades, external quote-referenced trades, and individual negotiated trades together when calculating the Market Price Signal Lock-in Factor, all three stages of pricing—price acceptance, price comparison, and price redefinition—are numerically represented. Accordingly, unlike simple volume indicators, the Market Price Signal Lock-in Factor functions as a non-random, verifiable technical indicator based on traders' actual pricing behavior.

[0158] The transaction brokerage unit (330) can mediate transactions by applying the predicted market price of waste resources and the predicted market price of new resources as transaction standards to coordinate transaction conditions between the seller and the collector. Here, the predicted market price is not merely reference information, but is applied as a standard value for determining transaction feasibility and forming transaction conditions.

[0159] The transaction brokerage unit (330) queries the predicted market price corresponding to the item to be traded based on the waste resource sales information and new resource purchase information registered from the seller terminal (100). Subsequently, the transaction brokerage unit (330) compares the desired transaction conditions presented by the seller with the predicted market price to determine whether it falls within the range where trading is possible. For example, the transaction is set as a transaction subject to brokerage only if the desired selling price presented by the seller falls within the acceptable range of the predicted market price.

[0160] The transaction brokerage unit (330) selects multiple candidates for the transaction to be brokered. In this process, the transaction brokerage unit (330) primarily determines whether the range of collectible items, the quantity that can be processed, and the collection schedule registered by each collector match the seller's transaction conditions. Through this, collectors for whom the transaction itself cannot be established are excluded in advance.

[0161] The transaction brokerage unit (330) evaluates the collectors who are likely to perform actual transactions based on predicted market prices among the first-stage selected collector candidates. The evaluation criteria include the distance between the collector's location and the seller's location, past transaction history, transaction success rate, transaction rating, and whether a transaction delay has occurred. The transaction brokerage unit (330) synthesizes the evaluation criteria and selects the collectors who are likely to perform transactions stably based on predicted market prices as priority matching candidates.

[0162] The transaction brokerage unit (330) automatically generates transaction conditions based on predicted market prices for selected matching candidates. The transaction conditions include the transaction amount, collection schedule, and settlement method, and can be set within a range where negotiation between transaction participants is possible using the predicted market price as a central value. The generated transaction conditions are provided to the seller terminal (100) so that the seller can check the conditions and then accept the automatic matching or select one of the multiple matching candidates.

[0163] When a seller accepts automatic matching or selects a specific collector, the transaction broker (330) converts the transaction into a confirmed transaction and transmits the transaction confirmation information to the collector terminal (200). Subsequently, the transaction broker (330) can manage the transaction progress status so that the schedule for waste resource collection and new resource delivery remains consistent with the logistics conditions reflected in the prediction market price calculation.

[0164] The transaction brokerage unit (330) applies the predicted market price as a standard value for establishing a transaction to prevent excessive deviation in transaction conditions and minimize the possibility of price disputes between transaction participants. As a result, the seller and the collector can proceed with the transaction within a predictable price range, and the waste resource transaction brokerage server (300) can provide a stable and fair transaction brokerage environment.

[0165] The transaction brokerage unit (330) can manage both transactions as a single transaction flow when waste resource collection transactions and new resource delivery transactions occur simultaneously or in conjunction. Through this, the transaction brokerage unit (330) prevents conflicts between transaction conditions in advance and controls the collection schedule and delivery schedule to ensure they are consistently coordinated. As a result, the transaction brokerage unit (330) can minimize transaction delays or settlement errors and simultaneously improve transaction efficiency and user satisfaction.

[0166] In this way, the transaction brokerage unit (330) can prevent indiscriminate price proposals or market distortion by utilizing predicted market prices not as simple display information, but as a standard value for determining transaction execution and generating conditions. Additionally, by combining with the selection of collectors based on transaction history and ratings, it can provide the effect of improving the transaction success rate and reducing transaction delays and cancellations.

[0167] The transaction brokerage unit (330) can calculate a transaction selection bias distortion index by reflecting the deviation between the selection frequency and average selection frequency of each collector and the degree of automatic matching intervention as a selection involuntariness coefficient. The transaction selection bias distortion index is an indicator calculated by comprehensively reflecting how excessively a specific collector among multiple collectors is selected or excluded compared to the average selection frequency in the waste resource trading brokerage platform system, and whether that selection is induced by systemic or environmental factors rather than the collector's voluntary judgment. In other words, the transaction selection bias distortion index simultaneously quantifies the degree of imbalance in the selection distribution among collectors and whether said imbalance is distorted by structural factors.

[0168] In existing trading platforms, factors such as automated matching logic, display order, price predictions, logistical convenience, and the highlighting of past success rates can lead to specific collectors being repeatedly selected regardless of actual quality. In the long term, this selection bias can result in the undermining of fair trade, barriers to entry for new collectors, market distortion, and reduced transaction stability.

[0169] Accordingly, the present invention enables the evaluation and control of the fairness and balance of the selection structure within the platform by introducing a transaction selection bias distortion index that weights and reflects deviation from the mean and involuntary selection factors, rather than simply the number of selections.

[0170] The trading selection bias distortion index is [ It can be calculated based on the formula of ].

[0171] The selection frequency of a collector refers to the number of times a specific collector is finally selected as a trading partner among multiple transactions that occurred on a trading platform during a predetermined analysis period, or the selection ratio. The transaction brokerage unit (330) sets an analysis period to extract transactions confirmed as completed during the analysis period, identifies the finally selected collector for each extracted transaction, aggregates the number of transactions in which a specific collector was selected, and normalizes the aggregated number of selections as a ratio relative to the total number of transactions or total number of selections during the period, thereby determining the selection frequency of the collector. For example, if the total number of transactions during the analysis period is 100 and the number of transactions in which collector A was selected is 30, the selection frequency of collector A can be calculated as 30 times or 0.30 relative to the total.

[0172] The average selection frequency is a value obtained by averaging the selection frequencies of all collectors who participated in the transaction within the same analysis period, and represents the ideal or standard selection level within the transaction platform. The transaction brokerage unit (330) can identify collectors who participated in the transaction during the analysis period, calculate the selection frequency for each collector, sum the calculated multiple selection frequencies, and divide the result by the number of collectors to calculate the average selection frequency. For example, if the number of collectors is 5 and the selection frequency of each collector is 30, 25, 20, 15, and 10 times, the average selection frequency can be calculated as 20.

[0173] The selection involuntariness coefficient is a coefficient indicating the degree to which a specific collector's selection is induced by systemic or environmental factors rather than the user's voluntary judgment among selected transactions. In other words, the selection involuntariness coefficient reflects the level of coercion, induction, and automation of the selection result. The transaction brokerage unit (330) determines whether the collector selection method for each transaction is automatic matching or manual selection, and if automatic matching is maintained, it increases the involuntariness score; if fixed exposure order, recommendation highlighting, or price advantage display is applied, it applies additional weights; if the collector changes the recommendation result, it decreases the involuntariness score; it calculates the level of selection involuntariness by combining the increased and decreased factors, and normalizes the calculated value to determine the selection involuntariness coefficient. For example, if the ratio of transactions maintaining automatic matching is 80%, the ratio of recommendation highlighting application is 70%, and the ratio of user manual change is 10%, the involuntariness score is calculated high, and the selection involuntariness coefficient can be set to 1.3. Conversely, when most selections are made manually, the selection non-volatility coefficient can be set low, such as 0.7.

[0174] The transaction selection bias distortion index according to the present invention is based on the deviation between the selection frequency and the average selection frequency for each collector, and corrects for whether the selection was made voluntarily by the user using a selection involuntariness coefficient, thereby providing the effect of quantitatively detecting structural selection bias and distortion phenomena occurring within a transaction platform. Through this, it can contribute to ensuring transaction fairness, improving recommendation algorithms, preventing market distortion, and enhancing platform reliability.

[0175] Various examples of calculating the trading selection bias distortion index are as follows.

[0176] (Analysis period: Same period (e.g., 1 month), Total transactions: 100, Number of participating collectors: 5, Average selection frequency: Total selections 100 / Collectors 5 = 20 times)

[0177] Examples Collector's choice frequency Average selection frequency deviation Selective involuntariness coefficient Bias contribution by collector Transaction Selection Bias Distortion Index Example 1 22, 19, 21, 18, 20 20, 20, 20, 20, 20 2, 1, 1, 2, 0 0.9, 1.0, 0.8, 1.0, 0.9 1.8, 1.0, 0.8, 2.0, 0.0 5.6 Example 2 40, 20, 18, 12, 10 20, 20, 20, 20, 20 20, 0, 2, 8, 10 1.2, 1.0, 1.0, 1.1, 1.1 24.0, 0.0, 2.0, 8.8, 11.0 45.8 Example 3 60, 15, 10, 8, 7 20, 20, 20, 20, 20 40, 5, 10, 12, 13 1.5, 1.2, 1.3, 1.3, 1.4 60.0, 6.0, 13.0, 15.6, 18.2 112.8

[0178] In Example 1, the selection frequency is evenly distributed around the average, and the influence of automatic matching and exposure induction is low, so the platform selection structure is in a fair and stable state. In Example 2, selection is concentrated on a specific collector, and selection bias is clearly detected due to the partial operation of system factors such as automatic matching maintenance and recommendation exposure, but it is not at a stage of complete distortion. In Example 3, a specific collector is excessively and repeatedly selected, the automatic matching retention rate is very high, and there is almost no user change, quantitatively proving that the platform structure itself causes selection distortion.

[0179] The Transaction Selection Bias Distortion Index proves that it is not randomly generated by calculating collector selection frequency based on actual confirmed transaction data, calculating average selection frequency using overall transaction statistics, and calculating the coefficient of involuntary selection based on automatic matching, recommendation exposure, and user change logs, thereby avoiding the use of fictitious psychological variables and abstract scores.

[0180] The transaction selection bias distortion index is structured not merely as a high number of selections, but as deviation from the mean * degree of involuntariness, allowing for a clear distinction between voluntary and system-induced selections even with the same number of selections.

[0181] In other words, the transaction selection bias distortion index according to the present invention can quantitatively evaluate the degree of distortion in the selection structure within a platform by weighting involuntary selection factors, such as automatic matching and recommendation exposure, to the deviations in selection frequency and average selection level derived from actual transaction logs. Accordingly, this index is not an arbitrarily set formula but an indicator that can be repeatedly verified by actual transaction data, thereby providing a technical basis for ensuring transaction fairness and improving brokerage algorithms.

[0182] The selection of individual variables for calculating the transaction selection bias distortion index described above has the following specific background and reasons, which supports the fact that the present invention is a systematic approach rather than a simple combination of data.

[0183] Selective Frequency by Collector: In waste resource trading platforms, a phenomenon may occur where specific collectors are repeatedly selected even for transactions with identical conditions. While this phenomenon may be attributed to the voluntary preferences of trading participants, there is also a possibility that it is induced by systemic factors such as automatic matching algorithms, recommendation exposure frequency, and initial assignment retention policies. Therefore, to quantitatively evaluate transaction selection bias, it is necessary to use the selective frequency by collector as a primary variable, which indicates how often each collector is actually selected. Selective frequency by collector can be calculated directly from data on confirmed transactions and is the most intuitive indicator reflecting the actual operational status of the trading platform. Furthermore, unlike simple exposure or recommendation counts, it reflects whether actual transactions were concluded, thereby clearly demonstrating the effectiveness of the selection results. By using selective frequency by collector as a variable, it is possible to objectively capture phenomena of selection concentration or selection alienation regarding specific collectors, and to evaluate the imbalance in the transaction selection structure based on empirical data.

[0184] Average Selection Frequency: It is difficult to determine whether transaction choices are distorted based solely on the selection frequency of individual collectors. For instance, an increase in a specific collector's selection frequency may be a normal phenomenon when the total transaction volume increases. Accordingly, a reference value is required to judge individual collectors' selection frequencies as a standard of comparison, and for this purpose, the average selection frequency for all collectors is introduced. The average selection frequency is the value obtained by averaging the selection frequencies of all collectors within the same analysis period, serving as a baseline for judging the relative selection levels among transaction participants. This allows for the quantitative determination of whether an individual collector's selection frequency is excessively high or low compared to the market average. By calculating the deviation based on the average selection frequency, the degree of distortion in the selection distribution can be stably compared regardless of increases or decreases in transaction volume, and a consistent standard of judgment can be maintained even if the platform scale or the number of transactions changes.

[0185] Coefficient of Involuntary Choice: It is difficult to accurately identify the causes of transaction selection bias based solely on differences in selection frequency. In particular, when user choices are effectively restricted by system factors such as automatic matching, retention of default values, and preferential exposure of recommendations, a choice may appear to have been made on the surface but could actually be involuntary. Accordingly, an indicator is needed to distinguish whether user intent is sufficiently reflected in the selection outcome. The coefficient of involuntary choice can be calculated based on actual user behavior logs, such as the automatic matching retention rate, the user selection change rate, the recommendation acceptance rate, and the frequency of manual selection intervention. This allows for the evaluation of voluntary versus involuntary choices even with the same selection frequency. By introducing the coefficient of involuntary choice, it is possible to clearly distinguish between simple popularity bias and selection distortion caused by system induction, and to diagnose issues in the transaction selection structure from the perspective of algorithm design.

[0186] By using the selection frequency by collector, the average selection frequency, and the selection involuntariness coefficient together, the transaction selection bias distortion index of the present invention can simultaneously reflect how often a selection was made, how much it deviates from the average, and how involuntary the selection was. Accordingly, this index is not merely a simple statistical indicator; it can precisely detect transaction selection distortions caused by the platform structure based on actual transaction data and provides technical effects that can be utilized for adjusting automatic matching logic, improving recommendation algorithms, and establishing fairness control policies.

[0187] The logistics management department (340) is configured to minimize the overall logistics burden by analyzing the logistics flow that occurs during the waste resource collection and new resource delivery processes based on waste resource transaction data and new resource transaction data, and by efficiently reorganizing it.

[0188] The logistics management department (340) can define transaction units for each transaction of waste resources and new resources based on the collection location, delivery location, scheduled time of collection or delivery, target items, and quantity. Since collection and delivery may take place at different locations and times even for the same transaction, the collection unit and delivery unit can be managed separately. Through this, the logistics management department (340) can prevent conflicts between transaction conditions in advance when analyzing logistics flow.

[0189] The logistics management department (340) can determine whether it is appropriate for each transaction unit to pass through a specific logistics base by referring to base logistics base information. For example, the logistics management department (340) can select an efficient method between a direct delivery method and a logistics base passing method by considering the distance between the collection location and the delivery location, the processing capacity of the logistics base, and the items that can be stored.

[0190] The logistics management department (340) can group transaction units based on the type and loading capacity of the transport vehicle. That is, the logistics management department (340) can group transactions that can be transported together within the loading capacity range among multiple waste resource collection transactions occurring in the same or adjacent areas into a single transport group. In this process, the logistics management department (340) can prevent unnecessary waiting time or additional transport by prioritizing the grouping of transactions with similar collection schedules.

[0191] The logistics management department (340) can determine whether the same transport vehicle can load and deliver new resources immediately after collecting waste resources in the case of a transaction where new resources are scheduled to be delivered after waste resources are collected. Through this, the logistics management department (340) can minimize empty vehicle operation and configure a complex transport route that continuously connects the collection route and the delivery route.

[0192] The logistics management department (340) can recommend optimal sales routes, purchase routes, collection routes, and delivery routes for each group for multiple transport groups. At this time, the recommended route is set so that at least one of the factors of travel distance, travel time, and number of transports is minimized, and the possibility of conflict with traffic conditions or available work time may also be considered.

[0193] The logistics management unit (340) can provide the recommended route to the transaction brokerage unit (330) and the collector terminal (200) so that it is reflected in the actual collection and delivery schedule. In this process, if a schedule change or transaction cancellation occurs during the transaction, the logistics management unit (340) can minimize the decrease in logistics efficiency by reorganizing the transport group and route to reflect the changes.

[0194] In this way, the logistics management department (340) can structurally eliminate redundant transportation by optimizing the logistics flow through the steps of defining transaction units, grouping transportation, linking routes, and reconfiguring. Accordingly, the transportation distance and the number of transportations are reduced, thereby reducing logistics costs, and the quality degradation and management burden caused by the long-term accumulation of waste resources can be reduced.

[0195] In addition, the configuration of a combined transport route linking collection and delivery reduces empty vehicle operation, thereby reducing fuel consumption and carbon emissions and providing the effect of alleviating environmental burden. This logistics optimization acts as a key factor in improving the overall transaction processing efficiency and sustainability of the waste resource trading brokerage server (300).

[0196] When the settlement and donation linkage unit (350) confirms that the waste resource collection and new resource delivery transaction has been completed by the transaction brokerage unit (330), it may convert the transaction into a transaction subject to settlement and initiate a financial settlement procedure. At this time, whether the transaction is completed may be determined when at least one of the following conditions is satisfied: confirmation of completion of collection, confirmation of completion of delivery, and approval of transaction termination between transaction participants.

[0197] The settlement and donation linkage unit (350) can calculate the brokerage fee generated during the transaction brokerage process. The brokerage fee can be calculated differentially based on the transaction amount, the transaction type (waste resource-only transaction, new resource-linked transaction, etc.), and the transaction success conditions (whether automatic matching is performed, whether emergency collection is performed, whether logistics optimization is applied, etc.). Through this, the settlement and donation linkage unit (350) can charge a reasonable brokerage fee for transactions with different transaction difficulty or platform contribution levels.

[0198] The settlement and donation linkage unit (350) can determine the settlement amount between the seller and the collector by deducting the brokerage fee from the transaction amount or separating it into a separate item once the brokerage fee is calculated. At this time, the settlement method can be set to at least one of immediate settlement, bulk settlement after waiting for settlement, and periodic settlement, and can be applied differently depending on the type of transaction participant or transaction conditions.

[0199] The settlement and donation linkage unit (350) may set at least a portion of the calculated brokerage fee as the donation target amount. The donation target amount may be fixed at a certain percentage according to the platform operation policy, or may be determined according to a percentage directly selected by the transaction participant when registering the transaction or when confirming the transaction.

[0200] Donation organizations can be configured to allow selection from multiple organizations registered on the platform in advance. In this case, donation organizations can be classified and managed according to different purposes, such as environmental protection, resource recycling, social welfare, and community support, and transaction participants can select a donation organization that aligns with the nature of their transaction or their values.

[0201] The settlement and donation linkage unit (350) can execute donations by automatically transferring the donation target amount to the relevant donation organization in conjunction with the settlement procedure, or by accumulating it over a certain period and then making a lump-sum donation. In this process, the settlement and donation linkage unit (350) can manage the donation execution time, donation amount, and donation organization information by storing them all so that they can be viewed later.

[0202] The settlement and donation linkage unit (350) can store and manage the settlement details and donation details for each transaction. Accordingly, through their transaction history screen, the transaction participant can check the brokerage fee generated from a specific transaction, the amount of the fee used for donation, and information about the donation organization.

[0203] In addition, the settlement and donation linkage unit (350) can provide a comprehensive list of the transaction participants' cumulative transaction amount, cumulative donation amount, and participating donation organizations. Through this, the transaction participants can intuitively recognize that their transaction activities are leading to the creation of social value beyond simple economic activities.

[0204] In this way, the settlement and donation linkage unit (350) is configured to manage the transaction settlement and donation procedures in a single flow so that the transaction itself is connected to social contribution without the need for a separate donation act. Accordingly, the burden of donation participation for transaction participants is reduced, and the reliability of the platform and the sustainability of participation can be improved.

[0205] In addition, by linking and managing transaction history and donation details, transparency of donation is ensured, and the waste resource trading brokerage server (300) can provide the effect of contributing to the establishment of a sustainable resource circulation ecosystem with public interest and reliability, going beyond a simple brokerage system.

[0206] The settlement and donation linkage unit (350) can calculate a social return duration resonance index by reflecting the period-based donation amount linked to the completion of the transaction, the strength of the transaction linkage, and the time duration weighting according to time. The social return duration resonance index is an indicator that quantitatively represents the degree to which social return activities occurring on the transaction platform do not remain as one-time events, but are linked to the transaction activities, continue over time, and resonate repeatedly. Specifically, the social return duration resonance index is calculated by comprehensively reflecting the period-based donation amount, the degree to which the donation is linked to the actual transaction, and the characteristics of the return activities being maintained and accumulated over time.

[0207] Existing donation evaluation methods are based primarily on simple cumulative amounts or frequency, which makes it difficult to assess connectivity with transaction activities, sustainability, and structural reproducibility. Accordingly, this invention proposes a Social Return Sustainability Resonance Index that combines the scale of donation, transaction connectivity, and temporal sustainability to determine whether social return is amplified voluntarily and repeatedly within a platform structure.

[0208] The social return sustainability resonance index can be calculated based on the formula [Period-specific donation amount * Transaction linkage strength * Time duration weighting].

[0209] The donation amount by period refers to the amount actually executed for the purpose of social return by a trading platform or a trading participant during a pre-set time unit. The settlement and donation linkage unit (350) can set the donation calculation period in units of days, weeks, or months, and can extract the amount that has been actually executed from the donation details that occurred during the donation calculation period and determine it as the donation amount by period. For example, if the donation amount for January is 500,000 won, the donation amount for February is 700,000 won, and the donation amount for March is 600,000 won, each monthly amount can be set as the donation amount by period.

[0210] The transaction linkage strength is a coefficient that quantitatively indicates whether the donation amount during the period is a purely voluntary donation or a donation that is directly linked to the occurrence of a transaction, the completion of a transaction, the transaction amount, etc. The settlement and donation linkage unit (350) classifies the donation details that occurred during the period according to whether they are linked to a transaction, assigns a high linkage weight to donations that are automatically generated upon completion of a transaction and assigns a low linkage weight to donations unrelated to the transaction, and can calculate the transaction linkage strength by combining the ratio and strength of the transaction-linked donations. For example, if 80% of the total donations are automatically generated upon completion of a transaction and the remaining 20% ​​are voluntary donations, the transaction linkage strength can be set to 0.8.

[0211] The time duration weight is a weighting coefficient intended to reflect the effect that donation and social return activities are not one-off events but continue and accumulate over time. The settlement and donation linkage unit (350) sets a reference point, tracks whether donations occur during each period, assigns a high weight to periods where donations occur continuously, gradually reduces the weight when there is a gap in donations, and normalizes the weight results to determine the time duration weight for the corresponding period. For example, if donations are maintained for three consecutive months, the time duration weight is set to 1.0; if donations are resumed after one interruption in the middle, the time duration weight is set to 0.7; and if donations are one-off, the time duration weight can be set to 0.4.

[0212] Various examples of calculating the social contribution sustainability resonance index are as follows.

[0213] Examples period Donation amount by period (won) Transaction linkage strength Time duration weight Social Return Sustainability Resonance Index Example 1 t1 1,000,000 0.3 0.4 120,000 Example 2 t1, t2 500,000, 500,000 0.8 0.7 560,000 Example 3 t1, t2, t3 400,000, 450,000, 500,000 0.9 1.0 1,215,000 Example 4 t1, t2 1,500,000, 1,200,000 0.2 0.5 270,000

[0214] Looking at Examples 1 and 2, Example 1 involved a relatively large donation amount over a single period; however, due to low transaction linkage intensity and time duration weighting, social return activities failed to spread structurally. In contrast, while the donation amount itself was dispersed in Example 2, the social return duration resonance index increased by more than 4.6 times compared to Example 1 as each donation occurred repeatedly in conjunction with the completion of actual transactions. This demonstrates that transaction linkage and repeatability have a greater influence on the resonance effect than the size of the donation. Looking at Examples 2 and 3, Example 3 not only maintains a high transaction linkage intensity but also maintains the time duration weighting at its maximum level as donations continue uninterrupted over a continuous period. Consequently, despite having identical or similar donation scales, the social return duration resonance index in Example 3 rises by more than twofold compared to Example 2. This indicates that the social return duration resonance index quantitatively reflects the social ripple effect based on sustainability, rather than simply the accumulated amount.

[0215] Looking at Examples 3 and 4, Example 4 has a large absolute donation amount, but because it has low transactional ties and takes the form of voluntary and non-traditional donations, the social return sustainability resonance index is calculated to be relatively low. In contrast, Example 3 shows a resonance index that is about 4.5 times higher due to transactional ties and sustainability, despite having a smaller donation amount. This means that the index of the present invention excludes the problem of concentration on large donations and evaluates structural contribution.

[0216] The Social Return Sustainability Resonance Index utilizes only objective and verifiable data elements, such as donation amounts derived from actual transaction data, transaction linkage strength determined by the occurrence and conditions of transactions, and time duration weights calculated based on the continuity of donation history. Therefore, this index does not rely on arbitrary statistical coefficients or subjective scores, but is reproducibly calculated using data automatically collected during the actual operation of the platform.

[0217] By calculating the Social Return Sustainability Resonance Index, the sustainability and structuring levels of transaction-based social returns can be quantified; one-time donations can be clearly distinguished from repetitive and linked donations; the effects of social returns can be compared and analyzed before and after changes in platform policies; and this can be extended to evaluation indicators for ESG, public interest, and reliability. Furthermore, social returns can become established as part of the transaction structure rather than a one-time promotional tool; voluntary contributions from transaction participants can be encouraged; the public interest and reliability of the platform can be strengthened; and a sustainable ecosystem for resource circulation and social contribution can be established.

[0218] The selection of individual variables for calculating the social return sustainability resonance index described above has the following specific background and reasons, which supports the fact that the present invention is a systematic approach rather than a simple combination of data.

[0219] Donation Amounts by Period: Existing donation evaluation methods focused on cumulative donation amounts over a specific period or the scale of single-event donations, which had limitations in that they failed to adequately reflect the connection with actual platform transaction activities or the context in which donations occurred. Particularly in platform environments where repetitive and decentralized transactions occur, such as waste resource trading, the distribution and patterns of change in donation amounts occurring in each period more accurately reflect the actual ripple effect of social contribution than a donation amount at a single point in time. Accordingly, this invention is configured to decompose donation amounts into period-specific donation amounts rather than simple cumulative values, enabling the individual evaluation of the scale of social contribution generated by transaction activities during each period. This allows for the precise tracking of changes in the level of social contribution based on platform operational status, such as increases or decreases in transactions or changes in donations following the introduction of specific policies. By using period-specific donation amounts as a variable, it is possible to clearly distinguish between one-time high-value donations and continuous small-value donations, quantitatively analyze the response to social contribution in response to changes in transaction activity, and verify the social effects of changes in platform policies and fee structures.

[0220] Transaction Linkage Strength: When social contribution activities occur arbitrarily, unrelated to transactions, such donations are closer to external, one-time acts than to structural values ​​within the platform ecosystem. Conversely, donations that occur automatically upon the completion of a transaction or are linked to transaction conditions signify that the platform's transaction structure itself generates social value. In this invention, transaction linkage strength is introduced as a variable to quantify the extent to which donations are directly linked to transaction activities. Transaction linkage strength can be calculated by reflecting factors such as whether donations are automatic upon transaction completion, the ratio of donations to transaction amounts, and the degree of linkage between the number of transactions and the frequency of donations. By using transaction linkage strength as a variable, it is possible to distinguish between voluntary and arbitrary donations and structural and systematic donations; it is possible to induce a structure where increased transactions lead to increased social contribution; and it is possible to objectively evaluate the platform's public nature and the creation of ESG value.

[0221] Time Duration Weight: Existing donation evaluation indicators rely solely on whether a donation occurs or the amount, failing to reflect whether the donation has been sustained over a long period or concentrated temporarily. However, social trust and public interest are formed through repetition and stability over time, and long-term, continuous donations have a greater social impact than short-term donations. Accordingly, this invention introduces a time duration weight to reflect the duration and continuity of donations. The time duration weight can be set to gradually increase when donations occur continuously and decrease as the gap in donations lengthens. By using the time duration weight as a variable, quantitative rewards for continuous social contribution activities are possible, the overvaluation of short-term, event-based donations is prevented, and the long-term trustworthiness and public interest levels of the platform can be evaluated.

[0222] The social return sustainability resonance index of the present invention reflects the scale of social return through periodic donation amounts, the structural contribution of donations through transaction linkage strength, and the stability and long-term nature of social return through time duration weighting. Unlike evaluation methods centered on simple total donation amounts or frequency, the social return sustainability resonance index calculated by this combination of variables can objectively and reproducibly evaluate the substantial sustainability resonance effect of transaction-based social return.

[0223] Although the present invention has been described above with reference to one embodiment, those skilled in the art will understand that various modifications and changes can be made to the invention without departing from the spirit and scope of the invention as described in the following claims. Explanation of the symbols

[0224] 100: Seller's terminal 200: Collector terminal 300: Waste Resource Trading Brokerage Server 310: Data Collection Unit 320: Demand Forecasting and Market Price Calculation Department 330: Transaction Brokerage Department 340: Logistics Management Department 350: Settlement and Donation Linkage Department

Claims

Claim 1 A seller terminal that registers waste resource sales information by inputting the type, quantity, and disposal location of waste resources, and registers new resource purchase information by inputting the item and quantity of new resources; a collector terminal that registers waste resource collection information for the collection of waste resources based on the waste resource sales information, and registers new resource delivery information for the delivery of new resources; It includes a waste resource trading brokerage server that provides information on market prices for waste resources and new resources, stores waste resource trading data and new resource trading data, predicts the demand for waste resources and new resources by analyzing the waste resource market price information, new resource market price information, waste resource trading data, and new resource trading data using artificial intelligence, and calculates the market prices of waste resources and new resources based on the prediction results to broker transactions; the waste resource trading brokerage server includes a data collection unit that collects waste resource trading data including waste resource sales information and waste resource collection information, and new resource trading data including new resource purchase information and new resource delivery information, and collects and stores waste resource market price information and new resource market price information; a demand forecasting and price calculation unit that predicts the demand for waste resources and new resources by analyzing past trading patterns and current market prices based on the waste resource trading data, new resource trading data, waste resource market price information, and new resource market price information, and calculates the predicted market price for waste resources and the predicted market price for new resources that serve as the basis for transactions based on the prediction results; and the waste resource predicted market price and the A transaction brokerage unit that mediates the trading and settlement of waste resources and new resources by considering the location, rating, and transaction history of sellers and collectors based on predicted market prices for new resources, automatically matching the optimal collector or enabling sellers to select a collector; and based on at least one of the waste resource trading data, the new resource trading data, base logistics bases, and transport vehicles, a sales route, a purchase route,Logistics management department that groups and recommends collection and delivery routes; It includes a settlement and donation linkage unit that settles brokerage fees generated upon completion of waste resource transactions, donates at least a portion of the said brokerage fees to a pre-registered donation organization, and manages the donation history in conjunction with the transaction history; the said demand forecasting and market price calculation unit calculates a forecast price confidence collapse sensitivity index that indicates how easily the forecast market price can collapse by reflecting the forecast price maintenance duration, forecast price fluctuation range, change time interval, and market response sensitivity coefficient; the said forecast price maintenance duration is calculated by setting the point in time when the waste resource forecast market price or new resource forecast market price is calculated as the reference point, setting a certain observation period after the reference point, and calculating the ratio or maintenance time of transactions that exist within the allowable error range relative to the corresponding forecast market price among actual transaction prices that occurred during the certain observation period, and normalizing the ratio or maintenance time of transactions; the said forecast price fluctuation range is calculated by setting the waste resource forecast market price or new resource forecast market price at the reference point as the reference forecast price, and setting the maximum or minimum value among the subsequently updated waste resource forecast market price or new resource forecast market price as the updated forecast price, and the reference forecast Calculated based on the difference between the market price and the updated predicted market price, and the above change time interval records the initial calculation time when the hidden resource predicted market price or new resource predicted market price was calculated, records the time of price change when the predicted market price was significantly changed, and is calculated based on the time difference between the initial calculation time and the time of price change, and the above market reaction sensitivity coefficient sets the period before the price change and the period after the price change respectively based on the point in time when a change occurred regarding the waste resource predicted market price or new resource predicted market price, compares the period before the price change and the period after the price change to calculate the rate of increase / decrease in trading volume, the rate of increase / decrease in the transaction cancellation rate, and user reaction indicators, and the calculated rate of increase / decrease in trading volume,A waste resource trading brokerage platform system using AI-based demand forecasting and market price information, characterized by calculating a market response level by combining the rate of increase / decrease in transaction cancellation rates and user response indicators according to preset weights, and calculating the market response level by normalizing it to a reference range. Claim 2 delete Claim 3 A waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information, wherein, in claim 1, the demand forecasting and market price calculation unit combines the waste resource trading volume, weights according to the trading time, and new resource market price volatility coefficients, and calculates a bidirectional raw material-linked demand pressure index indicating how much pressure the resource supply and demand is under according to the trading activity in the waste resource market and the purchasing pressure in the new resource market, considering the new resource purchase volume and new resource trading reliability, and the new resource trading reliability is calculated by reflecting the transaction success rate indicator, delivery performance stability indicator, transaction repeatability indicator, transaction dispute occurrence indicator, and transaction evaluation score indicator. Claim 4 delete Claim 5 In paragraph 1, the demand forecasting and price calculation unit calculates a market price signal locking coefficient that indicates how strongly actual completed transactions depend on the waste resource market price or new resource market price by reflecting the number of platform price reference transactions, the number of external price reference transactions, and the number of individual negotiation transactions; the number of platform price reference transactions is calculated by aggregating the number of transactions in which the transaction is completed without the transaction participant modifying the presented transaction price after the waste resource prediction market price, new resource prediction market price, waste resource standard market price, or new resource standard market price has been presented as the transaction price; the number of external price reference transactions is calculated by identifying transactions in which the transaction participant indicates whether to reference the external price or enters external price information during the process of modifying the transaction price, and by aggregating the number of transactions when the transaction is finally concluded; the number of individual negotiation transactions is calculated by identifying transactions in which the transaction price has been modified multiple times without information on the waste resource prediction market price, new resource prediction market price, waste resource standard market price, new resource standard market price, or external price reference price, and where there is a record of direct negotiation between the transaction participants, and when the transaction is finally concluded A waste resource trading brokerage platform system utilizing artificial intelligence-based demand forecasting and market price information, characterized by aggregating and calculating the number of transactions in a given case. Claim 6 In claim 1, the settlement and donation linkage unit calculates a social return duration resonance index that indicates the degree to which social return activities do not end as a one-time occurrence but are linked to transaction activities, continue over time, and resonate repeatedly by reflecting the period-specific donation amount linked to transaction completion, transaction linkage strength, and time duration weighting according to time; the period-specific donation amount is calculated by extracting the amount of actual execution completed among the donation details that occurred during the donation calculation period; the transaction linkage strength is calculated by classifying donation details according to whether they are linked to a transaction, assigning a linkage weighting according to whether the donation is automatically generated upon transaction completion or is unrelated to the transaction, and synthesizing the ratio and strength of transaction-linked donations; and the time duration weighting is calculated by setting a reference point and determining whether the reference point is a period in which donations occurred continuously or a period in which donation gaps occurred. This characterizes a waste resource trading brokerage platform system using artificial intelligence-based demand forecasting and market price information.

Images