System for Providing Search Data Transaction Platform through User-Centric Real-Time Data Valuation and Transaction Mediation

Abstract

A system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to the present invention may include: a data receiving unit that receives search data from a user terminal owned by a user; a value analysis unit that calculates a commercial value score of the search data received through the data receiving unit in real time; a transaction brokerage unit that generates a transaction request based on the commercial value score calculated by the value analysis unit and transmits it to a plurality of data demander servers, collects transaction proposals including transaction conditions through a transaction brokerage method between the plurality of data demander servers, transmits a list of collected transaction proposals to the user terminal, and performs transaction completion management for the selected transaction proposal; and a payment unit that pays a consideration including economic or non-economic value according to the transaction conditions of a specific transaction proposal selected by the user terminal to a user account.

Description

Technology Field

[0001] The present invention relates to a system for providing a search data trading platform through user-centric real-time data value evaluation and transaction brokerage, which calculates a commercial value score of search data received from a user terminal owned by the user in real time and enables the user to actively select optimal conditions and receive direct compensation through transaction brokerage between multiple data demander servers. Background Technology

[0002] In the modern digital economy, personal data is recognized as a new form of economic asset. In particular, user search data is high-value information that directly reflects real-time intentions and interests, making it a crucial resource for search platform companies, advertisers, and data analytics firms. According to the global market research firm IDC, the global data market is projected to reach $220 billion by 2024, exhibiting an average annual growth rate of 12%. Notably, the market for high-value search data, which reflects users' real-time purchasing intentions, is recording an explosive growth rate of 31.5% annually, serving as a key driving force for the overall digital advertising market.

[0003] However, the current data ecosystem harbors serious structural imbalances. The most fundamental problem is data value asymmetry. While ordinary users, the actual producers of data, provide it for free without fully recognizing the economic value their data generates, platform operators collect and analyze this data to generate massive profits. In fact, global platform companies such as Google and Facebook record hundreds of billions of dollars in annual revenue solely from advertising based on user data, yet they provide no direct compensation to the individual users who actually produce the data. From the perspective of data sovereignty, this structure is clearly an unfair situation.

[0004] Existing search platforms adopt a business model that collects user-generated search data for free and converts it into advertising revenue. In this process, users can utilize search services for free, but they receive no compensation whatsoever for the direct economic value generated by their data. Platform companies build sophisticated user profiles based on the collected data and sell them to advertisers, generating hundreds of billions of dollars in annual revenue. For example, Google recorded $175 billion in search advertising revenue alone as of 2023, yet the billions of users who provided the search data that serves as the source of this revenue are completely excluded from this value creation process.

[0005] Advertisers, as consumers of data, are also enduring massive inefficiencies under the current system. According to a 2023 study by the American Advertisers Association (ANA), approximately 23% of programmatic advertising spending is wasted on low-quality "mad-for-advertising" websites (Source: ANA). A more serious issue is sophisticated digital advertising fraud, which is projected to result in annual losses exceeding $100 billion (Source: Juniper Research).

[0006] From the perspective of advertisers, the current system also exhibits several limitations. Traditional keyword advertising requires platform intermediation, which compromises data quality reliability and faces serious issues that hinder advertising efficiency. The most significant problem is organized ad fraud perpetrated by sophisticated bots or click farms. According to an analysis by Pixelate, an average of 11% of global desktop web traffic consists of Invalid Traffic (IVT) caused by bots (Source: Pixelate), demonstrating that a significant portion of advertisers' spending fails to reach actual potential customers.

[0007] While the introduction of Real-Time Bidding (RTB) technology into the online advertising market has significantly improved the efficiency of advertising transactions, users remain merely objects of the transaction, unable to become active participants. In RTB systems, advertisers conduct real-time auctions based on users' cookie information, but all revenue is monopolized by platforms and media companies, with no compensation provided to users, who are the actual producers of the data. Furthermore, RTB remains a one-sided structure where advertisers "target" users, providing users with absolutely no authority to actively participate in their data transactions or select transaction conditions.

[0008] Although blockchain-based data exchanges and personal information reward platforms have recently emerged, most of them adopt a method of accumulating data over a long period and then providing rewards in a lump sum. For example, platforms such as Ocean Protocol and Datum operate on a structure where token-based rewards are paid out weeks to months after users provide data. This approach fails to address the core challenges of real-time valuation of individual data and immediate monetization, and also has limitations in inducing active user participation.

[0009] While reward applications and survey platforms also provide compensation for user behavior, their compensation systems are based on fixed amounts predetermined by the platform, failing to reflect the actual market value of the data. Platforms like Swagbucks and InboxDollars offer fixed rewards of $1 to $5 for participating in surveys or watching ads, but this is significantly different from the advertising revenue actually generated by that data. Furthermore, most operate by inducing artificial behavior, which is far removed from the authenticity of natural user data.

[0010] Existing systems also exhibit distinct limitations in terms of data quality management. Since users do not receive direct incentives for providing data, there is a lack of motivation to generate high-quality data. Consequently, the market is flooded with low-quality data, creating a vicious cycle that undermines the reliability and efficiency of the entire data ecosystem.

[0011] This differs fundamentally from the core concept of the present invention, in which the user, as the producer of the data, actively sells their 'real-time search data' itself and directly selects and receives payment from multiple consumers, and has an obvious limitation in that it fails to return data sovereignty to the user.

[0012] These structural problems are difficult to resolve through technical improvements alone, and a fundamental paradigm shift is required to transform users from passive data providers into active market participants.

[0013] Therefore, conventional technologies as described above contain the following complex problems, necessitating a fundamental paradigm shift.

[0014] (1) Data value asymmetry and economic alienation of users: Users, who are the actual producers of data, are completely alienated from the enormous economic value (hundreds of billions of dollars annually) generated by their data, which is a problem of structural unfairness.

[0015] (2) Inefficiency and distrust among data users: Data users, such as advertisers, pay high commissions (30-50%) to the platform, but they suffer from a decline in advertising efficiency (an average loss of 37%) due to predictive targeting based on historical data and low-quality data.

[0016] (3) Failure to reflect real-time value in existing reward models: Existing reward platforms fail to reflect the real-time market value of individual data and provide a fixed, uniform reward (at the level of 1-5 dollars), which causes a discrepancy between the actual value of the data and the reward.

[0017] (4) Vicious cycle of degraded data quality: Since users do not receive direct and fair compensation for providing data, there is a lack of motivation to generate high-quality data, which ultimately leads to a vicious cycle that lowers the credibility of the entire data ecosystem.

[0018] (5) Absence of user data sovereignty: Existing systems treat users only as passive data providers and have a fundamental limitation in that they fail to provide actual data sovereignty, such as the right to actively participate in transactions regarding their data, the right to determine prices, and the right to choose transaction conditions.

[0019] Therefore, innovative methods to solve these problems are urgently needed. Prior art literature

[0020] Korean Published Patent No. 10-2010-0044394 The problem to be solved

[0021] The present invention is devised to solve the problems of the aforementioned prior art and aims to provide a fair data trading mechanism that enables users, who are the actual producers of data, to recognize the value of their search data in real time and actively trade it.

[0022] In addition, the present invention aims to improve the efficiency and reliability of the entire data ecosystem by accurately reflecting the market value of individual data through real-time competition among multiple data users and by inducing the production of high-quality data through a quality management system.

[0023] The problems of the present invention are not limited to those mentioned above, and other unmentioned problems will be clearly understood by those skilled in the art from the description below. means of solving the problem

[0024] The system for providing a search data trading platform through user-centric real-time data value assessment and transaction brokerage according to the present invention for achieving the above-mentioned purpose may include: a data receiving unit that receives search data from a user terminal owned by a user; a value analysis unit that calculates a commercial value score of the search data received through the data receiving unit in real time; a transaction brokerage unit that generates a transaction request based on the commercial value score calculated by the value analysis unit and transmits it to a plurality of data demander servers, collects transaction proposals including transaction conditions through a transaction brokerage method between the plurality of data demander servers, transmits a list of collected transaction proposals to the user terminal, and performs transaction completion management for the selected transaction proposal; and a consideration payment unit that pays consideration including economic or non-economic value according to the transaction conditions of a specific transaction proposal selected by the user terminal to a user account.

[0026] **Specific implementation of the Value Analysis Department

[0027] In the present invention, the value analysis unit may perform a quality index calculation algorithm comprising the steps of: (a-1) executing a machine learning-based value analysis model; (a-2) calculating individual scores according to a plurality of pre-set evaluation criteria for the search data; and (a-3) determining a commercial value score of the search data by summing the individual scores calculated in step (a-2).

[0028] In particular, the above step (a-2) applies a multi-layered evaluation system to accurately measure the commercial value of the search data. Specifically, it includes step (a-2-1), which assigns a first threshold score if the search data contains a general product name; step (a-2-2), which assigns a second threshold score if the search data contains specific identification information regarding the general product name; and step (a-2-3), which assigns a third threshold score if the search data contains keywords indicating purchase intent. Through this stepwise evaluation method, the specificity and commercial potential of the search terms can be accurately quantified.

[0030] ** Comprehensive transaction mechanism of the transaction brokerage

[0031] The above transaction brokerage unit may perform a transaction brokerage integration algorithm comprising: (b-1) generating a transaction request including an anonymized data identifier based on the commercial value score to create an efficient and fair transaction environment between data demanders and users; (b-2) transmitting the transaction request in parallel to the plurality of data demander servers; (b-3) collecting transaction proposals including transaction conditions from the plurality of data demander servers; (b-4) sorting the collected plurality of transaction proposals in descending order according to the transaction conditions; (b-5) transmitting the list of transaction proposals sorted in step (b-4) to the user terminal; (b-6) receiving a selection signal for a specific transaction proposal from the user terminal; and (b-7) generating a unique transaction identification number for the selection signal.

[0032] At this stage, in step (b-1), anonymization processing is performed to fully protect the user's personal information while accurately conveying information regarding the commercial value of the search data; in step (b-2), a fair and competitive market environment is created by providing trading opportunities to multiple data demanders simultaneously for time efficiency. In step (b-3), various forms of trading proposals are collected through a reverse auction method, an AI-based automatic matching method, a fixed-price bidding method, or a hybrid method combining these; and in step (b-4), optimal trading options are presented to the user by applying a multi-criteria decision-making algorithm.

[0034] **Guarantee of User Sovereignty by Transaction Brokers

[0035] The above transaction brokerage unit processes and provides the transaction proposals collected in step (b-5) in an optimized form that the user can easily understand and compare, in order to grant the user complete control over data transactions, supports the user's voluntary and informed selection in step (b-6), and establishes a systematic tracking management system for the selected transactions in step (b-7).

[0036] Specifically, in step (b-5), information comprehensively analyzing the consideration amount, payment method, expected payment timing, reliability of data users, and additional benefits of each transaction proposal is provided through responsive web design, and the strengths and weaknesses of each proposal are objectively analyzed and presented using a machine learning-based recommendation system. In step (b-6), digital signature technology and a two-factor authentication system ensure that user selections are processed safely and validly, and in step (b-7), the transparency and traceability of transactions are fully secured by combining the UUID v4 standard with the blockchain hash chain method.

[0037] Through this, users can transparently verify all transaction conditions presented for their search data and exercise complete transaction sovereignty by actively selecting optimal conditions suited to their personal preferences and circumstances. Furthermore, the transaction brokerage unit flexibly supports various trading methods, such as reverse auctions, forward auctions, AI automatic matching, and fixed-price trading, enabling the adaptive application of optimal trading mechanisms based on market conditions and user requirements.

[0039] Settlement system of the payment department

[0040] The above payment unit may perform an algorithm for processing consideration including economic or non-economic value, comprising: (c-1) a step of calculating the amount of consideration including economic or non-economic value by analyzing the transaction conditions of a specific transaction proposal selected from the user terminal to ensure accurate and rapid payment of consideration according to selected transaction conditions; (c-2) a step of verifying the validity of the user account and whether payment of consideration including economic or non-economic value is possible; (c-3) a step of paying the consideration including economic or non-economic value to the user account; and (c-4) a step of transmitting a notification signal indicating the completion of payment of consideration including economic or non-economic value to the user terminal and storing a transaction record.

[0041] In addition, the payment unit receives evidence proving subsequent purchasing activity from the user terminal to verify the realization of the actual value of the data and to pay additional compensation, verifies the purchasing activity by analyzing the evidence, and can additionally pay additional compensation to the user account according to the verification result.

[0042] Specifically, an additional reward processing algorithm can be performed, comprising: a step (d-1) of receiving evidence materials proving subsequent purchase activity from the user terminal; a step (d-2) of verifying the purchase activity by analyzing the evidence materials received in step (d-1); a step (d-3) of determining a differentiated additional reward ratio based on the verification result of step (d-2); and a step (d-4) of paying additional rewards to the user account based on the ratio determined in step (d-3).

[0044] Ecosystem optimization by the Quality Control Department

[0045] The present invention may further include a quality management unit that manages a user-specific search data quality index in real time and dynamically adjusts the search data submission limit according to the quality index to ensure the quality improvement and sustainability of the entire data ecosystem.

[0046] The above quality management unit may perform a dynamic limit control algorithm comprising: a step (e-1) of querying the current quality index for the search data of a selected user; a step (e-2) of determining which grade among a plurality of grades classified according to a pre-set classification criterion the user belongs to based on the quality index queried in step (e-1); a step (e-3) of calculating a daily data submission limit based on the grade determined in step (e-2); and a step (e-4) of notifying the user terminal of the submission limit calculated in step (e-3).

[0047] Through this quality management system, more transaction opportunities can be provided to high-quality data providers, and the reliability and efficiency of the entire platform can be continuously improved by preventing the influx of low-quality data.

[0049] **Integrated System Operation

[0050] Each component of the present invention operates in an organically interconnected manner. When the data receiving unit securely receives the user's search data, the value analysis unit immediately quantifies the commercial value, and the transaction brokerage unit conducts comprehensive transaction brokerage with multiple data demanders based on this. The transaction brokerage unit systematically analyzes and sorts transaction proposals collected through various transaction methods (reverse auction, AI-based automatic matching, fixed-price bidding, hybrid methods, etc.) and presents them transparently to the user, thereby fully guaranteeing the right to make a final choice, while also integrally managing the completion of the selected transaction. The payment unit completes the transaction through prompt and accurate payment settlement based on the transaction information received from the transaction brokerage unit. Finally, the quality management unit promotes continuous quality improvement of the entire system to continuously enhance the reliability and efficiency of the transaction ecosystem.

[0051] Through such integrated operation, the present invention can establish an innovative data economy ecosystem where all participants can coexist by guaranteeing complete transaction sovereignty over their data and optimal compensation to users who are data producers, providing efficient access to high-quality data and various transaction options to companies that are data consumers, and providing a stable and scalable transaction brokerage system to platform operators. Effects of the invention

[0052] The system for providing a search data trading platform through user-centered real-time data value assessment and transaction brokerage according to the present invention, designed to solve the aforementioned problems, has the groundbreaking effect of solving the problems of the aforementioned prior art as follows.

[0053] (1) Solving the problem of data value asymmetry and economic alienation of users

[0054] The value analysis unit of the present invention accurately calculates the commercial value score of search data in real time through a machine learning-based value analysis model, and the transaction brokerage unit fundamentally solves the existing problem of data value asymmetry by allowing the user to directly select the highest price through a transaction brokerage mechanism between multiple data demanders.

[0055] This enables users, who are the actual producers of data, to recognize the economic value generated by their search data in real time and generate direct compensation, thereby having an innovative effect of fully realizing users' economic sovereignty in an annual data economy worth hundreds of billions of dollars.

[0056] (2) Solving the problems of inefficiency and distrust among data users

[0057] The present invention identifies high-quality search data in real time using multi-layered evaluation criteria including general product names, specific identification information, and purchase intent keywords through a quality index calculation algorithm of the value analysis unit, and enables the identification of user intent more accurately than conventional predictive targeting by utilizing a machine learning-based analysis model.

[0058] In addition, the additional compensation processing algorithm of the payment department verifies actual purchasing behavior through OCR technology, etc., thereby allowing data users to immediately access high-quality search data without platform brokerage fees (30-50%), significantly reducing advertising cost losses (average 37%) and enabling efficient marketing based on highly reliable data.

[0059] (3) Overcoming the problem of real-time value not being reflected in existing reward models

[0060] The present invention provides compensation that accurately reflects the market value of individual search data through a real-time transaction brokerage processing algorithm performed by a transaction brokerage unit, rather than a fixed amount (at the level of $1-5) predetermined by the platform.

[0061] The transaction selection management algorithm of the transaction brokerage department sorts multiple transaction proposals in descending order according to compensation conditions, allowing users to directly select the optimal conditions. Unlike existing uniform and standardized compensation systems, this offers the advantage of enabling differential payment based on the unique value of each data. This has an innovative effect of completely resolving the discrepancy between the actual commercial value of data and the compensation.

[0062] (4) Resolving the vicious cycle of data quality degradation

[0063] The present invention manages the search data quality index for each user in real time through a dynamic limit adjustment algorithm of the quality management department, and dynamically adjusts the search data submission limit according to the quality index to grant more transaction opportunities to high-quality data providers.

[0064] Furthermore, the additional compensation processing algorithm of the payment unit provides users with a strong and continuous incentive to generate high-quality data by paying differentiated additional compensation based on the verification results of evidence proving subsequent purchasing activities. This fundamentally improves the existing vicious cycle structure where low-quality data floods the market, thereby enhancing the efficiency and reliability of the entire data ecosystem.

[0065] (5) Full realization of user data sovereignty

[0066] The present invention transforms the user from a passive data provider into an active market participant, allowing the user to check a list of transaction proposals and directly select the optimal conditions, thereby fully guaranteeing the right to initiate a transaction, the right to check the price, and the right to select the final transaction for their data assets.

[0067] The transaction brokerage guarantees the transparency and traceability of all transactions by generating unique optional identification numbers, while the payment department establishes a reliable transaction environment through secure payment verification along with user account validation. This has significant socio-economic effects by technically realizing 'Data Sovereignty,' in which users exercise complete control and ownership over their own data.

[0069] **Overall ripple effect

[0070] As a result, the present invention dramatically improves the efficiency and transparency of the entire data trading market, breaks away from the existing platform-centric data monopoly structure, and realizes a new user-centric data economy ecosystem.

[0071] Through this, by providing users with the right to receive fair compensation for data, data consumers with immediate access to high-quality data, and the entire market with a fair and efficient value distribution mechanism, it has the outstanding effect of creating an innovative platform ecosystem where all participants in the global data market can coexist and prosper.

[0072] 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 in the claims. Brief explanation of the drawing

[0073] FIG. 1 is a diagram showing each component of a system for providing a search data trading platform through user-centered real-time data value assessment and transaction brokerage according to an embodiment of the present invention. FIG. 2 is a diagram showing each process of a quality index calculation algorithm performed by a value analysis unit in a system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention. FIG. 3 is a diagram showing the detailed process of step (a-2) of the quality index calculation algorithm in a system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to one embodiment of the present invention. FIG. 4 is a diagram showing each process of the transaction brokerage processing and transaction selection management algorithm performed by the transaction brokerage unit in a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention. FIG. 5 is a diagram showing each process of an algorithm for processing compensation including economic or non-economic value performed by a compensation payment unit in a system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention. FIG. 6 is a diagram showing each process of an additional compensation processing algorithm performed by a payment unit in a system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention. FIG. 7 is a diagram showing each process of a dynamic limit adjustment algorithm performed by a quality management unit in a system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention. FIG. 8 is a diagram illustrating an exemplary search interface provided to a user terminal by a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention. FIGS. 9 and 10 are diagrams illustrating, in an exemplary manner, commercial value scores based on differences in search terms in a search interface provided to a user terminal by a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention. FIG. 11 is a diagram exemplarily illustrating the process of payment including economic or non-economic value in a search interface provided to a user terminal by a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention. FIG. 12 is a diagram exemplarily illustrating the process of providing additional rewards in a search interface provided to a user terminal by a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention. Specific details for implementing the invention

[0074] In this specification, where a component (or region, layer, part, etc.) is described as being "on," "connected," or "combined" with another component, it means that it may be directly placed / connected / combined with the other component, or that a third component may be placed between them.

[0075] Identical reference numerals denote identical components. Additionally, in the drawings, the thicknesses, proportions, and dimensions of the components are exaggerated for the effective illustration of the technical content.

[0076] "And / or" includes all one or more combinations that the associated configurations can define.

[0077] Terms such as "first," "second," etc., may be used to describe various components, but said components should not be limited by said terms. These terms are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be named the second component, and similarly, the second component may be named the first component. A singular expression includes a plural expression unless the context clearly indicates otherwise.

[0078] Additionally, terms such as "below," "lower side," "above," and "upper side" are used to describe the relationships between the components depicted in the drawings. These terms are relative concepts and are described based on the directions indicated in the drawings.

[0079] Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as generally understood by those skilled in the art to which the present invention pertains. Additionally, terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and are explicitly defined herein unless interpreted in an ideal or overly formal sense.

[0080] Terms such as "include" or "have" are intended to indicate 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.

[0081] Furthermore, when it is stated in this specification that a first component operates or is executed on (ON) a second component, it should be understood that the first component operates or is executed in an environment where the second component operates or is executed, or operates or is executed through direct or indirect interaction with the second component.

[0082] Where any component, device, or system is described as including a component consisting of a program or software, it should be understood that, even without explicit mention, that component, device, or system includes hardware (e.g., memory, CPU, etc.) or other programs or software (e.g., an operating system or drivers required to run the hardware) necessary for the execution or operation of that program or software.

[0083] Furthermore, unless otherwise specified regarding the implementation of a component, it should be understood that the component may be implemented in software, hardware, or in any form that combines both software and hardware.

[0084] Furthermore, the terms used herein are for describing the embodiments and are not intended to limit the invention. In this specification, the singular form includes the plural form unless specifically stated otherwise in the text. As used herein, 'comprises' and / or 'comprising' do not exclude the presence or addition of one or more other components to the mentioned components.

[0085] Additionally, in this specification, terms such as 'part', 'device', etc., may be intended to refer to hardware and the functional and structural combination of software driven by said hardware or for driving said hardware. For example, the hardware here may be a data processing device including a CPU or other processor. Furthermore, the software driven by the hardware may refer to a running process, object, executable, thread of execution, program, etc.

[0086] Furthermore, it can be easily inferred by an average expert in the art of the present invention that the above terms may refer to a specific code and a logical unit of hardware resources for executing the said specific code, and do not necessarily refer to physically connected code or a single type of hardware.

[0088] **Overall System Structure

[0089] FIG. 1 is a diagram showing each component of a system for providing a search data trading platform through user-centered real-time data value assessment and transaction brokerage according to an embodiment of the present invention.

[0090] As illustrated in FIG. 1, in this embodiment, a system for providing a search data trading platform through user-centered real-time data value assessment and transaction brokerage may include a platform providing server (100), a data receiving unit (110), a value analysis unit (120), a transaction brokerage unit (130), a payment unit (140), and a quality management unit (150). At this time, the platform providing server (100) may perform the role of mediating between a user terminal (10) and a data demander server (20).

[0092] **User terminal (10) configuration

[0093] The user terminal (10) enables the user to input search data and communicate with the platform providing server (100). For example, the user terminal (10) may include a smartphone, a tablet PC, a laptop computer, or a desktop computer. However, the user terminal (10) may be implemented in various forms and is not limited to such forms.

[0094] All data communication between the user terminal (10) and the platform providing server (100) is securely performed using an encryption protocol of TLS (Transport Layer Security) 1.3 or higher, thereby fundamentally preventing theft or tampering during the data transmission process. In addition, the user's search data is transmitted using AES-256 encryption, and user authentication is performed using an OAuth 2.0-based token method.

[0096] **Data user server (20) configuration

[0097] Multiple data demander servers (20) may be configured as server systems operated by search platform companies, advertisers, data analysis companies, etc. Such data demander servers (20) can perform the function of receiving a transaction request from a platform provider server (100) and transmitting a transaction proposal including transaction conditions.

[0098] And the data demand server (20) can be implemented as a distributed system structure including, for example, a web server, an application server, and a database server. However, the data demand server (20) can be implemented in various forms and is not limited to only such forms.

[0100] **Configuration of platform providing server (100)

[0101] The platform providing server (100) can perform the role of mediating search data transactions between a user terminal (10) and a plurality of data demander servers (20). The platform providing server (100) can be physically implemented as a single server or multiple servers and can be equipped to operate in a cloud-based distributed computing environment.

[0103] **Specific implementation of the data receiving unit (110)

[0104] The data receiving unit (110) can receive search data from a user terminal (10) owned by the user. That is, the data receiving unit (110) can communicate with the user terminal (10) via a network to receive and store search terms or search-related information entered by the user.

[0105] The received search data may include keywords in text form, search sentences, or various forms of information indicating search intent. The data receiving unit (110) can verify the integrity of the data during the reception process and verify that the data was transmitted from a legitimate user through user authentication.

[0106] Additionally, the data receiving unit (110) can decrypt the search data transmitted in an encrypted form for transmission security and convert it into a processable form, and can perform the role of transmitting the received search data to the value analysis unit (120) for subsequent processing.

[0107] The data receiving unit (110) can also efficiently process large volumes of search data by utilizing a message queue system for real-time data processing, and can reliably receive and transmit tens of thousands of search data per second by linking with a distributed streaming platform such as Apache Kafka.

[0109] **Detailed implementation of the value analysis unit (120)

[0110] The value analysis unit (120) can calculate the commercial value score of the search data received through the data receiving unit (110) in real time. That is, the value analysis unit (120) can analyze the received search data and quantify the commercial value that the data can have for the data demander servers (20).

[0112] Quality Index Calculation Algorithm

[0113] FIG. 2 is a diagram showing each process of a quality index calculation algorithm performed by a value analysis unit (120) in a system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention.

[0114] As illustrated in FIG. 2, in this embodiment, the value analysis unit (120) can perform a quality index calculation algorithm comprising a step (a-1) of executing a machine learning-based value analysis model, a step (a-2) of calculating individual scores according to a plurality of evaluation criteria set in advance for search data, and a step (a-3) of determining a commercial value score of search data by summing the individual scores calculated in step (a-2).

[0116] ** (a-1) Step: Run the machine learning model

[0117] Step (a-1) is the process of loading and executing a pre-trained machine learning-based value analysis model. In this step (a-1), the artificial intelligence model can extract and analyze features of the search data and perform predictive analysis based on past data patterns.

[0118] The above machine learning-based value analysis model can be trained using a training dataset consisting of millions of anonymized 'search term-product information-final purchase status' data pairs. The training of the model is performed with the goal of predicting the probability that the search term will lead to an actual purchase or the expected transaction price within a certain period of time by receiving the textual characteristics of a given search term ('product name recognition rate (P1)', 'specificity index (D2)', and 'intention keyword density (I3)').

[0119] In addition, the machine learning-based value analysis model of step (a-1) in this embodiment can be implemented by combining a Transformer-based BERT (Bidirectional Encoder Representations from Transformers) model, a decision tree, and a support vector machine in an ensemble form. In particular, the BERT model is used to understand the contextual meaning of a search term, the decision tree is responsible for clear rule-based classification, and the support vector machine is responsible for complex pattern recognition.

[0121] **(a-2) Step: Application of multi-layered evaluation criteria

[0122] FIG. 3 is a diagram showing the detailed process of step (a-2) of the quality index calculation algorithm in a system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to one embodiment of the present invention.

[0123] As illustrated in FIG. 3, in this embodiment, step (a-2) may include step (a-2-1) of assigning a first standard score when the search data contains a general product name, step (a-2-2) of additionally assigning a second standard score when the search data contains specific identification information for the general product name, and step (a-2-3) of additionally assigning a second standard score when the search data contains keywords indicating purchase intent.

[0125] **(a-2-1) Step: General Product Name Evaluation

[0126] Step (a-2-1) is the process of assigning a first threshold score if the search data contains a general product name. In this step (a-2-1), it can be determined whether the search term includes a general name for a product or service.

[0127] That is, in this embodiment, step (a-2-1) can identify whether a general product name is included through Named Entity Recognition technology utilizing the Stanford CoreNLP library and matching with a product category database, and can assign a first threshold score if a major category product name, such as electronic products, clothing, food, or cosmetics, is included in the search term. The first threshold score is a score representing the commercial base value of the search term, and can be set, for example, within a range of 30% to 50% based on a pre-set maximum score.

[0129] **(a-2-2) Step: Evaluation of Specific Identification Information

[0130] Step (a-2-2) is the process of assigning an additional second threshold score if the search data contains specific identifying information regarding a general product name. In other words, Step (a-2-2) can evaluate whether specific identifying information, such as brand name, model name, color, and size, is included in addition to the simple product name.

[0131] In particular, in this embodiment, step (a-2-2) can determine whether specific identification information is included through regular expression pattern matching and cross-validation with a brand database and a product specification database. Therefore, step (a-2-2) can evaluate commercial value highly by additionally assigning a second threshold score when the search term indicates a clearer purchase intent. For example, the second threshold score can be set within a range of 30% to 50% based on a pre-set maximum score.

[0133] **(a-2-3) Step: Evaluate Purchase Intent Keywords

[0134] Step (a-2-3) is the process of assigning an additional third threshold score if the search data contains keywords indicating purchase intent. In this step (a-2-3), it is possible to analyze whether keywords indicating direct purchase intent, such as purchase, order, price, discount, and delivery, are included.

[0135] In addition, in this embodiment, step (a-2-3) can search for the relevant keywords within the search term using a purchase intent keyword dictionary utilizing the TF-IDF (Term Frequency-Inverse Document Frequency) vectorization technique, and calculate a third reference score based on the number and importance of the matching keywords. Furthermore, step (a-2-3) can distinguish expressions indicating urgency ("today," "immediately"), expressions indicating comparative purchase ("vs," "comparison"), and expressions indicating price sensitivity ("lowest price," "discount") and apply different weights to them.

[0137] **(a-3) Step: Determine Final Value Score

[0138] Step (a-3) is the process of determining the commercial value score of the search data by summing the individual scores calculated in Step (a-2). In Step (a-3), the final commercial value score can be derived by selectively applying pre-set weights to the individual scores calculated for each evaluation criterion.

[0139] For example, step (a-3) may perform score integration based on a weighted average method, but using a linear combination method that is dynamically adjusted according to market trends. Additionally, the commercial value score determined in step (a-3) can be transmitted to the transaction brokerage unit (130) and used as basic data for generating transaction requests.

[0141] **Key technical implementation of the value analysis department (120)

[0142] 1. Technical Overview

[0143] The value analysis unit (120) is a core component that calculates the commercial value score of the search data received through the data receiving unit (110) in real time. The value analysis unit (120) of the present invention precisely quantifies the commercial value that the search data can have for data demander servers (20) through the following quantitative mathematical model, rather than simple keyword matching.

[0144] 2. Formula for Calculating the Total Commercial Value Score

[0145] 2.1 Main Formula

[0146] CVS_final(q) = w₁S₁ + w₂S₂ + w₃S₃ + w₄ML_Score(q)

[0147] 2.2 Variable Definition

[0148] CVS_final(q): Final commercial value score of search term q (0 ≤ CVS_final(q) ≤ 100)

[0149] Weight variables (w₁, w₂, w₃, w₄): Coefficients representing the importance of each evaluation item

[0150] * w₁ = 0.3 (Product name weight)

[0151] * w₂ = 0.4 (specificity weight)

[0152] * w₃ = 0.2 (purchase intention weight)

[0153] * w₄ = 0.1 (ML model calibration weight)

[0154] Constraint: w₁ + w₂ + w₃ + w₄ = 1.0

[0155] Evaluation score variable:

[0156] * S₁: 1st threshold score (General product name evaluation, 0 ≤ S₁ ≤ 100)

[0157] * S₂: Second threshold score (Evaluation of specific identification information, 0 ≤ S₂ ≤ 100)

[0158] * S₃: Third threshold score (Purchase intent keyword evaluation, 0 ≤ S₃ ≤ 100)

[0159] * ML_Score(q): Machine learning model analysis score (0 ≤ ML_Score(q) ≤ 100)

[0160] 3. Mathematical Definition of the Quality Index Calculation Algorithm

[0161] 3.1 (a-1) Step: Running the Machine Learning Model

[0162] Ensemble model formula

[0163] ML_Score(q) = α₁ × BERT(q) + α₂ × DT(q) + α₃ × SVM(q)

[0164] Variable Definition:

[0165] * BERT(q): Semantic analysis score of the BERT model (0 ≤ BERT(q) ≤ 100)

[0166] * DT(q): Rule-based classification score of the decision tree (0 ≤ DT(q) ≤ 100)

[0167] * SVM(q): Support Vector Machine pattern recognition score (0 ≤ SVM(q) ≤ 100)

[0168] * α₁ = 0.5, α₂ = 0.3, α₃ = 0.2 (ensemble weights)

[0169] Constraint: α₁ + α₂ + α₃ = 1.0

[0170] 3.2 (a-2) Step: Application of multi-layered evaluation criteria

[0171] 3.2.1 (a-2-1) Step: General Product Name Evaluation Formula

[0172] S₁ = min((P₁ × M₁ × C₁), 1.0) × 100

[0173] Variable Definition:

[0174] * S₁: First cutoff score (0 ≤ S₁ ≤ 100)

[0175] * P₁: Product name recognition rate = Σ(Number of matched product names) / Total number of tokens (0 ≤ P₁ ≤ 1)

[0176] * M₁: Category matching coefficient

[0177] o 1.0 (complete matching)

[0178] o 0.7 (partial matching)

[0179] o 0 (Missing)

[0180] * C₁: Commercial importance coefficient by category

[0181] o Electronic products: 1.2

[0182] o Fashion: 1.0

[0183] o Food: 0.8

[0184] Calculation Example: Search Term: "Samsung Galaxy Smartphone"

[0185] * P₁ = 2(Product Name: Galaxy, Smartphone) / 3(Total Tokens) = 0.67

[0186] * M₁ = 1.0 (complete matching)

[0187] * C₁ = 1.2 (electronic products)

[0188] S₁ = min(0.67 × 1.0 × 1.2, 1.0) × 100 = min(0.804, 1.0) × 100 = 80.4 points

[0189] 3.2.2 Step (a-2-2): Specific Identification Information Evaluation Formula

[0190] S₂ = min((D₂ × R₂ × B₂), 1.0) × 100

[0191] Variable Definition:

[0192] * S₂: Second cutoff score (0 ≤ S₂ ≤ 100)

[0193] * D₂: Specificity Index = (Number of detected identifiers) / (Maximum number of identifiers) (0 ≤ D₂ ≤ 1)

[0194] * R₂: Regular expression matching rate = Number of matched patterns / Total number of applied patterns (0 ≤ R₂ ≤ 1)

[0195] * B₂: Brand Reliability Coefficient

[0196] o Premium Brand: 1.5

[0197] o General Brand: 1.0

[0198] Calculation Example: Search Term: "iPhone 15 Pro 128GB Gold"

[0199] * D₂ = 4(iPhone, 15 Pro, 128GB, Gold) / 4 = 1.0

[0200] * R₂ = 4(Brand, Model, Capacity, Color) / 4 = 1.0

[0201] * B₂ = 1.5 (Premium Brand)

[0202] * S₂ = min(1.0 × 1.0 × 1.5, 1.0) × 100 = 1.0 × 100 = 100 points

[0203] 3.2.3 Step (a-2-3): Purchase Intent Keyword Evaluation Formula

[0204] S₃ = min(I₃ × TF-IDF₃ × U₃, 1.0) × 100

[0205] Variable Definition:

[0206] * S₃: Third cutoff score (0 ≤ S₃ ≤ 100)

[0207] * I₃: Intent Keyword Density = Number of Purchase Intent Keywords / Total Number of Keywords (0 ≤ I₃ ≤ 1)

[0208] * TF-IDF₃: Weighted Frequency Score = Σ(Keyword's TF-IDF Value)

[0209] * U₃: Urgency multiplier

[0210] o 1.0 (General)

[0211] o 1.5 (if urgency keywords exist)

[0212] TF-IDF Calculation Formula:

[0213] TF-IDF(keyword) = TF(keyword) Х log(N / DF(keyword))

[0214] Variable Definition:

[0215] * TF: Keyword frequency within search terms

[0216] * N: Total number of documents (search database size)

[0217] * DF: Number of documents containing the keyword

[0218] Calculation Example: Search Term: "Lowest iPhone purchase price today"

[0219] * I₃ = 2(purchase, lowest price) / 4 = 0.5

[0220] * TF-IDF₃ = TF-IDF (Purchase) + TF-IDF (Lowest Price) = 2.1 + 1.8 = 3.9

[0221] * U₃ = 1.0 + 0.5(existence of 'today') = 1.5

[0222] * S₃ = min(0.5 × 3.9 × 1.5, 1.0) × 100 = min(2.925, 1.0) × 100 = 1.0 × 100 = 100 points

[0223] 3.3 Step (a-3): Determining the Final Value Score

[0224] Calculation of basic final score

[0225] CVS_final = w₁S₁ + w₂S₂ + w₃S₃ + w₄S₄

[0226] Here, S₄ = ML_Score(q)

[0227] Dynamic adjustment reflecting market trends

[0228] CVS_adjusted = min(CVS_final Х Market_Trend_Factor, 100)

[0229] Market_Trend_Factor = 1.0 + (Current_Category_Demand_Index - 1.0) × 0.2

[0230] Adjustment Example:

[0231] * When demand for electronic products surges: Market_Trend_Factor = 1.0 + (1.5 - 1.0) × 0.2 = 1.1

[0232] * Final adjusted score: CVS_adjusted = CVS_final × 1.1

[0233] 4. Specific Integrated Calculation Example

[0234] Search term: "Lowest price to buy iPhone 15 Pro 128GB"

[0235] Step-by-step calculation results

[0236] * Step (a-1) Result: S₄ = ML_Score = 85 points

[0237] * (a-2-1) Step: S₁ = 90 points (Product name clear)

[0238] * (a-2-2) Step: S₂ = 100 points (Rich in specific identifying information)

[0239] * (a-2-3) Step: S₃ = 100 points (strong purchase intention)

[0240] Final calculation

[0241] CVS_final = (0.3 Х 90) + (0.4 Х 100) + (0.2 Х 100) + (0.1 Х 85)

[0242] = 27 + 40 + 20 + 8.5

[0243] = 95.5 points

[0244] Market adjustment applied

[0245] Reflecting the electronics boom: If CVS_final is 95.5 points, CVS_adjusted = min(95.5 × 1.1, 100) = 100 points.

[0246] 5. Technical Features and Effects

[0247] The quantitative formula-based approach of the present invention has the following technical features:

[0248] 1. Objectivity: The entire evaluation process is defined by mathematical formulas, excluding subjective judgment.

[0249] 2. Reproducibility: Guarantees that the same result is always produced for the same input.

[0250] 3. Scalability: Flexible expansion possible through weight adjustments when adding new evaluation criteria

[0251] 4. Precision: Detailed differentiation of commercial value through a multi-layered evaluation system

[0252] This quantitative formula-based approach is a core technical feature of the present invention that is clearly distinct from conventional abstract keyword analysis methods, and enables the measurement of the commercial value of search data in an objective and reproducible manner.

[0254] ** Comprehensive transaction system of the transaction brokerage department (130)

[0255] The transaction brokerage unit (130) can perform the role of comprehensively brokering search data transactions between multiple data demander servers (20) and user terminals (10) based on commercial value scores calculated by the value analysis unit (120). That is, the transaction brokerage unit (130) can collect transaction proposals including transaction conditions from data demanders through various transaction methods, and transmit the list of collected transaction proposals to the user terminal (10) to provide a brokerage service so that the user can select the optimal transaction conditions.

[0257] **Transaction Brokerage Integrated Algorithm

[0258] FIG. 4 is a diagram showing each process of a transaction brokerage integrated algorithm performed by a transaction brokerage unit (130) in a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention.

[0259] As illustrated in FIG. 4, in this embodiment, the transaction broker (130) can perform a transaction brokerage integration algorithm comprising: (b-1) a step of generating a transaction request including an anonymized data identifier based on a commercial value score; (b-2) a step of transmitting the transaction request in parallel to a plurality of data demander servers (20); (b-3) a step of collecting transaction proposals including transaction conditions from a plurality of data demander servers (20); (b-4) a step of sorting the collected plurality of transaction proposals in descending order according to transaction conditions; (b-5) a step of transmitting the list of transaction proposals sorted in step (b-4) to a user terminal (10); (b-6) a step of receiving a selection signal for a specific transaction proposal from the user terminal (10); and (b-7) a step of generating a unique transaction identification number for the selection signal.

[0261] **(b-1) Step: Create anonymized transaction request

[0262] Step (b-1) is a process of generating a transaction request that includes an anonymized data identifier based on a commercial value score. In this step (b-1), the commercial value score received from the value analysis unit (120) can be included in the transaction request message, and an anonymized data identifier can be generated to protect the user's personal information.

[0263] In this embodiment, the anonymized data identifier of step (b-1) is generated using the SHA-256 hash function and can be configured to include the content and value information of the search data while completely removing the user's actual identity. Additionally, a unique UUID (Universally Unique Identifier) ​​is assigned to each data to ensure traceability while protecting personal information.

[0264] When generating a transaction request in step (b-1), the transaction brokerage unit (130) may select and apply at least one of the following methods: a reverse auction method, an AI-based automatic matching method, a fixed price offering method, or a hybrid method combining these. For example, in the case of a reverse auction method, a transaction request is generated to induce data demanders to compete with each other and offer a higher price, and in the case of an AI-based automatic matching method, a transaction request for optimal matching is generated by analyzing the past transaction patterns and preferences of each data demander.

[0266] **(b-2) Step: Parallel transmission system

[0267] Step (b-2) is a process of transmitting transaction requests in parallel to multiple data demander servers (20). In Step (b-2), transaction requests can be transmitted simultaneously to multiple data demander servers (20) for time efficiency, and the network status and response time of each server can be monitored in real time.

[0268] For example, in this embodiment, step (b-2) can handle thousands of concurrent connections by utilizing a Node.js event loop based on asynchronous I / O (Asynchronous Input / Output), and can ensure stable communication through the HTTP / 2 protocol and Keep-Alive connection. Additionally, step (b-2) can learn the past response patterns of each data demander server (20) using machine learning and send to the server with a high probability of a response first.

[0269] The transaction broker (130) can transmit customized transaction requests by utilizing the preference profiles of data consumers in step (b-2). To this end, the transaction broker (130) can build a preference database that analyzes the search keyword patterns, preferred price ranges, and transaction success rates that each data consumer server (20) has previously shown interest in, and can generate and transmit the most suitable form of transaction request to each data consumer based on this.

[0271] **(b-3) Step: Collection and Verification of Deal Proposals**

[0272] Step (b-3) is a process of collecting transaction proposals including transaction conditions from multiple data demander servers (20). In this step (b-3), information such as the amount of consideration, payment method, special benefits, service conditions, and expected time of transaction completion proposed by each data demander can be systematically collected.

[0273] In addition, in this embodiment, step (b-3) can process all transaction proposals received within a set deadline (typically 500ms to 2 seconds) in real time and verify their validity through JSON schema validation, and can automatically filter out invalid responses or responses that do not match the format.

[0274] The transaction brokerage unit (130) can perform multidimensional analysis on the transaction proposals collected in step (b-3). Specifically, it can calculate a comprehensive score for each transaction proposal by comprehensively evaluating the immediacy (time of payment), reliability (past transaction fulfillment rate of data users), added value (additional benefits or services), and risk (complexity of transaction conditions) of each transaction proposal.

[0276] **(b-4) Step: Intelligent Alignment System

[0277] Step (b-4) is the process of sorting the collected multiple transaction proposals in descending order according to transaction conditions. In this step (b-4), the proposals can be arranged from the highest amount to the lowest amount based on the compensation offered by each data demander.

[0278] Furthermore, step (b-4) may not only sort by simple amount but also apply a Multi-Criteria Decision Making (MCDM) algorithm to calculate an integrated score that comprehensively considers payment conditions, additional benefits, and reliability scores. Additionally, personalized sorting criteria can be applied by analyzing the user's past selection patterns using Collaborative Filtering techniques.

[0279] The transaction brokerage unit (130) can reflect user preferences when sorting transaction proposals in step (b-4). For example, for a user who prefers quick payment, transaction proposals with immediate payment conditions can be placed at the top; for a user who prefers high compensation, they can be sorted by compensation amount; and for a user who values ​​stability, transaction proposals from data demanders with high reliability can be placed first.

[0281] **(b-5) Step: Provide an optimized user interface

[0282] Step (b-5) is the process of transmitting the list of trade proposals sorted in Step (b-4) to the user terminal (10). That is, in Step (b-5), the list of trade proposals sorted can be processed and transmitted in a form that the user can easily understand and compare.

[0283] For example, step (b-5) can clearly display the consideration amount, expected payment time, data demander information, and special conditions of each transaction proposal through responsive web design, and can be provided as an optimized interface tailored to the screen size and resolution of the user terminal (10). In addition, the advantages and disadvantages of each transaction proposal can be automatically analyzed through a machine learning-based recommendation system and recommendation information can be provided to the user.

[0284] The transaction brokerage unit (130) can provide detailed analysis information for each proposal along with a list of transaction proposals in step (b-5). This analysis information may include the data user's past transaction fulfillment rate, average payment time, user satisfaction evaluation, history of providing additional benefits after the transaction, etc., and supports the user in making a more informed decision.

[0286] **(b-6) Step: Receive safe selection signal

[0287] Step (b-6) is the process of receiving a selection signal for a specific transaction proposal from the user terminal (10). In this step (b-6), when the user selects a desired condition from the list of presented transaction proposals, the corresponding selection information can be safely received.

[0288] In addition, in this embodiment, step (b-6) utilizes digital signature technology to verify whether the user's choice is valid and can check in real time whether the selected transaction proposal remains valid. Furthermore, it can verify that the user's choice is legitimate through a two-factor authentication system.

[0289] The transaction broker (130) can immediately send a transaction confirmation signal to the corresponding data demander server (20) after receiving the user's selection in step (b-6). At this time, the transaction broker (130) can prevent duplicate transactions by sending a transaction closing notification to other data demanders, and can proceed with a procedure to finally confirm the detailed transaction conditions with the selected data demander.

[0291] **(b-7) Step: Generation of Unique Identifiers and Transaction Management

[0292] Step (b-7) is the process of generating a unique transaction identification number for the selection signal. Specifically, in Step (b-7), a number capable of uniquely identifying each transaction is generated so that it can be used for future transaction tracking and dispute resolution.

[0293] For example, the unique transaction identification number in step (b-7) is generated according to the UUID v4 standard and can be created by combining a timestamp, a user hash value, and a transaction proposal identifier. Additionally, data integrity and tamper prevention can be guaranteed by applying a blockchain hash chain method.

[0294] The transaction brokerage unit (130) can systematically manage the entire transaction process using the unique transaction identification number generated in step (b-7). This allows tracking all steps from the start to the completion of the transaction, and, if necessary, allows viewing transaction details or using them as evidence in the event of a dispute. Additionally, after the transaction is completed, performance data of the transaction can be collected and used to improve the transaction brokerage algorithm in the future.

[0296] **Support for multiple transaction methods of the transaction brokerage department (130)

[0297] The transaction brokerage unit (130) can flexibly support various transaction methods based on the integrated algorithm described above. Specifically, the transaction brokerage unit (130) can broker transactions using at least one of a reverse auction method, an AI-based automatic matching method, a fixed price offering method, or a hybrid method combining these.

[0298] In the case of a reverse auction method, the transaction brokerage unit (130) can create a competitive transaction environment that induces data demanders to compete with each other and offer higher prices. At this time, the transaction request generated in step (b-1) may include information such as the auction closing time, minimum transaction unit, and current highest price, and in step (b-3), the transaction status that is updated in real time can be collected to create a competitive atmosphere.

[0299] In the case of an AI-based automatic matching method, the transaction broker (130) can automatically perform optimal matching by utilizing a machine learning algorithm to analyze the characteristics of the user's search data and the preferences of each data user. At this time, in step (b-1), a targeted transaction request based on the AI ​​analysis results is generated, and in step (b-3), transaction proposals can be collected preferentially from data users with high matching scores.

[0300] In the case of a fixed price presentation method, the transaction brokerage unit (130) calculates an appropriate price based on the commercial value score calculated by the value analysis unit (120) and presents it to several data consumers to select a trading partner on a first-come, first-served basis or based on additional conditions.

[0301] In the case of a hybrid method, the transaction broker (130) can utilize a combination of the above methods depending on the situation. For example, for high-value search data, a reverse auction method can be applied to secure the highest possible price, for general search data, AI automatic matching can be used to facilitate a quick transaction, and for data of a specific category, a fixed-price method can be applied to ensure a stable transaction.

[0303] **Real-time processing capability of the transaction brokerage unit (130)

[0304] The transaction broker (130) adopts a high-performance architecture for real-time data processing. Specifically, the transaction broker (130) can process tens of thousands of transaction requests per second simultaneously by linking with a distributed streaming platform such as Apache Kafka, and can guarantee response times in milliseconds through in-memory caching using Redis Cluster.

[0305] In addition, the transaction broker (130) is designed based on a microservices architecture, so that each functional module can be scaled independently. Automatic scaling in response to increased traffic is possible through Docker containers and Kubernetes orchestration, and the load can be efficiently distributed among multiple server instances through a load balancer.

[0307] **Security and privacy of the transaction brokerage department (130)

[0308] The transaction broker (130) applies strict security protocols to all transaction processes. All communication between the user and the data user is performed through encryption protocols of TLS 1.3 or higher, and the search data is protected by AES-256 encryption.

[0309] In addition, the transaction broker (130) applies a Zero Trust security model to apply end-to-end encryption to communication between all components. The user's personal information is processed using homomorphic encryption technology so that computation is possible even while encrypted, and fully complies with international personal information protection regulations such as GDPR, CCPA, and personal information protection laws.

[0310] The transaction brokerage department (130) also operates an AI-based abnormal transaction detection system to perform 24-hour real-time security monitoring. Through this, it can detect and block abuse, false data submission, manipulated search behavior, and bot-based invalid traffic in real time.

[0312] **Settlement system of the payment department (140)

[0313] The consideration payment unit (140) can pay consideration including economic or non-economic value according to the transaction conditions of a specific transaction proposal selected by the user terminal (10) to the user account.

[0314] Here, the term 'consideration including economic or non-economic value' encompasses not only direct economic consideration such as cash, points, and cryptocurrencies, but also all forms of consideration that provide tangible value to the user, such as digital coupons, service discount vouchers, game items, NFTs, platform tier upgrades, VIP benefits, and priority access rights. The aforementioned non-economic value includes, but is not limited to, all forms of digital assets, service rights, and special benefits that provide tangible utility to the user.

[0316] **Algorithm for processing compensation including economic or non-economic value

[0317] FIG. 5 is a diagram showing each process of an economic or non-economic value-inclusive payment processing algorithm performed by a payment payment unit (140) in a system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention.

[0318] As illustrated in FIG. 5, in this embodiment, the payment unit (140) can perform an economic or non-economic value payment processing algorithm comprising: a step (c-1) of calculating the amount of payment including economic or non-economic value by analyzing the transaction conditions of a specific transaction proposal selected from a user terminal (10); a step (c-2) of verifying the validity of the user account and whether payment of payment including economic or non-economic value is possible; a step (c-3) of paying the payment including economic or non-economic value to the user account; and a step (c-4) of transmitting a payment completion notification signal for the payment including economic or non-economic value to the user terminal (10) and storing a transaction record.

[0320] **(c-1) Step: Calculation of accurate compensation

[0321] Step (c-1) is a process of analyzing the transaction conditions of a specific transaction proposal selected from the user terminal (10) to calculate the amount of consideration including economic or non-economic value. In this step (c-1), the amount of consideration, payment method, and additional conditions presented in the transaction can be comprehensively analyzed based on the user's selection information received from the transaction brokerage unit (130).

[0322] For example, in this embodiment, step (c-1) may calculate the exact amount of consideration including economic or non-economic value to be actually paid by applying factors such as system operating costs (typically 3-5%), value-added tax, and discount benefits to the basic consideration amount. Additionally, the transparency and automation of the consideration calculation process can be ensured by utilizing Smart Contract technology.

[0324] **(c-2) Step: Comprehensive Validation

[0325] Step (c-2) is a process of verifying the validity of the user account and the possibility of payment of consideration, including economic or non-economic value. To this end, Step (c-2) can verify whether the user account is active and whether account information or payment methods are correctly registered.

[0326] For example, Step (c-2) can verify the legality of payment through various verification items, including the Know Your Customer (KYC) authentication status, identity verification completion, Anti-Money Laundering (AML) compliance, and whether the transaction is subject to legal sanctions. Additionally, it can automatically detect and block abnormal transaction patterns through a Real-time Fraud Detection System.

[0328] **(c-3) Step: Safe Payment

[0329] Step (c-3) is the process of paying consideration, including economic or non-economic value, to the user account. In this step (c-3), the amount of consideration, including economic or non-economic value, calculated can be actually transferred or credited to the verified user account.

[0330] For example, step (c-3) can support various forms of payment, including cash payment, digital point accumulation, coupon issuance, and cryptocurrency transfer. In particular, the payment unit (140) can safely pay the payment to the user's preferred payment method (e.g., Kakao Pay, Toss Pay, PayPal, etc.) by linking with the API of a Payment Gateway (PG) operator, in addition to the method of performing a direct transfer to a financial account registered by the user.

[0331] In addition, through Double Spending Prevention algorithms and Atomic Transaction processing, duplicate payments caused by system errors can be prevented and data consistency can be guaranteed.

[0333] **(c-4) Step: Completion Notifications and History Management

[0334] Step (c-4) is a process of transmitting a notification signal indicating the completion of payment of consideration, including economic or non-economic value, to the user terminal (10) and storing the transaction record. In this step (c-4), the user is immediately notified that the payment of consideration has been successfully completed, and detailed information including the payment amount, payment date and time, and transaction number can be provided.

[0335] Furthermore, in step (c-4), payment completion notifications can be transmitted via various channels, including SMS, email, web push notifications, and in-app messages. Additionally, all transaction records can be permanently stored in an encrypted database compliant with GDPR and personal data protection laws, allowing them to be managed for future retrieval, analysis, and dispute resolution.

[0337] **Additional reward processing algorithm

[0338] FIG. 6 is a diagram showing each process of an additional compensation processing algorithm performed by a payment unit (140) in a system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention.

[0339] As illustrated in FIG. 6, in this embodiment, the payment unit (140) can perform an additional reward processing algorithm including a step (d-1) of receiving evidence materials proving subsequent purchase activity from a user terminal (10), a step (d-2) of verifying the purchase activity by analyzing the evidence materials received in step (d-1), a step (d-3) of determining a differentiated additional reward ratio according to the verification result of step (d-2), and a step (d-4) of paying additional rewards to the user account according to the ratio determined in step (d-3).

[0340] The reason for the difference in terminology between 'consideration' and 'compensation' regarding consideration including economic or non-economic value and additional compensation is as follows.

[0341] This is because consideration including economic or non-economic value arises from a data user purchasing a user's search data—essentially, the transaction of the user's search data—and is therefore a consideration rather than compensation; additional compensation, on the other hand, is an additional reward for the user uploading supporting documents.

[0343] **(d-1) Step: Receiving various supporting documents

[0344] Step (d-1) is the process of receiving evidence from the user terminal (10) that proves subsequent purchasing activity. In this step (d-1), various forms of evidence that can prove actual purchasing activity related to the search data provided by the user may be accepted.

[0345] In addition, the supporting documents that can be received in step (d-1) may include, for example, images of purchase receipts, screenshots of online transaction screens, delivery completion notifications, credit card payment records, bank transfer statements, etc. Furthermore, step (d-1) supports a Multipart File Upload function to allow the user to submit various types of supporting documents simultaneously, and can collect metadata (file size, upload time, file format, etc.) for each supporting document.

[0347] **(d-2) Step: AI-based evidence analysis

[0348] Step (d-2) is a process of verifying the purchase activity by analyzing the supporting documents received in Step (d-1). Specifically, in Step (d-2), structured information can be extracted from the supporting documents by utilizing Optical Character Recognition (OCR) technology and computer vision algorithms.

[0349] Accordingly, for receipt images, store information, transaction date and time, transaction amount, and purchased product name can be extracted through the Tesseract OCR engine and regular expression parsing; for online transaction screens, shopping mall information, order number, and purchased product information can be extracted through web scraping pattern recognition; for delivery completion notifications, shipping company information, tracking number, and delivery completion date and time can be extracted using natural language processing (NLP) technology; and for payment history, transaction date and time, merchant information, and transaction amount can be extracted through financial data standard format parsing.

[0350] In addition, step (d-2) can cross-verify the authenticity of the extracted information through external API integration, and evaluate the correlation between the search data provided by the user and the actual purchased product through semantic similarity analysis.

[0352] **(d-3) Step: Determining Differentiated Reward Rates

[0353] Step (d-3) is the process of determining differentiated additional compensation ratios based on the verification results of Step (d-2). In this Step (d-3), different compensation ratios may be applied depending on the completeness and reliability of the verification.

[0354] For example, a method may be applied in which 100% of the basic fee is paid if cross-verification through multiple supporting documents is successful, and 70% to 90% of the basic fee is paid differentially depending on the type of supporting document if verification is performed using only a single supporting document. Additionally, the reliability of the verification can be scored using a machine learning-based reliability evaluation model, and the fee ratio can be dynamically adjusted accordingly.

[0356] **(d-4) Step: Additional rewards paid

[0357] Step (d-4) is the process of paying additional rewards to the user account based on the ratio determined in Step (d-3). In this Step (d-4), the calculated additional reward amount can be actually transferred to or credited to the user account.

[0358] In addition, in this embodiment, step (d-4) can support multiple payment methods including consideration with economic or non-economic value, and can provide consideration in various forms such as cash payment, point accumulation, or coupon issuance. Furthermore, step (d-4) can transmit detailed payment details to the user after the additional compensation payment is completed, and transparently disclose the verification process and the basis for calculating the consideration.

[0360] **Ecosystem optimization of the Quality Control Department (150)

[0361] The present invention may further include a quality management unit (150) that manages a search data quality index for each user in real time and dynamically adjusts the search data submission limit according to the quality index.

[0363] Dynamic limit control algorithm

[0364] FIG. 7 is a diagram showing each process of a dynamic limit control algorithm performed by a quality management unit (150) in a system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to one embodiment of the present invention.

[0365] As illustrated in FIG. 7, in this embodiment, the quality control unit (150) can perform a dynamic limit control algorithm comprising: a step (e-1) of querying the current quality index for the search data of a selected user; a step (e-2) of determining which grade the user belongs to among a plurality of grades classified according to a pre-set classification criterion based on the quality index queried in step (e-1); a step (e-3) of calculating a daily data submission limit based on the grade determined in step (e-2); and a step (e-4) of notifying the user terminal (10) of the submission limit calculated in step (e-3).

[0367] ** (e-1) Step: View Real-time Quality Index

[0368] Step (e-1) is the process of querying the current quality score for the selected user's search data. In this step (e-1), when a specific user requests data submission, the user's latest quality score information can be retrieved in real-time from the Redis in-memory cache.

[0369] In addition, in this embodiment, step (e-1) can retrieve a quality index calculated using a Weighted Moving Average method, which includes comprehensive information such as accumulated past transaction history, data quality evaluation results, success rate of subsequent activity verification, and community ratings for each user.

[0371] **(e-2) Step: Intelligent Classification

[0372] Step (e-2) is a process of determining which grade a user belongs to among multiple grades classified according to pre-set classification criteria, based on the quality index retrieved in Step (e-1). In this Step (e-2), the quality index can be divided into intervals to classify users into multiple grades (e.g., Bronze, Silver, Gold, Platinum, Diamond).

[0373] In addition, in this embodiment, step (e-2) may consider additional factors such as the user's subscription period, activity frequency, compliance history, and community contribution through **Multivariate Analysis**, in addition to the quality index.

[0375] **(e-3) Step: Dynamic Limit Calculation

[0376] Step (e-3) is the process of calculating the daily data submission limit based on the grade determined in Step (e-2). In other words, in Step (e-3), the specific number of submissions allowed for the user can be calculated based on the submission limit criteria pre-set for each grade.

[0377] To this end, step (e-3) can determine the final submission limit by applying a tier-specific multiplier (e.g., Bronze 1.0x, Silver 1.5x, Gold 2.0x) to the basic submission limit, and can also perform dynamic adjustments to reduce the limit during peak hours and increase it during non-peak hours to distribute traffic by time of day.

[0379] **(e-4) Step: User-friendly Guide

[0380] Step (e-4) is a process of providing the submission limit calculated in Step (e-3) to the user terminal (10). In this Step (e-4), the calculated submission limit information can be processed into a form that the user can easily understand and transmitted to the user terminal (10).

[0381] In the case of this embodiment, step (e-4) may provide detailed information including the current rating, quality index, daily submission limit, and remaining number of submissions in the form of a dashboard, and may also provide specific guidelines for quality improvement. Additionally, information updated in real time may be provided to the user by utilizing Progressive Web App technology.

[0383] **Key technical implementation of the quality control department (150)

[0384] 1. Technical Overview

[0385] The quality management department (150) manages the search data quality index for each user in real time and performs the role of dynamically adjusting the search data submission limit according to the quality index. To this end, the quality management department (150) uses a mathematically clearly defined formula for calculating a comprehensive quality index to evaluate the user's activities from various angles.

[0386] 2. Formula for Calculating the Overall Quality Index

[0387] 2.1 Main Formula

[0388] The user quality index Q(t) at time t is defined by the following **weighted linear combination**:

[0389]

[0390] 2.2 Definition of Weights

[0391] Weighting coefficients: α = 0.4, β = 0.3, γ = 0.2, δ = 0.1

[0392] Constraint: α + β + γ + δ = 1.0

[0393] Score range: 0 ≤ Q(t) ≤ 100

[0394] 2.3 Detailed Formulas by Component

[0395] 2.3.1 V(t) (Value Contribution Score)

[0396] It represents the average economic value of the data submitted by users and is calculated using the following formula:

[0397]

[0398] Variable Definition:

[0399] * n: Number of search data entries submitted during the evaluation period

[0400] * Winning Bid: Winning bid (KRW) of the i-th search data

[0401] * Conversion Rate: Purchase conversion rate of the i-th search data (0 ≤ Conversion Rate ≤ 1)

[0402] Constraint: 0 ≤ V(t) ≤ 100

[0403] Specific calculation example: Data for 5 search terms submitted over the past 30 days

[0404] * Data 1: Winning bid 500 won, conversion rate 0.7 → 500 × 0.7 = 350

[0405] * Data 2: Winning bid 800 won, conversion rate 0.85 → 800 × 0.85 = 680

[0406] * Data 3: Winning bid 1,200 won, conversion rate 0.92 → 1,200 × 0.92 = 1,104

[0407] * Data 4: Winning bid 600 won, conversion rate 0.6 → 600 × 0.6 = 360

[0408] * Data 5: Winning bid 900 won, conversion rate 0.8 → 900 × 0.8 = 720

[0409] V(t) = (350 + 680 + 1104 + 360 + 720) / 5 × 0.1 = 3214 / 5 × 0.1 = 64.28 points

[0410] 2.3.2 C(t) (Purchase Conversion Confidence)

[0411] This represents the percentage of submitted data that led to actual purchases, calculated as follows:

[0412] C(t) = (Actual_Purchase_Count / Total_Search_Data_Submission_Count) × 100

[0413] Variable Definition:

[0414] * Actual_Purchase_Count: Number of search data entries that led to actual purchases during the evaluation period

[0415] * Total_Search_Data_Submissions: Total number of search data submitted during the evaluation period

[0416] Constraint: 0 ≤ C(t) ≤ 100

[0417] Calculation example:

[0418] * Search data submitted in the last 30 days: 25

[0419] * Number of cases resulting in actual purchases: 18

[0420] * C(t) = (18 / 25) × 100 = 72.0 points

[0421] 2.3.3 R(t) (Recency Weight)

[0422] It is calculated using the exponential decay function:

[0423]

[0424] Weight function:

[0425]

[0426] Variable Definition:

[0427] * m: Number of transactions subject to evaluation

[0428] * w: Time weight of the i-th transaction

[0429] * q: Quality score of the i-th transaction (0 ≤ q ≤ 100)

[0430] * d: Number of days elapsed since the i-th transaction

[0431] * λ = 0.1 (damping constant)

[0432] Constraint: 0 ≤ R(t) ≤ 100

[0433] Calculation example:

[0434] * Transaction 7 days ago (Quality Score 85): w₁ = e^(-0.1×7) = e^(-0.7) = 0.497

[0435] * Transaction 3 days ago (Quality Score 92): w₂ = e^(-0.1×3) = e^(-0.3) = 0.741

[0436] * Transaction 1 day ago (Quality Score 88): w₃ = e^(-0.1×1) = e^(-0.1) = 0.905

[0437] R(t) = (0.497Х85 + 0.741Х92 + 0.905Х88) / (0.497 + 0.741 + 0.905)

[0438] = (42.245 + 68.172 + 79.64) / 2.143

[0439] = 190.057 / 2.143 = 88.69 points

[0440] 2.3.4 F(t) (Transaction Frequency Normalized Score)

[0441] F(t) = min(Total_Transactions / Reference_Transactions, 1) × 100

[0442] Variable Definition:

[0443] * Total_Transactions: Total number of transactions within the evaluation period (monthly)

[0444] * Reference_Transaction_Count = 20 (Monthly Reference Value)

[0445] Constraint: 0 ≤ F(t) ≤ 100

[0446] Calculation example:

[0447] * In the case where the monthly number of transactions is 25: F(t) = min(25 / 20, 1) × 100 = min(1.25, 1) × 100 = 100 points

[0448] * When the monthly transaction count is 15: F(t) = min(15 / 20, 1) × 100 = min(0.75, 1) × 100 = 75 points

[0449] 3. Dynamic Limit Calculation Algorithm

[0450] 3.1 Step 1: Classification Algorithm

[0451] Based on the calculated quality index Q(t), the following conditional classification is performed:

[0452] IF Q(t) ≥ 90: Grade = "Diamond", Multiplier = 3.0

[0453] ELIF Q(t) ≥ 80: Grade = "Platinum", Multiplier = 2.5

[0454] ELIF Q(t) ≥ 70: Rank = "Gold", Multiplier = 2.0

[0455] ELIF Q(t) ≥ 60: Grade = "Silver", Multiplier = 1.5

[0456] ELIF Q(t) ≥ 50: Rank = "Bronze", Multiplier = 1.0

[0457] ELSE: Rank = "Newbie", Multiplier = 0.5

[0458] 3.2 Step 2: Limit Calculation Formula

[0459] Daily_Limit = Base_Limit × Tier_Multiplier × Time_Time_Adjustment_Factor

[0460] Variable Definition:

[0461] * Basic limit = 10

[0462] * Grade_Multiplier: Value determined in Step 1 (0.5 ≤ Grade_Multiplier ≤ 3.0)

[0463] * Time zone adjustment factor:

[0464] o Peak hours (09:00-18:00): 0.8

[0465] o Non-peak hours (18:01-08:59): 1.2

[0466] Constraint: Daily_Limit ≥ 1 (Minimum Guarantee)

[0467] Specific calculation example: For a Gold rank user (Q(t) = 75 points, multiplier = 2.0):

[0468] * Access at 2 PM: Daily_Limit = 10 × 2.0 × 0.8 = 16

[0469] * 10 PM login: Daily_limit = 10 × 2.0 × 1.2 = 24

[0470] 4. Example of Calculating Overall Quality Index

[0471] 4.1 User A's Overall Evaluation

[0472] Individual scores:

[0473] * V(t) = 64.28 points (Value contribution)

[0474] * C(t) = 72.0 points (Purchase conversion confidence)

[0475] * R(t) = 88.69 points (Recency weight)

[0476] * F(t) = 100 points (transaction frequency, 25 cases per month)

[0477] Final calculation:

[0478] Q(t) = 0.4 Х 64.28 + 0.3 Х 72.0 + 0.2 Х 88.69 + 0.1 Х 100

[0479] = 25.712 + 21.6 + 17.738 + 10

[0480] = 75.05 points

[0481] Rank Determination: Gold Rank (70 ≤ Q(t) < 80)

[0482] Limit calculation:

[0483] * Peak time: Daily limit = 10 × 2.0 × 0.8 = 16 items

[0484] * Off-peak hours: Daily limit = 10 × 2.0 × 1.2 = 24 items

[0485] 5. Technical Basis for Weight Setting

[0486] 5.1 α = 0.4 (Value contribution top priority)

[0487] Basis for setting:

[0488] Key Performance Indicators: Creation of direct economic value by the platform

[0489] * Revenue Generation Ability: The Importance of High Bids and Conversion Rates

[0490] * Sustainability: Encouraging long-term participation from users with high economic contributions

[0491] 5.2 β = 0.3 (Emphasis on reliability)

[0492] Basis for setting:

[0493] The Core of Data Quality: Verifying Authenticity Through Actual Purchase Conversion Rates

[0494] * Fraud Prevention: Prevention of submitting false data

[0495] Advertiser Satisfaction: A key metric directly linked to improved ROI

[0496] 5.3 γ = 0.2 (reflecting recency of fit)

[0497] Basis for setting:

[0498] * Time Sensitivity: Predictive superiority of recent behavioral patterns

[0499] * User Activity: Reflects current reliability

[0500] * Gradual Improvement: Inducing quality improvement at a moderate level

[0501] 5.4 δ = 0.1 (Participation-assisted assessment)

[0502] Basis for setting:

[0503] * Measurement of Participation: Minimum recognition of contribution to ecosystem revitalization

[0504] * Auxiliary Indicators: A supporting role under the quality-first principle

[0505] * Maintaining Balance: Maintaining a policy of prioritizing quality over quantity

[0506] 6. Technical Features and Effects

[0507] The quality control unit (150) of the present invention has the following technical features:

[0508] 1. Quantitative Evaluation: All quality elements are objectified using mathematical formulas

[0509] 2. Dynamic Adjustment: Automatic adjustment of limits based on real-time quality changes

[0510] 3. Multidimensional Analysis: Comprehensive Evaluation of Economic Value, Reliability, Recency, and Participation

[0511] 4. Guarantee of Fairness: Clear Standards and Transparent Calculation Process

[0512] 5. Scalability: Flexible expansion through weight readjustment when adding new evaluation factors

[0513] This quantitative quality management system is a core technical feature of the present invention that is clearly distinct from existing subjective evaluation methods, and enables the objective and fair evaluation of data quality for each user.

[0515] **Gamification System (Optional Implementation)

[0516] The present invention may further include gamification functions to induce continuous user participation and quality improvement. For such gamification functions, the present invention may further include a leaderboard management module, a badge issuance engine, a reward multiplier calculator, an event management module, and a user achievement tracker.

[0518] **Leaderboard Management Module

[0519] The leaderboard management module can calculate rankings by aggregating weekly quality indices for each user and provide additional bonuses to top users. The leaderboard management module can generate ranking information by aggregating user-specific quality index data calculated by the quality management department (150) through an Apache Spark cluster over a set period, and can independently operate weekly, monthly, and quarterly leaderboards.

[0521] **Badge Issuance Engine

[0522] The badge issuance engine can enhance users' sense of accomplishment by issuing badges for achieving continuous high-quality data delivery. Based on an event streaming architecture, the badge issuance engine monitors users' real-time activities and can immediately issue digital badges upon meeting specific conditions. For example, regarding continuous high-quality data delivery, different grades of badges can be awarded for achieving 7, 30, 90, and 365 consecutive days.

[0524] **Additional Reward Multiplier Calculator

[0525] The compensation multiplier calculator can provide incentives to users by applying a compensation multiplier based on improvements in the quality index. The compensation multiplier calculator can apply an additional multiplier to the base compensation amount by analyzing the user's current quality index and past quality index change trends using **Time Series Analysis**. For example, a 1.1x multiplier can be applied if the quality index increases by 10% or more compared to the previous month, and a 1.2x multiplier can be applied if it increases for three consecutive months.

[0527] **Data User Matching Optimization System (Advanced Features)

[0528] The present invention may further include a data user matching optimization function that is linked with a transaction brokerage unit (130) to learn preferences for each data user and perform optimized matching. To this end, the present invention may further include a preference learning engine, a matching score calculation engine, a priority transaction management module, and a feedback collection module.

[0530] **Preference learning engine

[0531] The preference learning engine can learn and analyze the preferred data quality standards for each data user server (20). The preference learning engine can comprehensively analyze search term patterns, transaction participation frequency, and satisfaction evaluation after winning the bid, which each data user has valued highly in the past, by utilizing a deep learning-based recommendation system algorithm. In particular, it can learn the pattern of preference change over time through an RNN (Recurrent Neural Network)-based LSTM (Long Short-Term Memory) model.

[0533] **Matching score calculation engine

[0534] The matching score calculation engine can calculate matching scores in real time by analyzing user search data according to the preferences of each data user. The matching score calculation engine utilizes a **Vector Space Model** to convert the content characteristics of search terms into high-dimensional vectors and can quantify the level of interest and consistency for each data user within a range from 0 to 100 points through Cosine Similarity measurement.

[0536] User group-based collective negotiation system (extension feature)

[0537] The present invention may further include a user group-based collective negotiation function that strengthens collective bargaining power by grouping users with similar interests. For such a collective negotiation function, the present invention may further include a group formation engine, a bargaining power analysis module, a collective transaction brokerage unit, and a consideration distribution processing unit.

[0539] **Group Formation Engine

[0540] The group formation engine can automatically group users with similar tendencies by analyzing the characteristics and interests of users' search data. In this process, the group formation engine combines the K-means clustering algorithm with hierarchical clustering techniques to comprehensively analyze users' past search history, major interest categories, and search time zone patterns, thereby deriving the optimal group composition.

[0542] **Negotiation Power Analysis Module

[0543] The bargaining power analysis module can quantitatively analyze and evaluate the collective bargaining power of a formed user group. Through **Multiple Regression Analysis**, this module can calculate a bargaining power index for each group by synthesizing the number of users within the group, the group's overall average quality index, and the past transaction success rates of group members.

[0545] **User Interface Implementation Example

[0546] FIG. 8 is a diagram illustrating an exemplary search interface provided to a user terminal (10) by a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention.

[0547] As illustrated in FIG. 8, the search interface displayed on the user terminal (10) in this embodiment can be optimized for various screen sizes by applying **Responsive Web Design**. At the top, a logout button and a corporate dashboard view button are placed along with the platform name, and in the central area, a "search data input" section and a search term input field are located.

[0549] **Real-time Value Display

[0550] FIGS. 9 and 10 are diagrams illustrating, in an exemplary manner, commercial value scores according to differences in search terms in a search interface provided to a user terminal (10) by a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to an embodiment of the present invention.

[0551] As can be seen in FIGS. 9 and 10, the value analysis unit (120) can calculate and display different commercial value scores in real time depending on the characteristics of the input search terms. This is an important function that allows the user to immediately recognize the economic value of their search data.

[0553] **Indication of payment process

[0554] FIG. 11 is a diagram exemplarily illustrating the process of payment including economic or non-economic value in a search interface provided to a user terminal (10) by a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to one embodiment of the present invention.

[0555] As can be seen in Fig. 11, a message such as “Congratulations! 268 won has been paid!” is displayed according to the economic or non-economic value-inclusive payment processing algorithm of the payment unit (140), allowing the user to immediately confirm the completion of the transaction.

[0557] **Purchase Proof Upload Interface

[0558] FIG. 12 is a diagram exemplarily illustrating the process of providing additional rewards in a search interface provided to a user terminal (10) by a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to one embodiment of the present invention.

[0559] As can be seen in Fig. 12, through the “Upload Purchase Proof” section, the user can submit proof materials proving subsequent purchase activities, and through the guidance “Immediately verified by OCR automatic analysis,” the automatic analysis function of the payment unit (140) can be known.

[0561] System Scalability and Performance Optimization

[0562] The system of the present invention is designed based on a microservices architecture, allowing each component to be scaled independently. Automatic scaling in response to increased traffic is possible through Docker containers and Kubernetes orchestration, and optimized services can be provided to users worldwide by utilizing a Content Delivery Network (CDN).

[0563] Furthermore, real-time data processing performance is guaranteed through an event streaming architecture utilizing Apache Kafka, and response speeds can be improved through distributed caching using Redis Cluster. The database is configured as a PostgreSQL cluster with sharding applied, enabling the stable processing of large-scale transaction data.

[0565] **Security and Privacy Protection

[0566] This invention applies end-to-end encryption to communication between all components by utilizing a Zero Trust security model. User personal information is processed using homomorphic encryption technology to enable computation even while encrypted, and fully complies with international personal information protection regulations such as GDPR, CCPA, and personal information protection laws.

[0567] In addition, it ensures the transparency and integrity of all transaction processes through a blockchain-based audit log system and performs 24-hour real-time security monitoring by operating an AI-based abnormal transaction detection system.

[0569] conclusion

[0570] For the above, a system for providing a search data trading platform through user-centric real-time data value assessment and transaction brokerage according to an embodiment of the present invention has been described in detail. By having each component operate in an organically interconnected manner, the present invention can realize an innovative data economy ecosystem that provides optimal value to users as data producers, companies as data consumers, and platform operators alike.

[0571] The technical concept and scope of the present invention are not limited to the above embodiments, and those skilled in the art will understand that various modifications and improvements are possible within the scope of the technical concept and scope of the present invention without departing from it.

[0573] Below, various functions additionally applicable to the present invention will be described.

[0574] First, the system for providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage according to the present invention may further include a gamification function linked with a quality management unit (150) that manages the search data quality index for each user in real time to induce continuous user participation and quality improvement.

[0575] For such gamification functions, the present invention may further include a leaderboard management module, a badge issuance engine, a reward multiplier calculator, an event management module, and a user achievement tracker.

[0576] The leaderboard management module can calculate rankings by aggregating weekly quality indices for each user and provide additional rewards to top users. The leaderboard management module can generate ranking information by aggregating user-specific quality index data calculated by the quality management department (150) over a set period, and can independently operate weekly, monthly, and quarterly leaderboards. In addition, the leaderboard management module can provide not only overall user rankings but also segmented leaderboards by age group, region, and interest, thereby encouraging competitive participation from diverse user groups.

[0577] The badge issuance engine can enhance user satisfaction by issuing badges for achieving continuous high-quality data delivery. The badge issuance engine can issue digital badges based on the number of consecutive days of high-quality data delivery, the monthly quality improvement rate, and expertise by specific categories.

[0578] For example, regarding the continuous provision of high-quality data, different grades of badges may be awarded upon short-term, medium-term, long-term, and ultra-long-term continuous achievement, respectively, and an improvement achievement badge may be issued if the monthly quality index improvement rate exceeds a predetermined standard. Additionally, the badge issuance engine may issue a field expert badge to users who consistently provide high-quality search data in a specific product category.

[0579] The compensation multiplier calculator can provide incentives to users by applying compensation multipliers based on improvements in the quality index. The compensation multiplier calculator can apply additional multipliers to the base compensation amount by analyzing the user's current quality index and past trends in quality index changes.

[0580] For example, if the quality index has increased compared to the previous month, a predetermined multiplier may be applied, and if it has increased for a certain period of time, a higher multiplier may be applied, and for users holding a specific badge, an additional multiplier may be provided according to the badge grade. Accordingly, the compensation multiplier calculator is linked with the compensation payment unit (140) so that the calculated multiplier can be automatically applied when paying compensation including economic or non-economic value and additional compensation.

[0581] The event management module can plan and operate special bonus events for premium users. This module can host additional reward events for a certain percentage of users in the top monthly quality index, or run limited-time events tailored to specific anniversaries or seasons.

[0582] Specifically, the event management module can set participation conditions, compensation details, and event period, automatically select target users, and send event participation notices to user terminals (10). In addition, the event management module can operate various types of events, including seasonal theme events, new user welcome events, and events to induce the return of dormant users.

[0583] User achievement trackers can continuously track and manage the achievement of individual users. Such trackers can comprehensively record and analyze the cumulative number of transactions, total value of earnings, trends in quality index changes, and the status of acquired badges per user.

[0584] Furthermore, the user achievement tracker can provide guidelines for achieving the next goal based on individual achievement records and generate personalized engagement messages by analyzing users' past activity patterns. Additionally, the user achievement tracker can identify users who have been inactive for a long period and automatically send notifications to encourage their return.

[0585] This gamification function can be linked in real-time with quality index data managed by the quality management department (150) to continuously provide the user with motivation to improve data quality. All ranking information, badge status, and event participation results provided by the gamification function are visually displayed through the user terminal (10) to maximize the user's sense of participation and achievement.

[0586] Next, the system for providing a search data trading platform through user-centered real-time data value assessment and transaction brokerage according to the present invention may further include a data demander matching optimization function that is linked with a transaction brokerage unit (130) to learn preferences for each data demander and perform optimized matching. To this end, the present invention may further include a preference learning engine, a matching score calculation engine, a priority transaction management module, and a feedback collection module.

[0587] The preference learning engine can learn and analyze the preferred data quality standards for each data user server (20). The preference learning engine can comprehensively analyze the preference tendencies of each data user regarding search term patterns, transaction participation frequency, satisfaction evaluation after winning the bid, and subsequent activity verification results that each data user has previously valued highly.

[0588] Furthermore, the preference learning engine can learn transaction activity by time of day, interest in specific categories, the distribution of transaction amounts by quality index, and preferences regarding user demographic characteristics through machine learning algorithms. The data trained by the preference learning engine can be structured and stored as preference profiles for each data user, and can be utilized as foundational data for calculating matching scores for new search data.

[0589] The matching score calculation engine can calculate a matching score in real time by analyzing the user's search data according to the preferences of each data user. The matching score calculation engine can quantify the level of interest and consistency for each data user by synthesizing the content characteristics of the search term, the user's profile information, the commercial value score calculated by the value analysis unit (120), and the quality index managed by the quality management unit (150).

[0590] Matching scores can be calculated in a range from, for example, 0 to 100 points, and a higher score may indicate the likelihood that the data user has a high level of interest in the search data. The matching score calculation engine can utilize various similarity measurement algorithms, including cosine similarity, Euclidean distance, and Pearson correlation coefficient.

[0591] First, the transaction management module can improve overall transaction efficiency by prioritizing the transmission of transaction requests to data demander servers (20) with high matching scores. This priority transaction management module ranks data demanders based on calculated matching scores and can prioritize the transmission of transaction requests to a certain percentage of top data demanders. If there is a lack of transaction proposals in the priority group, the transaction requests can be transmitted by gradually expanding to groups with lower matching scores.

[0592] Furthermore, the transaction management module can determine priority by considering not only the matching score but also the past transaction proposal speed, transaction success rate, and reliability index.

[0593] The feedback collection module can gather evaluations from data users regarding data quality satisfaction, expectation compliance, and achievement of business objectives after the bid is awarded. This feedback information is then transmitted back to the preference learning engine to be utilized in improving the accuracy of the learning model. Furthermore, the feedback collection module can collect both qualitative evaluations and quantitative scores, and can convert text-based feedback into structured data using natural language processing technology.

[0594] This data demander matching optimization function can be linked with the transaction brokerage processing algorithm of the transaction brokerage unit (130) to make the transaction request transmission process more efficient. That is, by prioritizing data demanders with high matching scores to participate in the transaction, the overall transaction proposal rate can be increased, and more competitive and appropriate transaction conditions can be provided to users. Therefore, the data demander matching optimization function can improve the transaction efficiency and user satisfaction of the entire system.

[0595] Next, the system for providing a search data trading platform through user-centric real-time data valuation and transaction brokerage according to the present invention may further include a user group-based collective negotiation function that strengthens collective bargaining power by grouping users with similar interests. To support such collective negotiation functions, the present invention may further include a group formation engine, a bargaining power analysis module, a collective transaction brokerage unit, and a compensation distribution processing unit.

[0596] The group formation engine can automatically group users with similar tendencies by analyzing the characteristics and interests of users' search data. In this process, the group formation engine can derive the optimal group composition by comprehensively analyzing the user's past search history, major interest categories, search time zone patterns, and quality index levels.

[0597] For example, users can be classified by category, such as groups with a high interest in electronics, groups that frequently search for fashion, and groups with a strong interest in food and cooking; each group must consist of a predetermined minimum number of members to secure collective bargaining power. The group formation engine can manage not only static groups but also dynamic groups formed in real-time, and supports the creation of temporary groups tailored to specific events or seasons.

[0598] The bargaining power analysis module can quantitatively analyze and evaluate the collective bargaining power of formed user groups. This module can calculate a bargaining power index for each group by synthesizing the number of users within the group, the group's overall average quality index, the past transaction success rates of group members, and the level of interest from data consumers in the relevant category.

[0599] Groups with higher bargaining power indices have the potential to elicit better transaction terms from data consumers, and this information can be provided to group members in real time. Additionally, the bargaining power analysis module can propose measures to enhance bargaining power through group expansion or quality improvement.

[0600] The collective transaction brokerage can organize large-scale data packages at the group level and conduct collective negotiations with data demanders. Additionally, the collective transaction brokerage can integrate individual search data from group members, process it into a package format, evaluate the total value of the package, and generate transaction requests.

[0601] Furthermore, unlike individual data transactions, group transactions allow for the utilization of additional bargaining chips, such as bulk purchase discounts, long-term contract terms, and exclusive data access rights. To this end, the group transaction brokerage includes a function to select a group leader, designating the user with the highest quality index or the most active participation within the group as the group leader; the group leader holds the authority to approve the final negotiation terms.

[0602] The compensation distribution processing unit can fairly distribute the compensation secured through collective negotiation to group members. In other words, the compensation distribution processing unit can calculate individual compensation ratios by comprehensively considering each member's data contribution, quality index, and activity level within the group.

[0603] The compensation distribution processing unit basically operates on the principle of equal distribution, but may provide additional bonuses to members or group leaders who provide high-quality data. In addition, the compensation distribution processing unit may operate a system that allows group members to raise objections regarding the distribution results by disclosing transparent distribution criteria in advance.

[0604] In addition, the compensation distribution processing unit can provide a group joint accumulation function to support members in accumulating a certain percentage of their individual compensation into a group common fund and utilizing it for activities aimed at improving the overall quality of the group or strengthening bargaining power.

[0605] In addition, the collective bargaining function may include a group-specific quality management system. Such a system monitors the average quality index of the entire group in real time and can provide improvement support at the group level for members with poor quality.

[0606] In addition, the group-based quality management system can be linked with the quality management department (150) to operate a group-based quality improvement program and can also provide a peer learning function in which excellent members mentor other members. When the quality index of the entire group improves above a certain level, a bonus benefit can be provided to the entire group to motivate collective quality improvement.

[0607] This user group-based collective negotiation function can be operated in parallel with the existing individual transaction process of the transaction brokerage unit (130). Users can choose between individual data transactions and group-based collective transactions, and depending on the situation, they can maximize profits by utilizing both methods. Explanation of the symbols

[0608] 10: User terminal 20: Data Demand Server 100: Platform provider server 110: Data receiver 120: Value Analysis Department 130: Transaction Brokerage Department 140: Payment of Consideration 150: Quality Control Department

Claims

Claim 1 A data receiving unit that receives search data from a user terminal owned by a user; a value analysis unit that calculates a commercial value score of the search data received through the data receiving unit in real time; a transaction brokerage unit that collects transaction proposals including transaction conditions from a plurality of data demander servers based on the commercial value score calculated by the value analysis unit, and provides a list of collected transaction proposals to the user terminal to broker a transaction; and a payment unit that pays a consideration including economic or non-economic value according to the transaction conditions of a specific transaction proposal selected by the user terminal to a user account; wherein the value analysis unit includes: (a-1) a step of executing a machine learning-based value analysis model; and (a-2) a step of calculating individual scores for the search data according to a plurality of preset evaluation criteria. A system for providing a search data trading platform through user-centric real-time data value evaluation and transaction brokerage, comprising: a quality index calculation algorithm including a step (a-3) of determining a commercial value score of the search data by summing individual scores calculated in step (a-2); wherein step (a-2) includes a step (a-2-1) of assigning a first standard score when the search data contains a general product name of a product or service; a step (a-2-2) of additionally assigning a second standard score when the search data contains specific identification information including a brand name, model name, color, and size for the general product name; and a step (a-2-3) of additionally assigning a third standard score when the search data contains keywords indicating purchase intent including purchase, order, price, discount, and delivery. Claim 2 delete Claim 3 delete Claim 4 A system providing a search data trading platform through user-centric real-time data value evaluation and transaction selection management algorithm, wherein the transaction broker performs the following steps: (b-1) generating a transaction request including an anonymized data identifier based on the commercial value score; (b-2) transmitting the transaction request in parallel to the plurality of data demander servers; (b-3) collecting transaction proposals including transaction conditions from the plurality of data demander servers; (b-4) sorting the collected plurality of transaction proposals in descending order according to the transaction conditions; (b-5) transmitting the list of transaction proposals sorted in step (b-4) to the user terminal; (b-6) receiving a selection signal for a specific transaction proposal from the user terminal; and (b-7) generating a unique selection identification number for the selection signal. Claim 5 In claim 1, the consideration payment unit performs an algorithm for processing consideration including economic or non-economic value, comprising: (c-1) a step of calculating the amount of consideration including economic or non-economic value by analyzing the transaction conditions of a specific transaction proposal selected from the user terminal; (c-2) a step of verifying the validity of the user account and whether the consideration including economic or non-economic value can be paid; (c-3) a step of paying the consideration including economic or non-economic value to the user account; and (c-4) a step of transmitting a signal notifying the completion of payment of the consideration including economic or non-economic value to the user terminal and storing a transaction record; a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage. Claim 6 In claim 1, the payment unit receives evidence proving subsequent purchasing activity from the user terminal, verifies the purchasing activity by analyzing the evidence, and additionally pays additional compensation to the user account according to the verification result, a system for providing a search data trading platform through user-centric real-time data valuation and transaction brokerage. Claim 7 In claim 6, the payment unit performs an additional reward processing algorithm comprising: (d-1) receiving evidence materials proving subsequent purchasing activity from the user terminal; (d-2) verifying the purchasing activity by analyzing the evidence materials received in step (d-1); (d-3) determining a differentiated additional reward ratio according to the verification result of step (d-2); and (d-4) paying additional rewards to the user account according to the ratio determined in step (d-3); a system providing a search data trading platform through user-centric real-time data valuation and transaction brokerage. Claim 8 A system for providing a search data trading platform through user-centric real-time data value assessment and transaction brokerage, further comprising a quality management unit that manages a search data quality index for each user in real time and dynamically adjusts the search data submission limit according to the quality index in claim 1. Claim 9 In claim 8, the quality management unit performs a dynamic limit control algorithm comprising: (e-1) a step of querying the current quality index for the search data of a selected user; (e-2) a step of determining which grade among a plurality of grades classified according to a pre-set classification criterion the user belongs to based on the quality index queried in step (e-1); (e-3) a step of calculating a daily data submission limit based on the grade determined in step (e-2); and (e-4) a step of notifying the user terminal of the submission limit calculated in step (e-3); a system providing a search data trading platform through user-centered real-time data value evaluation and transaction brokerage.

Images