Residential remodeling decision support system based on transaction price data and lifestyle trait indicators

Abstract

The present invention relates to a residential remodeling decision support system based on actual transaction price data and lifestyle tendency indicators. The system may include: an input unit that collects basic information, tendency and lifestyle information, and process and / or material selection information of a target residence from a user; and a processing unit that calculates a plurality of lifestyle tendency indicators based on the input information collected from the input unit, calculates the process configuration and / or budget allocation of at least one remodeling option according to the plurality of lifestyle tendency indicators, and calculates an estimate for said remodeling option.

Description

Technology Field

[0001] The present invention relates to a computer-implemented decision support system and method that supports the selection of processes, materials, and budget by utilizing user lifestyle tendency indicators, basic information about the residence (square footage, age, region, etc.), and actual transaction price data together during the process of planning remodeling or partial construction of a residence (apartment, house, etc.).

[0002] More specifically, this relates to a technology that scores lifestyle tendencies to reflect spatial importance (or priority by space) and process configuration, calculates an estimated High Value Increasingness (HVI) and Return on Investment (ROI) based on data such as actual transaction prices from the Ministry of Land, Infrastructure and Transport, evaluates asset value, maintenance, and defect risks, and provides the basis for such judgments to the user in an understandable format. Background Technology

[0003] While interior remodeling may appear to be a "design choice" that changes the atmosphere of a home, in reality, it is closer to a "complex decision-making" process where the selection of materials and processes (wallpapering, flooring, electrical / lighting, bathrooms, kitchens, windows, carpentry, etc.) are intertwined. Fixing one process often causes the scope of other processes to expand or contract, frequently affecting the construction period and costs as well. Even for apartments of the same size (32 pyeong), the difficulty and cost vary significantly depending on the building's age, the condition of existing finishes, the deterioration of plumbing and electrical systems, insulation levels, structural constraints, and management office regulations; consequently, it is difficult for consumers to make decisions suitable for their homes based solely on average unit prices or package descriptions.

[0004] Traditionally, estimates were mostly generated through offline consultations and site visits. While this method is beneficial if the person in charge is highly experienced, the quality of the consultation varies by individual, and if the customer's requirements are not clearly defined, the cycle of "question-answer-re-estimate" is likely to be repeated multiple times. In particular, adding a single process item or changing the material grade by just one level often requires re-explaining the scope of inclusions and exclusions, indirect costs, construction duration, and work sequence. During this process, customers fail to systematically understand "why the price has changed," ultimately leading them to make judgments based solely on the total amount or delay their decision.

[0005] To alleviate such inconvenience, online quote comparison platforms and simplified quoting services have increased, but they have another limitation. Even if quotes are received from multiple companies at once, the itemization and calculation standards vary by firm, meaning the actual scope of work can differ even if the projects appear to be the same on the surface. For instance, items such as demolition, waste disposal, protection, site cleanup, and supervision fees may be included in some places but listed separately in others. In such cases, consumers find it difficult to "compare under the same conditions," which ultimately leads to problems where they are swayed by the "lower total quote" or, conversely, experience growing distrust.

[0006] While simplified quotes quickly provide prices based on limited input values ​​such as square footage, age, and scope of work, they struggle to reflect the diverse living conditions of individuals. For instance, single-person households may prioritize convenience and minimize storage space, whereas families with young children may place greater importance on safety, durability, ventilation, and insulation. Dual-income households may prefer options that reduce the burden of maintenance, while noise, lighting, and workflow become critical variables for remote workers. However, conventional services fail to consistently incorporate these "differences in lifestyle patterns" into process priorities or budget allocation, often leaving users with the feeling of choosing an "average package" rather than a "decision tailored to their needs."

[0007] Recently, tools such as photo-based style recommendations, mood boards, 3D drawings, and AI interior image generation are also widely used. While these tools are useful for quickly establishing a mood or taste, they can cause confusion in decision-making if the actual process selection and cost structure are not firmly integrated. In particular, if a user wants partial renovations but is presented with images that look like a full renovation, or if the selected processes do not correspond one-to-one with the visual results, misguided expectations may be formed. Ultimately, trust issues can arise during the consultation phase as explanations such as "more processes are needed or the budget is higher to achieve the look of the image" are repeated.

[0008] Furthermore, interior design is not merely about "making things look pretty," but is a decision that influences residential satisfaction, maintenance costs, and even asset value. Many users actually ask questions such as, "How much will the value of the house increase if I do this renovation?" or "Is there value left over compared to the construction costs?" However, conventional interior design services often struggled to provide quantitative evidence for these questions or frequently relied on the experiential advice of the person in charge. If there is a lack of comparison methods that consider actual transaction data or factors such as region, age, and market conditions, users are more likely to over-expect or under-judge from an investment perspective.

[0009] Choosing to remodel also entails 'risks.' Some materials are difficult to manage, which can lead to increasing maintenance costs over time, while certain processes increase the likelihood of defects (leaks, mold, cracks, etc.) if they conflict with the existing condition of the home (e.g., old plumbing, aging electrical systems, waterproofing). Furthermore, selecting specific finishes or non-standard options may deviate from market preferences, potentially acting as a depreciation factor when selling the property. However, traditionally, there has often been a lack of systems to systematically warn about these risks by categorizing them into "asset value, maintenance, and defects," or to provide guidance on risk levels in stages based on user conditions.

[0010] Particularly in existing residential buildings, invisible elements within the walls, such as plumbing, electrical wiring, and rebar, significantly impact the safety and quality of construction. While verification is typically performed using expert equipment, blueprints, and on-site experience, proceeding with drilling, demolition, or installation processes without sufficient verification can lead to rework, project delays, or safety issues. Even if some services offer pre-construction inspections or blueprint verification, it is difficult to effectively incorporate these results into decision-making unless they are seamlessly integrated into a continuous flow of process selection, cost estimation, and risk information.

[0011] Ultimately, the core limitation of conventional technology is that the elements users actually need—lifestyle patterns and preferences, spatial inconveniences, process and material selection, quotation item structure, visual aids, asset value assessment, maintenance and defect risks, and explanations of "why such a recommendation was made"—have not been connected into a single system and consistently updated. Consequently, even slight changes in user choices have resulted in inconsistent overall outcomes, the reasons for recommendations have been unclear and difficult to accept, and inefficiencies have arisen due to repeated consultations and re-quoting. Against this backdrop, there has been a continuous demand for decision support technology that organizes and presents costs, value, risks, and rationale together, tailored to individual conditions and changes in choices. Prior art literature

[0012] Korean Published Patent Application No. 10-2023-0035750 (March 14, 2023) The problem to be solved

[0013] The objective of the present invention is to receive input regarding a user's lifestyle inclination indicators and residential conditions, and to provide the results necessary for remodeling decision-making in an integrated flow. Specifically, the invention aims to (i) automatically calculate priorities by space and process based on lifestyle inclination scores and inconvenience factors, (ii) propose a composition of processes, materials, and budget according to the priorities, and (iii) ensure that relevant scores and estimates are automatically updated in a consistent manner when the user selects or changes processes or materials.

[0014] Furthermore, the present invention aims to enable users to make "rational" judgments rather than merely "pretty" ones by presenting expected changes in asset value (estimated increase in housing prices) and return on investment (ROI) after construction based on actual transaction price data, categorized by scenario. In addition, it aims to provide explainable decision-making grounds by evaluating the risks of process and material combinations from the perspectives of asset value (ASSET), maintenance (MAINTENANCE), and defect (DEFECT), presenting them in a PASS / WARN / BLOCK format, and explaining "why this conclusion was reached" for each judgment through the correlation between lifestyle tendency indicators and selection factors. An additional objective is to assist in making safe and convincing remodeling decisions by supporting the identification and avoidance of construction risk zones in advance using smartphone sensors and references to drawings when necessary. means of solving the problem

[0015] The present invention, aimed at solving the above-mentioned problems, relates to a system for calculating interior space priorities and recommending processes, materials, and estimates. The system may include: an input unit that collects basic information, tendencies, and lifestyle information of a target residence, and information on the selection of processes and / or materials from a user; and a processing unit that calculates a plurality of lifestyle tendency indicators based on the input information collected from the input unit, calculates the process configuration and / or budget allocation of at least one remodeling option according to the plurality of lifestyle tendency indicators, and calculates an estimate for the remodeling option.

[0016] According to an embodiment of the present invention, the processing unit may be configured to select some of a plurality of tendency analysis questions based on basic information of the target residence and present them to a user, and to collect the user's response to the selected questions to calculate the plurality of lifestyle tendency indicators.

[0017] According to an embodiment of the present invention, the plurality of lifestyle tendency indicators may be configured to be calculated based on a many-to-many mapping rule in which one question response contributes to a plurality of lifestyle tendency indicators or multiple question responses contribute cumulatively to a single lifestyle tendency indicator.

[0018] According to an embodiment of the present invention, the processing unit may correct the plurality of lifestyle tendency indicators by referring to cross-influence information indicating correlation or cross-influence between lifestyle tendency indicators, so that other lifestyle tendency indicators are corrected in conjunction with a change in a specific lifestyle tendency indicator.

[0019] According to an embodiment of the present invention, the processing unit calculates (i) a survey-based lifestyle tendency indicator vector calculated based on survey responses included in the tendency and lifestyle information, and (ii) an action-based lifestyle tendency indicator vector calculated based on the action history regarding the selection, deselection, grade change, and / or option confirmation of the user's process and / or material selection information, respectively; calculates a final lifestyle tendency indicator vector by weightedly combining the survey-based lifestyle tendency indicator vector and the action-based lifestyle tendency indicator vector, while dynamically adjusting weights according to the degree of accumulation of the action history or the recency of the survey responses; updates the action-based lifestyle tendency indicator vector and the final lifestyle tendency indicator vector in response to changes in the process and / or material selection information; is configured such that the process configuration and / or budget allocation and estimate of the remodeling option are updated in conjunction with the updated final lifestyle tendency indicator vector; and is configured such that additional changes in process and / or material selection information performed by the user based on the updated remodeling option and / or estimate are received again to update the action-based lifestyle tendency indicator vector, thereby [according to] the process and / or material selection It can be configured to implement a bidirectional feedback loop in which propensity recalibration, option reconfiguration, and estimate updating are repeatedly performed.

[0020] According to an embodiment of the present invention, the processing unit may be configured to calculate a space priority score for each of a plurality of spaces including at least some of a kitchen, bathroom, living room, and bedroom by reflecting addition / subtraction factors based on a lifestyle tendency index and tendency / lifestyle information starting from a reference value, and to calculate the process configuration and / or budget allocation according to the space priority score.

[0021] According to an embodiment of the present invention, the processing unit may be configured to apply a classification rule that classifies a process into at least one of mandatory, recommended, or optional based on at least one of the space priority score and the construction year or family composition of the target residence, and the classification result may be reflected in the process configuration of the remodeling option.

[0022] According to an embodiment of the present invention, the processing unit may be configured to generate and provide a plurality of remodeling options in which at least one of different budget levels, material grades, or process ranges differs, and to provide process configurations and estimates for each option together so that they can be compared.

[0023] According to an embodiment of the present invention, the processing unit calculates basic costs based on area and region, applies a distribution ratio for each process, and calculates an estimate by applying a weighting factor for each process based on a lifestyle tendency index, and may be configured to reflect material costs by referring to cost or unit price data corresponding to the selected process and / or material.

[0024] According to an embodiment of the present invention, the processing unit may be configured to calculate a similarity for a plurality of predefined interior style candidates based on the plurality of lifestyle tendency indicators to match at least one style, classify lifestyle routine types based on lifestyle patterns included in the tendency and lifestyle information, and calculate and provide a scenario suitability score for each remodeling option based on the matched style and / or lifestyle routine type.

[0025] According to an embodiment of the present invention, the system estimates the possibility of the existence of a metal structure inside a wall using a magnetic field sensor or a direction sensor of a user terminal, and applies a fallback strategy to estimate the location of the metal structure by referring to a drawing database when the sensor reliability is below a predetermined threshold, and the estimation result may be configured to be reflected in the selection of a process related to drilling or demolition or the configuration of the process.

[0026] According to an embodiment of the present invention, the processing unit refers to process-image mapping information corresponding to process and / or material selection information to generate and provide pre- and post-construction images so that change elements corresponding to the selected process are reflected, wherein the post-construction image may be configured so that change elements corresponding to the unselected process are not reflected.

[0027] According to an embodiment of the present invention, the processing unit may be configured to calculate a market price difference related to a target residence based on actual transaction price data, classify transaction data of similar areas and similar floor plans into at least a first group and a second group according to a predetermined building age standard, calculate the market price difference based on the average transaction price difference of the first group and the second group, and calculate and provide a value index including an expected increase in asset value after construction and an investment recovery rate based on the market price difference and the estimate.

[0028] According to an embodiment of the present invention, the processing unit may be configured to apply at least one of a regional grade or regional weight, a building age coefficient, a process range coefficient, and a style match coefficient when calculating the value indicator, and to calculate and provide a plurality of scenario-specific value indicators based on different coefficient settings.

[0029] According to an embodiment of the present invention, the processing unit evaluates a plurality of risks including at least one of asset value risk, maintenance risk, and defect occurrence risk based on a combination of processes and / or materials and conditions of the target dwelling, assigns a judgment of one of recommendation (PASS), caution (WARN), or non-recommendation (BLOCK) based on the risk evaluation result, and generates and provides the basis for the judgment in natural language including a correlation with lifestyle tendency indicators and user input information, wherein the basis may be configured to provide a basis chain to the user that includes a connection relationship between a question response, a lifestyle tendency indicator, and a judgment result.

[0030] According to an embodiment of the present invention, the processing unit may be configured to identify process or estimation parameters linked by a predefined dependency relationship with the changed process and / or material when a partial change in process and / or material selection information occurs, and to selectively recalculate only for the identified process or estimation parameters to partially update the estimation and / or remodeling options.

[0031] According to an embodiment of the present invention, the processing unit may be configured to generate at least one of (i) a first constraint that forces the inclusion of a specific process, (ii) a second constraint that blocks the selection of a specific process or a combination of specific processes, and (iii) a third constraint that replaces a specific process or material with an alternative process or material, in conjunction with the determination, reconfigure the process configuration of the remodeling option or control selectable items on the user terminal based on the constraint, and update the estimate in conjunction with the reconfigured process configuration. Effects of the invention

[0032] According to the present invention, a user's lifestyle habits, family composition, inconvenience factors, and daily routines are quantified into a lifestyle tendency indicator (score). Since this score is configured to be directly linked to the importance of space (or priority by space) and the allocation of processes, materials, and budget, "what needs to be done first and why" can be summarized at a glance. As a result, even with the same floor area and similar budget, core criteria that differ for each user—such as those prioritizing cleaning / maintenance, storage / flow, or atmosphere / presentation—are reflected in personalized decision-making results rather than an average package, thereby increasing satisfaction.

[0033] Furthermore, since the present invention can apply a bidirectional (feedback) structure in which the user's selection of a process or change of options influences the propensity score to readjust the value, recommendations, priorities, and quotes are updated together based on the same criteria even when a change in selection occurs. Accordingly, users can immediately compare "what inconveniences remain and how costs and priorities change if a particular process is excluded," and service providers or consultants can reduce the repetitive tasks of re-quoting and re-explaining. Moreover, by automatically generating AI-mapped Before / After images for each process, users can intuitively verify "what actually changes with the newly selected process," thereby reducing expectation discrepancies and communication costs.

[0034] In addition, the present invention calculates the Housing Value Increase (HVI) and Return on Investment (ROI) by reflecting the market gap and regional and age-related weighting based on data such as actual transaction prices from the Ministry of Land, Infrastructure and Transport, and provides comparisons through scenarios such as conservative, realistic, and optimistic. This enables users to approach remodeling not merely as "emotional consumption" but as a rational decision that includes an asset perspective. In particular, this helps to quickly determine "whether spending this amount of money is excessive or appropriate" and "which choice is advantageous when planning to sell or rent."

[0035] Furthermore, the present invention can provide Explainable Decision Making (XAI) by evaluating the risks of process and material combinations from the perspectives of asset value (ASSET), maintenance (MAINTENANCE), and defect (DEFECT), presenting them as PASS / WARN / BLOCK judgments, and explaining the reasons for the judgment based on the connection between propensity indicators and material characteristics. Accordingly, users can make selections while understanding "why a warning is triggered" and "what needs to be changed to ensure safety," thereby reducing the likelihood of disputes and increasing the reliability of decision-making. If necessary, the location of metal structures inside walls can be pre-estimated through a three-stage fallback strategy combining smartphone sensors and drawing references, which is expected to induce pre-avoidance and safety measures during hazardous processes such as drilling and demolition. Brief explanation of the drawing

[0036] FIG. 1 is a block diagram schematically showing the overall configuration and processing flow of a residential remodeling decision support system based on actual transaction price data and lifestyle trend indicators according to the present invention. FIG. 2 is a block diagram schematically showing the detailed configuration of the user input collection unit and the lifestyle tendency indicator calculation unit according to the present invention, and the relationship linked thereto to remodeling decision-making / valuation / risk and basis provision. FIG. 3 is a block diagram schematically showing the detailed configuration and interconnected flow of the remodeling decision-making output unit, the actual transaction price-based valuation unit, and the risk and basis provision unit according to the present invention. Specific details for implementing the invention

[0037] Hereinafter, specific details for implementing the present invention will be described with reference to the attached drawings. Furthermore, in describing the present invention, detailed descriptions of related known functions are omitted if they are deemed obvious to a person skilled in the art and could unnecessarily obscure the essence of the invention.

[0038] FIG. 1 is a block diagram schematically showing the overall configuration and processing flow of a residential remodeling decision support system based on actual transaction price data and lifestyle trend indicators according to the present invention.

[0039] Referring to FIG. 1, the residential remodeling decision support system based on actual transaction price data and lifestyle tendency indicators according to the present invention includes a user input collection unit (100), a lifestyle tendency indicator calculation unit (200), a remodeling decision calculation unit (300), an actual transaction price-based valuation unit (400), and a risk and basis provision unit (500).

[0040] At this time, each component of the present invention may be classified into an input unit and / or a processing unit. Specifically, a user input collection unit (100) may be included in an input unit that collects basic information, tendencies and lifestyle information, and process and / or material selection information of the target residence from the user. Additionally, a lifestyle tendency index calculation unit (200) and a remodeling decision calculation unit (300) may be included in a processing unit that calculates a plurality of lifestyle tendency indices based on the input information and calculates the process configuration and / or budget allocation of at least one remodeling option to calculate an estimate. Furthermore, a real transaction price-based valuation unit (400) and a risk and basis provision unit (500) may be additionally included in the processing unit.

[0041] The above input unit may be implemented as an input interface provided on a user terminal (e.g., smartphone, tablet, laptop, desktop computer, etc.) and may include a touchscreen, keypad, mouse, voice input unit, camera, sensor (e.g., location sensor, environment sensor, etc.), and / or API input means linked with an external system. Additionally, the above input unit may be configured to acquire information not only through direct user input but also through the reception of data from an external database, real estate platform, process / material catalog server, etc.

[0042] In addition, the processing unit may be implemented as a computing device comprising one or more processors (e.g., CPU, GPU, NPU, microcontroller, etc.) and memory (e.g., RAM, ROM, storage device), and may be configured to perform functions such as calculating the lifestyle tendency index, calculating remodeling options, calculating estimates, evaluating value, and providing risk and basis by having the processor execute program instructions stored in the memory. For example, the processing unit may be performed within a user terminal, on a separate server (e.g., cloud server, web server, database server, etc.), or configured to be performed in a distributed processing manner by sharing roles between the user terminal and the server.

[0043] FIG. 2 is a block diagram schematically showing the detailed configuration of the user input collection unit and the lifestyle tendency indicator calculation unit according to the present invention, and the relationship linked thereto to remodeling decision-making / valuation / risk and basis provision.

[0044] Referring to FIG. 2, the user input collection unit (100) includes a residential basic information input unit (110), a tendency / lifestyle information input unit (120), and a process selection input unit (130).

[0045] The housing basic information input section (110) performs the function of receiving input from the user for "objective conditions of the target housing" which serve as the basis for calculating the lifestyle tendency indicator, calculating the construction configuration, calculating the estimate, evaluating the value based on actual transaction prices, and determining the risk, and generating and providing housing basic information in a normalized form.

[0046] The residential basic information input section (110) can be configured to prioritize the collection of key items so that the user can perform analysis with minimal input, for example, receiving input of at least some of the following: the size (or exclusive area) of the target residence, regional information, year of construction (or year of completion), family composition (or cohabitation type), and space configuration information (e.g., number of rooms, number of bathrooms). Here, since the size or exclusive area is used as a "scale parameter" for calculating the basic amount of the estimate and the distribution criteria for each process, the residential basic information input section (110) can be configured to store the size selected by the user (e.g., 20 pyeong, 25 pyeong, 32 pyeong, etc.) as numerical data, or to convert the pyeong unit into square meters and store it in a unified unit. In addition, to prepare for cases where the user does not know the exact area, the residential basic information input section (110) can be configured to reduce the input burden by using a representative size selection method (button / list selection) or a range selection method.

[0047] Since regional information is a key item in the valuation based on actual transaction prices and the calculation of the difference in market prices, the residential basic information input section (110) can be configured to receive input of stepwise administrative districts such as city / province, city / county / district, town / village / neighborhood, or receive input of an address string and convert it into a standardized code form internally and store it.

[0048] For example, when a user enters an address or complex name, the residential basic information input unit (110) normalizes the input to obtain a legal district code or a similar regional identifier, and can provide it as a regional parameter required for querying actual transaction price data in a subsequent step. In addition, since regional information can serve as a standard for regional tier classification and regional weighting application, the residential basic information input unit (110) can be configured to map a predefined regional tier based on the entered regional information, or at least transmit the minimum information required for regional tier mapping to the processing unit.

[0049] Since the year of construction is an item that directly affects the "necessity of construction (essential / recommended / optional)," "possibility of defect occurrence," and "value reflection coefficient," the housing basic information input section (110) can be configured to receive input in the form of the year of completion (e.g., 2010) or to receive input in the form of user-friendly age ranges (e.g., new construction 5 years or less, semi-new construction 10 to 20 years, older construction 20 years or more, 30 years or more, etc.).

[0050] At this time, the housing basic information input unit (110) can store the entered year as is, and in addition, reclassify and store it into a "year category" or "building age coefficient application range" so that subsequent algorithms can easily use it. For example, if the user selects "20 years or more," the housing basic information input unit (110) can label this as an aging range and generate flag information indicating that the need for basic construction work such as plumbing / equipment may be relatively high, which can be used as a base value for determining essential or recommended processes in subsequent process classification logic.

[0051] Since family composition can be used as a major variable in calculating life satisfaction and space priority, the housing basic information input section (110) can be configured to receive input from the user in an intuitive living unit, such as single-person household, newlywed / couple, infant children, adolescent children, adult children, living with elderly parents, living with pets, etc.

[0052] Since the input of family composition does not simply mean the number of people but can be used as an indicator representing the "representativeness of lifestyle patterns," the housing basic information input section (110) may generate and store lifestyle type tags (e.g., childcare household, large family, pet household, etc.) together when a family composition item is selected, or provide them so that they can be used to generate hypotheses for question selection during the subsequent tendency / lifestyle information input process. Additionally, since there are cases where two or more characteristics coexist at the same time (e.g., married couple + pet), the housing basic information input section (110) may be configured to allow multiple selections or to receive input by distinguishing between representative items and additional items.

[0053] The housing basic information input section (110) can be configured to improve input convenience by estimating the basic spatial configuration even if the user does not provide detailed drawings, such as "automatic recommendation of expected structure." For example, it can be implemented in a way that automatically sets default values ​​for the number of rooms and bathrooms according to the area and housing type, and allows the user to modify them if the actual structure is different.

[0054] Since this spatial configuration information is not merely for display purposes but serves as basic data for determining "which space to use as a premise to configure process options" in space priority calculation and process classification, the residential basic information input section (110) can receive additional input of minimum structural characteristics, such as the type of entrance, living room, and kitchen (e.g., open type / separated type), in addition to the number of rooms / bathrooms. However, since the present invention can have reducing the input burden as one of its technical effects, the residential basic information input section (110) can be configured to guide input step by step by distinguishing between mandatory input items and optional input items.

[0055] Additionally, the residential basic information input section (110) may include input verification and consistency check functions to increase the reliability of the input information. For example, if the number of rooms / bathrooms relative to the area significantly deviates from the usual range, the user may be asked to reconfirm, or if there is a discrepancy between the selected year range and the year of completion input, the user may be guided to make corrections with priority. Furthermore, the residential basic information input section (110) may be configured to selectively collect additional information effective for drawing identification, such as address, building name, building / unit number, or floor number, in preparation for cases where a drawing database reference is required in subsequent configurations. The identification information collected at this time can be used as a key for architectural drawing lookup in subsequent processing, such as "drawing-based correction or fallback judgment."

[0056] The tendency / lifestyle information input section (120) is configured to collect "subjective and behavioral conditions" such as individual user preferences, lifestyle habits, purpose of use, perceived inconvenience, priority of values, and lifestyle routines as structured data, distinct from the housing basic information input section (110) which collects "objective conditions" of the target housing.

[0057] That is, the tendency / lifestyle information input unit (120) performs the role of securing and providing user response data in the form required by the subsequent lifestyle tendency indicator calculation unit (200) and remodeling decision calculation unit (300) so that the structure in which "lifestyle habits and family types are scored and the scores directly influence actual processes, materials, budgets, and priorities," which is a differentiating feature of the present invention, is established.

[0058] The tendency / lifestyle information input unit (120) may be configured to present the selected questions step by step on the user terminal screen and collect responses to the questions when the processing unit (or lifestyle tendency indicator calculation unit (200)) selects a "set of questions to be presented" based on basic information such as area, year of construction, and family composition received from the housing basic information input unit (110). For example, the tendency / lifestyle information input unit (120) may extract only a portion of a number of question banks (e.g., around 6 or a predetermined range of numbers) to reduce the user's response burden, and each question may be provided as a two-option multiple choice, multiple choice, priority choice, or intensity choice, so that the response can be collected in a form that can be immediately quantified.

[0059] In order for a many-to-many mapping to be possible at the later stage, where "one question and answer contributes to multiple lifestyle tendency indicators or multiple question and answer contributes cumulatively to one lifestyle tendency indicator," it is desirable that the tendency / lifestyle information input unit (120) be configured to store and transmit question identification information (question ID) and answer identification information (response code) together, rather than simple text, and to store metadata such as the time of response and response change history together, thereby serving as a basis for recalibrating the tendency indicator or providing a basis chain.

[0060] Specifically, the personality / lifestyle information input section (120) can directly collect the user's interior standards and preferences by providing a flow of personality questions such as "find your taste first." For example, in the situation where "guests suddenly come," the user can receive a choice such as whether to prioritize hidden storage or the sense of openness of the house, or in the bathroom, whether to prioritize "convenience of cleaning over beauty" or "atmosphere even if there is a burden of maintenance." This allows for the collection of responses that reveal the user's personality, cleaning sensitivity, preference for visibility / openness, sensitivity to trends, preference for durability, preference for aesthetics, etc.

[0061] In addition, by having the user select a single or multiple items of their top priority values ​​(e.g., storage, cost-effectiveness, durability, trendiness, etc.) through questions such as “I absolutely cannot give up on this,” the space priority calculation unit (310) or process classification and option configuration unit (320) can be provided with reference data to determine “what should be reflected first.” In this way, the preference / lifestyle information input unit (120) forms an input structure that is distinguished from simple image / atmosphere recommendation type services in that it does not receive user preferences merely as a simple “style name,” but collects them in a repeatable format based on choices and provides input materials that can be scored.

[0062] In addition, the tendency / lifestyle information input section (120) can be configured to collect "lifestyle information" based on actual lifestyle patterns rather than simple demographics. For example, by collecting inputs on the balance of values, such as whether the user mainly rests at home on weekends, whether they frequently invite friends over for home parties, or which value is more important between "good-looking design" and "practical storage," users with different actual usage scenarios can be distinguished even if they prefer the same style.

[0063] These inputs serve as basic data necessary to classify lifestyle routine types (e.g., morning type, evening type, weekend type, etc.) at the later stage and to connect to scenario cards or the calculation of option suitability scores. That is, the tendency / lifestyle information input section (120) secures not only responses to "what kind of house is pretty" but also responses to "how the house is used," thereby enabling the pipeline of the present invention, which connects "lifestyle pattern-based analysis -> process, materials, budget, and image," to operate.

[0064] Furthermore, the preference / lifestyle information input section (120) allows the user to upload a photo of the interior they prefer ("photo analysis" flow), thereby collecting visual preference clues such as color, material, pattern, and lighting atmosphere that are difficult to capture with text surveys alone.

[0065] At this time, even if the analysis of the photo itself and style matching are performed in the subsequent logic, the tendency / lifestyle information input unit (120) provides an input structure that allows for branching to different results even for users who have entered the same photo, by combining and delivering at least photo data (or reference information therefor) with the response to a value question selected by the user (e.g., design priority / storage priority, home-stay healing / home party, etc.), so that even users who have entered the same photo can have different lifestyle tendencies. In addition, the tendency / lifestyle information input unit (120) may be configured to store which response was given to which question in an immutable log form to replay and provide the user's response in the form of a "reference replay," or at least maintain question-response history information so that the subsequent basis generation unit (530) can form a connection relationship between question-response-indicator-judgment.

[0066] That is, the tendency / lifestyle information input section (120) presents questions suitable for the user based on the housing conditions obtained from the housing basic information input section (110), and collects and records user responses in a standard format that can be quantified, thereby providing an input basis that enables the calculation of indicators (question-indicator mapping calculation section (210)), reflection of cross-indicator influence (indicator cross-influence reflection section (220)), and recalibration of tendency indicators based on fair selection (tendency indicator recalibration section (230)) of the lifestyle tendency indicator calculation section (200) to be performed accurately. At the same time, through linkage with the fair selection input section (130) and the risk / evidence provision section (500), it functions as an input hub that accumulates the user's lifestyle and preference data in stages and provides it in an updateable form so that a bidirectional structure such as "tendency score -> fair selection -> recalibration of tendency value again" can be realized.

[0067] The process selection input section (130) is an input configuration that allows the user to select and confirm "what construction work to do and within what scope," thereby forming the criteria for the process configuration (process combination) and budget allocation that the subsequent remodeling decision-making output section (300) and estimate calculation / update section (330) actually need to calculate. That is, the process selection input section (130) is not a simple checklist, but functions as an input hub that connects the pipeline of the present invention, which leads from "Tendency -> Process -> Estimate / Value / Risk / Image," to the actual decision-making stage by combining with the lifestyle tendency indicator (lifestyle tendency indicator output section (200)) derived from the tendency / lifestyle information input section (120).

[0068] The process selection input unit (130) can first collect input for selecting the remodeling scope from the user terminal screen. For example, by allowing the user to select the scope of construction (a scope input that affects the process scope coefficient), such as whether it is "full remodeling" or "partial construction," it is possible to distinguish whether the same detailed process is viewed as "design / inter-process linkage from the perspective of the overall process" or as "local improvement of only a specific space / process."

[0069] Since this range input is a direct prerequisite for the application of process range coefficients in the subsequent actual transaction price-based valuation unit (400), particularly in the HVI / ROI calculation and scenario comparison unit (430), or for the option configuration method of the remodeling decision calculation unit (300), it is desirable for the process selection input unit (130) to be configured to store the range selected by the user as an identifiable code value (e.g., whole / part, etc.) and transmit it to the subsequent unit.

[0070] Next, the process selection input unit (130) collects selection information for specific process items. For example, it provides multiple process items such as kitchen, bathroom, flooring, tile, wallpapering, built-in closet (including carpentry), sash, interior door, painting, electrical / lighting, film, and demolition, and the user can select the necessary process in a multiple selection manner.

[0071] At this time, the process selection input section (130) can be configured to provide pre-set process bundles (presets) such as "basic remodeling," "kitchen + bathroom focus," and "select all processes" so that the user can make a quick decision, and when a preset is selected, individual processes included in that bundle are automatically selected. Conversely, when the user directly selects individual processes, the list of selected processes, the number of selections, and flags indicating whether each process is selected can be updated and displayed in real time so that the user can clearly recognize "what was selected and what was excluded."

[0072] An important feature of the process selection input section (130) is that it receives additional "detailed requirements for each process" for the selected process, thereby enabling branching of actual material and construction options beyond simply "doing / not doing a kitchen." For example, when a kitchen process is selected, the process selection input section (130) can additionally collect "detailed attribute values ​​of the kitchen process," such as values ​​considered important in the kitchen (cleaning / maintenance, storage / organization, open type for family communication, sensory design, etc.), textures of countertops / doors (matte solid color, wood / stone texture, glossy accent, etc.), and cooking tool storage methods (all hidden / open / prioritizing movement path, etc.).

[0073] Even when a bathroom process is selected, attribute values ​​such as whether the focus is on cleaning and mold management, hotel-like atmosphere, preference for bathtub and spa functions, and color tone (white / beige / marble pattern / deep tone, etc.) can be input. These process-specific attribute values ​​serve as input materials that can be converted into material grades, construction difficulty, and option branches (essential / recommended / optional or E / Standard / Opus, etc.) within the process in the process classification and option configuration section (320) and the estimate calculation / update section (330) at the end. Additionally, the complex risk evaluation section (510) and judgment assignment section (520) of the risk and basis provision section (500) provide the "context of choice" necessary for the maintenance risk (MAINTENANCE) or defect occurrence risk (DEFECT).

[0074] The process selection input section (130) does not simply end at the stage of selecting a process, but may additionally collect "option selection input" to allow the user to make a final selection of one of the multiple remodeling options (options with different budget levels, material grades, and process ranges) generated by the system when presented to the user.

[0075] For example, when different grades of options such as cost-effective, balanced, or premium types are presented for the same process combination, the user can confirm and select a specific option through the process selection input unit (130), and the result of the selection can be reflected in the subsequent detailed quotation configuration, value indicator (HVI / ROI) display, risk assessment (PASS / WARN / BLOCK), and basis generation (basis generation unit (530)). That is, the process selection input unit (130) serves as a step for receiving input on "what the user wants" and simultaneously as a step for confirming "which of the options proposed by the system to adopt," thereby performing the role of converging the decision support results of the present invention into the final decision input.

[0076] Additionally, the process selection input unit (130) may be configured to provide a warning indication or guide additional verification procedures regarding the selection of a high-risk process, such as drilling or demolition, by referring to the "result of estimation of the possibility of existence and location of metal structures (pipes / wires, etc.) within the wall" calculated by the hardware sensor linkage unit. For example, if a user selects a demolition process or a process with a high probability of drilling, and the sensor-based estimation result indicates a risk zone, the process selection input unit (130) may inform the user of the need to avoid the risk zone or perform additional safety measures (reinforcement / relocation, etc.) on the process selection screen, and guide the input flow so that the selection is reflected in the subsequent process configuration. This means that the process selection input unit (130) can be expanded to assist in input from a safety and risk perspective so that the process selection converges into a "method that can be done," rather than merely receiving "the construction that is desired."

[0077] Additionally, the process selection input unit (130) may be configured to transmit to the processing unit as a "process selection event" whenever inputs such as user process selection, process release, material / grade change, or option confirmation occur. In one embodiment, the process selection event may function as a recalculation trigger for the tendency indicator recalculation unit (230), and may be configured so that the spatial priority (310) is recalculated in conjunction with the recalculated lifestyle tendency indicator (and the corresponding DNA type in the case of an embodiment that already includes the tendency DNA type (240),) the process classification and option configuration (320) are reconfigured, and the estimate of the estimate calculation / update unit (330) is updated accordingly.

[0078] Furthermore, the above event processing flow can be linked to the recalculation of the value indicator (HVI / ROI) of the actual transaction price-based valuation unit (400) and the updating of the risk evaluation / judgment result of the risk and basis provision unit (500). In an embodiment where the judgment result is reflected as a process configuration constraint (520), the constraint is fed back to the process selection input unit (130) and the process classification and option configuration unit (320) to form a cyclic structure that allows the user's selection to converge into a safe and consistent process configuration.

[0079] The lifestyle tendency indicator calculation unit (200) includes a question-indicator mapping calculation unit (210), an indicator cross-influence reflection unit (220), a tendency indicator recalculation unit (230), and a tendency DNA type determination unit (240).

[0080] The question-indicator mapping output unit (210) is configured to produce multiple lifestyle tendency indicators (e.g., 12 core indicators or 15 tendency tags (T01~T15), etc.) by applying a predefined many-to-many mapping rule to the question responses (e.g., 6 or 7 questions, or question bank-based expanded questions) collected from the tendency / lifestyle information input unit (120).

[0081] That is, the question-indicator mapping output unit (210) performs the role of quantifying how much each option (response) of a question influences a particular lifestyle tendency indicator, rather than the fact that "a question was asked," and converting it into an indicator vector that can be used by the subsequent remodeling decision output unit (300), the actual transaction price-based valuation unit (400), and the risk and basis provision unit (500).

[0082] The question-indicator mapping output unit (210) first receives "question identification information" and "user response" transmitted from the user input collection unit (100) as input. Here, some questions may be selected and presented based on the basic information of the target residence (square footage, year of construction, family composition, etc. entered in the residence basic information input unit (110)), and the question-indicator mapping output unit (210) is configured to standardize and process the response in the same way, whether there are 6 questions, 7 questions, or more extended questions actually presented.

[0083] To this end, the question-indicator mapping output unit (210) may refer to the "question bank" and "question-indicator mapping information" stored within the system, and the question-indicator mapping information may include (i) question ID, (ii) option (response) ID, (iii) affected indicator ID (multiple possible), (iv) contribution weight per indicator (including increase / decrease direction), (v) rules for processing duplication, absorption, and detailed attributes. Such mapping information is designed to support a many-to-many mapping structure in which, for example, "one response to one question contributes to multiple indicators simultaneously" or "responses to multiple questions contribute cumulatively to one indicator," and this becomes a characteristic processing structure of the present invention that is distinguished from simple survey scoring (one question -> one indicator).

[0084] Specifically, the question-index mapping output unit (210) normalizes the user's response into a form that allows for numerical calculation. For example, the choice-type response (either-choice, multiple-choice, one-priority choice, etc.) is converted into a score code corresponding to each choice, and if necessary, additional weights may be assigned according to the response intensity (emphasis response such as "highest priority," or question with high determinism such as "most regrettable").

[0085] Next, the question-indicator mapping output unit (210) queries a mapping table (question-indicator mapping information) using a normalized response as a key, and calculates a "raw score per indicator" by accumulating the contribution value for each indicator derived from the query result. At this time, the indicators can be defined as items directly connected to interior decision-making, such as spatial sense, visual sensitivity, auditory sensitivity, cleaning tendency, tidiness, budget sense, activity flow, daily routine, family composition influence, health factor, inconvenience factor, sleep pattern, etc. Depending on the implementation, the system can be configured so that a 12-core indicator system and a 15-tendency tag (T01~T15) system are used in parallel or interchangeably. Importantly, the question-indicator mapping output unit itself performs many-to-many mapping and cumulative calculation on the same principle regardless of "which system is used," and finally, the indicator value is scaled to a predetermined range (e.g., 1~100 points) or the upper and lower limits are clipped to determine the "lifestyle tendency indicator."

[0086] Additionally, the question-indicator mapping output unit (210) may be configured to perform mapping that considers "overlap, absorption, and detailed attributes." Here, overlap refers to cases where different questions repeatedly measure virtually the same tendency axis (e.g., preference for tidying up / storage), and absorption may refer to processing where, when there is a question with higher determinism (e.g., a question where one directly selects the top priority criterion, such as "I can never give up on this!"), the result partially replaces the contribution of other questions on similar axes or reduces their weight.

[0087] In cases where a detailed attribute is not a single concept but contains sub-attributes (e.g., sensitivity to contamination / allowing wiping water / avoiding mold, etc. within "cleaning tendency"), the question-indicator mapping output unit (210) can be implemented by first accumulating the sub-attribute contributions and then integrating them into a higher-level indicator, or by directly weighting and accumulating them into a higher-level indicator (T04, etc.) while recording the sub-attribute tags together. Through this, a "coherent indicator vector" reflecting semantic relationships between responses is produced rather than a simple summation, and a stable input basis is established for reflecting cross-influences (e.g., a correlation structure in which a sleep problem response simultaneously increases auditory sensitivity and independence preference) in the subsequent indicator cross-influence reflection unit (220).

[0088] The operation of the question-metric mapping output unit (210) can be intuitively explained through actual UI question examples. For instance, in the question "Guests are coming over suddenly! What is your first concern?", if the user selects "I have nowhere to hide clutter," this can be mapped to contribute positively (+) to the storage / organization-related metric (e.g., organization (T05) or storage-related metric), and conversely, if the user selects "I'm worried the house will look cramped and stuffy," it can be mapped to contribute positively (+) to the spatial sense or openness-related metric (e.g., spatial sense (T01)). Additionally, in the case where the user selects "Easy cleaning is better than beauty" in the question "When I want a hotel-like bathroom but am worried about maintenance," it can be mapped to contribute positively (+) to the cleaning tendency (T04) and maintenance avoidance tendency, and if the user selects "I will realize my dream even if I have to wipe water," it can be mapped to contribute positively (+) to the aesthetic / design priority tendency, sensory sensitivity, or style preference axis.

[0089] In this way, the response to a single question is structured to simultaneously influence multiple indicators that move together in the actual decision-making context (e.g., cleaning tendency and maintenance preference, or design tendency and trend acceptance), rather than raising "only one indicator," thereby enabling subsequent process recommendations, budget allocation, and risk assessment to operate more consistently.

[0090] In addition, the question-indicator mapping output unit (210) may be configured not only to output the final score but also to generate intermediate outputs for the "foundation chain (explainability)." That is, it stores or outputs a contribution log (e.g., question ID, response ID, contribution indicator, amount of contribution, applied weight) regarding how much contribution each indicator score received from which response to which question to form it, thereby enabling the subsequent foundation generation unit (530) or decision tracking logic to provide an explanation to the user in the form of a "selection foundation replay." At this time, the contribution log generated by the question-indicator mapping output unit (210) goes beyond simple explanation text generation and becomes the basic data for the foundation chain connecting "question response - lifestyle tendency indicator - judgment result," thereby structurally supporting the implementation of explainable decision-making (XAI) in the present invention.

[0091] The indicator cross-influence reflection unit (220) is configured to correct indicator values ​​by reflecting the correlation or cross-influence between the lifestyle tendency indicators so that the multiple lifestyle tendency indicators calculated by the question-indicator mapping calculation unit (210) in the first step sufficiently reflect "real-life problems."

[0092] In other words, the primary output in the question-indicator mapping output unit (210) creates a score based on the direct contribution of "user response -> corresponding indicator," but in real life, a problem or habit often does not appear as only one tendency but changes along multiple axes. The indicator cross-influence reflection unit (220) models these accompanying changes as "cross-influence information (cross-influence matrix or rule set)" within the system, and refines the user tendency vector into a more consistent and explainable form by correcting other lifestyle tendency indicators to rise or fall in conjunction with changes in specific lifestyle tendency indicators. This serves as a step for implementing a process (correction structure) that corrects indicators by referring to cross-influence information indicating correlations or cross-influences between lifestyle tendency indicators.

[0093] Specifically, the indicator cross-influence reflection unit (220) receives as input the "raw score (or primary indicator value)" generated by the question-indicator mapping calculation unit (210) based on the response to the tendency / lifestyle information input (120) collected from the user input collection unit (100). Here, the indicators may be expressed as a 12-indicator system (e.g., space efficiency, movement pattern, noise, health, family impact, budget elasticity, etc.) or a 15-tendency tag system (T01~T15) depending on the implementation, and the indicator cross-influence reflection unit (220) is configured to apply the influence relationship between "indicator ID-indicator ID" using the same principle regardless of the system. For example, among the 15 tags, auditory sensitivity (T03), cleaning tendency (T04), and tidiness (T05) may be directly calculated from different questions, but in real life, they may become entangled with sleep problems, childcare environment, whether working from home, weekend activity type, etc., and a single living condition may give a signal to multiple indicators simultaneously. The indicator cross-influence reflection unit (220) corrects such "entanglement" by unraveling it into predefined cross-influence information.

[0094] The cross-influence information referenced by the indicator cross-influence reflection unit (220) can be stored, for example, in the form of a "cross-influence matrix." This matrix arranges multiple indicators in rows and columns, and the influence coefficient (w) of a change in a specific indicator (i) on another indicator (j) ij It can be implemented as a structure that records ), and the influence coefficient can have a positive (+) or negative (-) sign, and the magnitude represents the index correction strength.

[0095] In addition, cross-impact information may include not only simple linear coefficients but also "rules triggered only under certain conditions" or "conditional rules applied only when a threshold is exceeded." For example, it can be implemented as a segment-specific function (gradual increase, upper saturation, etc.) to minimize cross-correction when a specific indicator is at an intermediate level, and to apply stronger cross-correction in segments where the indicator is very high or very low, clearly revealing daily inconveniences. This prevents the phenomenon of scores spiked due to excessive accumulation of cross-correction, while more sensitively capturing the accompanying inconveniences / demands of "users with distinct problems."

[0096] To easily explain the operation flow of the indicator cross-influence reflection unit (220), it is a method of first detecting a "companion signal" that might be missed when the primary indicator value (210) is left as is, and then adding a correction amount equal to that companion signal to another indicator.

[0097] For example, "sleep problems" do not simply alter a single indicator related to sleep patterns or rest; they can simultaneously be accompanied by a tendency to experience increased discomfort from noise in daily life (increased auditory sensitivity) and a desire to separate one's space (increased preference for independence).

[0098] In this case, when the question-indicator mapping output unit (210) first raises the sleep-related indicator from the sleep-related response, the indicator cross-influence reflection unit (220) can be configured to additionally raise the auditory sensitivity indicator (T03) in conjunction with the increase in the corresponding sleep-related indicator based on the cross-influence information, and to raise the indicator of the preference for independence (e.g., detailed indicator of the purpose of space use / life routine series) together. As a result, when a user inputs a single life problem such as "sleep is sensitive," the system can interpret it as a preference signal requiring "noise response (soundproofing / door / window / bedroom arrangement)" and "space separation (bedroom / study / middle door / separation of movement paths)."

[0099] As another example, if a user selects "freedom from cleaning / maintenance" as the top priority or responds that they are most concerned about "mold / maintenance" in the bathroom, the question-indicator mapping output unit (210) can reflect a direct contribution that significantly increases the cleaning propensity indicator (T04). However, in actual interior decision-making, the higher the cleaning propensity, the more likely it is that sensitivity to contamination / moisture / germs increases along with the cleaning propensity, going beyond simply "cleaning frequently," and this may also be accompanied by demands for health factors (ventilation, insulation, antibacterial properties, anti-slip properties, etc.). The indicator cross-influence reflection unit (220) can be set so that if the increase in cleaning propensity (T04) exceeds a certain threshold based on cross-influence information, the health factors or maintenance-related indicators increase together, and conversely, the "acceptability to options that are difficult to maintain" is adjusted downward. In this way, cross-correction serves to "expand the meaning of an indicator into a multidimensional signal of actual living needs" and is distinguished from the mechanical summation of simple survey scores.

[0100] The indicator cross-influence reflection section (220) can also reflect structural linkages such as between "organization / storage" and "movement / space efficiency." For example, if the user repeatedly selects "storage / organization" or a strong "hidden storage orientation" is observed, and the organization (T05) increases, this suggests not only that the user wants to increase storage furniture, but also that there is a possibility that there is a demand for "movement organization" or "space efficiency" so that frequently used items do not interfere with daily movement.

[0101] The indicator cross-influence reflection unit (220) can provide a foundation for automatically raising the priority of "organization bottleneck spaces" such as kitchens, entrances, pantries, and utility rooms in the subsequent space priority calculation unit (310) by partially transferring and correcting the increase in organization (T05) to the movement path / space efficiency indicators according to cross-influence information. In other words, cross-correction strengthens the connection between "cause (tendency) - effect (space)" so that the subsequent space priority algorithm operates more persuasively.

[0102] Meanwhile, the indicator cross-influence reflection unit (220) may include stabilization logic to prevent cross-correction from being amplified infinitely. The cross-influence matrix may have a cyclic relationship in which indicator A raises indicator B, and indicator B raises indicator A again; however, if this is applied repeatedly as is, a problem may occur where the score becomes excessively large.

[0103] Accordingly, the indicator cross-influence reflection unit (220) can be configured to reflect cross-correction "only as much as is needed" by applying an upper limit on the amount of correction (e.g., within ±n points per indicator in a single cross-reflection), a limit on the total amount of correction (e.g., a limit on the total fluctuation range relative to the primary indicator), or a convergence condition when applied repeatedly (e.g., terminate when the amount of change is below a certain threshold). Additionally, the final indicator value is clipped or normalized to a predetermined range (e.g., 1 to 100 points) and refined into an input form that can be reliably used by the estimate calculation / update unit (330) or the decision rule application unit.

[0104] The indicator cross-influence reflection unit (220) may generate a "cross-correction log" for explainability (providing evidence). That is, by recording how much a change in a certain causal indicator (e.g., sleep-related indicator) corrected a certain target indicator (e.g., auditory sensitivity (T03)) and by which cross-influence rule (or matrix item) the correction occurred, the subsequent evidence generation unit (530) can replay (replay) a "evidence chain" in natural language starting from the user's question response (120) and leading to the primary indicator (210), cross-correction (220), and final recommendation / warning (520).

[0105] For example, it becomes possible to provide an explanation such as, "Sleep-related responses were detected at a high level, causing sleep-related indicators to rise, and since these indicators tend to be associated with noise sensitivity, auditory sensitivity (T03) was further corrected to increase," which supports the fact that the present invention is not a simple 'recommendation' but a 'data-based decision-making algorithm'.

[0106] The tendency indicator recalculation unit (230) is configured to recalculate the lifestyle tendency indicator, which is calculated and corrected in the first step through the question-indicator mapping calculation unit (210) and the indicator cross-influence reflection unit (220), so that it is once again adjusted to a "realistic decision-making state" by the process and / or material selection information that the user actually selects and confirms.

[0107] That is, the tendency derived solely from the question and answer is close to the user's 'verbal preference (declarative tendency)' (210), and the reflection of cross-influence between indicators is a step of aligning that tendency to the living context (220). Then, the tendency indicator recalculation unit (230) implements a bidirectional structure leading from tendency score -> process selection -> recalculation of tendency value by automatically adjusting the tendency indicator again using the 'behavioral preference (behavioral tendency)' revealed by the process (e.g., kitchen / bathroom / wallpapering / built-in wardrobe, etc.) or material grade / option (e.g., Essential / Standard / Opus, cost-effective / balanced / premium type, etc.) actually selected by the user in the quotation flow as additional evidence.

[0108] In one embodiment of the present invention, the propensity indicator recalculation unit (230) can manage survey-based propensity (declarative propensity) and behavior-based propensity (behavioral propensity) in a dual manner. For example, a survey-based propensity vector calculated by answering questions V survey Let be denoted as , and V be the behavior-based propensity vector accumulated from the user's process / material / grade selection history. action If so, the processing unit is the final lifestyle tendency indicator vector V final It can be configured to be calculated by combining in the following form (weights are examples).

[0109] V final = α·V survey + (1-α)·V action

[0110] Here, α can be dynamically adjusted according to the degree of accumulation of the user's choice history (e.g., number of choice events, choice stability, frequency of change) or the recency of the survey response, and, for example, can be configured to gradually increase the weight of (1-α) as the choice history accumulates sufficiently.

[0111] Also, the above V finalIt acts as a reference signal for process classification and option configuration (320) and estimate calculation / update (330) to present the updated option / estimate to the user, and since the user can view it again and change the process / material selection, a bidirectional feedback loop can be implemented in which “process selection event (130) → tendency recalculation (230) → option reconfiguration (320) → estimate update (330)” is repeated.

[0112] The lifestyle tendency index, recalibrated in this manner, serves as supporting data for the process configuration and / or budget allocation of remodeling options and the subsequent updating of estimates; it also becomes a core function for realizing "index recalibration and option / estimate updates in response to changes in process and / or material selection information."

[0113] Specifically, the tendency indicator recalculation unit (230) may be configured to receive events such as user's process selection, release, material / grade change, and option confirmation from the process selection input unit (130) among the user input collection unit (100), and to recalculate the current lifestyle tendency indicator vector (e.g., T01~T15 or 12 extended indicator systems) whenever such events occur. At this time, the initial value serving as the basis for recalculation may be the indicator value calculated by the question-indicator mapping calculation unit (210) and the indicator value corrected by the indicator cross-influence reflection unit (220) reflecting cross-influence, and the tendency indicator recalculation unit (230) updates the indicator in the form of "current indicator value + selection-based correction amount" to maintain it in the latest state.

[0114] Therefore, even if the user adds or excludes processes in the middle, the system of the present invention can form a cyclic structure in which the selection does not merely change the quotation items but readjusts the user's propensity indicator itself, and the adjusted propensity indicator then influences the process recommendation and budget allocation logic.

[0115] The recalibration logic of the propensity indicator recalibration unit (230) can be performed by referring to a recalibration rule (or recalibration mapping information) that connects "process / material selection -> propensity indicator change".

[0116] The above recalibration rule defines which tendency indicators should be strengthened or weakened when a specific process or material is selected, and may be configured to reflect, for example, (i) the nature of the selected process (storage / cleaning / movement / noise / health / budget, etc.), (ii) the intensity of the selection (whether the process is classified as essential / recommended / optional, or whether the user voluntarily made an additional selection), (iii) options that reveal cost, quality, and maintenance tendencies together, such as material grades (cost-effective / balanced / premium or Essential / Standard / Opus), and (iv) conditions for combination with basic information such as the age / square footage / family composition of the target residence. According to such recalibration rule, the tendency indicator recalibration unit (230) regards the selection of the process / material as an "additionally observed user tendency signal" and dynamically adjusts the already calculated indicator value rather than keeping it as a fixed value.

[0117] For example, if a user selects storage-centered processes such as built-in closets, storage cabinets, or pantry expansion, or repeatedly selects options that enhance a sense of order rather than an open type, the tendency indicator recalibration unit (230) can adjust the order / storage-related tendency indicator (e.g., T05 series) upward, and can also adjust the space efficiency or movement-related indicator linked thereto within a limited range. Conversely, if a user selects options that enhance a sense of openness, such as "increasing the proportion of open space," "reducing built-in closets," or "securing visual space," the openness / minimalist-related indicators can be strengthened, and the tendency for excessive storage expansion can be adjusted to gradually decrease.

[0118] In addition, when the user confirms an option that prioritizes budget efficiency (cost-effective or Essential grade), the indicators related to budget sensitivity or budget elasticity are realigned to match the direction of the choice, and when the user confirms a choice that has a high proportion of aesthetics and presentation, such as premium materials, designer lighting, marble / imported finishes, the indicators can be corrected in a direction that strengthens the aesthetics / design priority tendency. In this way, the tendency indicator recalculation unit (230) converts a simple survey-based profile into a "decision-based profile" by reflecting back "where the user actually allocated money and process scope" into the tendency indicators.

[0119] In addition, the propensity indicator recalculation unit (230) can apply a recalculation strategy so that the conflict itself becomes a meaningful signal, rather than unilaterally following only one side even when the user's choice conflicts with the initial propensity. For example, if a user (T04 series) with high cleaning and maintenance sensitivity selects materials / configurations that are difficult to maintain, this may be (a) a sign that the user has chosen aesthetics even at the cost of maintenance burden, or (b) a signal that the user may be underestimating the actual difficulty of maintenance.

[0120] In this situation, the tendency indicator recalculation unit (230) can provide an input state in which the subsequent judgment granting unit (520) or complex risk evaluation unit (510) can persuasively generate WARN / BLOCK or supporting sentences by strengthening the aesthetic indicator while relaxing the cleaning tendency indicator within a certain range, or by maintaining the cleaning tendency indicator while reflecting the axis corresponding to 'management tolerance (maintenance tolerance)' as a separate correction amount. In other words, recalculation is not simply about matching scores, but rather a function that records "trade-offs revealed by the user's choice (e.g., aesthetics vs. management, openness vs. storage, cost vs. quality)" in the indicator vector.

[0121] For the stability of the recalibration process, the propensity indicator recalibration unit (230) may limit the correction range or perform normalization / clipping so that the cumulative correction does not become excessive. For example, an upper limit may be placed on the correction amount per process so that a specific indicator does not shift drastically with a single process selection, and a damping coefficient or reflection rate may be applied so that the indicator does not fluctuate even with repeated input of selection / deselection (130), and the final indicator value may be limited to a predetermined range (e.g., 1 to 100 points).

[0122] In addition, to maintain consistency between indicators, the indicator cross-influence information may be referenced when necessary for the result of the recalibration by the propensity indicator recalibration unit (230), and fine-tuned so that the indicator vector after recalibration does not contradict the living context. Through this, the entire indicator structure naturally follows whenever the user's choice changes, while ensuring that the system does not become unstable.

[0123] Furthermore, the propensity indicator recalculation unit (230) may be configured to generate and store "recalculation basis data" along with the recalculated lifestyle propensity indicator. For example, if a log is maintained regarding how much and in what direction a certain process selection (130) influenced a certain propensity indicator and what the application rule is, the basis generation unit (530) can construct a basis chain leading to "question response (120) -> primary indicator (210) -> cross-correction (220) -> recalculation based on process selection (230) -> final recommendation / warning (520)" and provide it to the user in a form that explains "why this recommendation was given."

[0124] The tendency DNA type determination unit (240) receives as input the final vector of a plurality of lifestyle tendency indicators (e.g., T01~T15) calculated through the question-indicator mapping calculation unit (210), the indicator cross-influence reflection unit (220), and the tendency indicator recalculation unit (230), and determines a tendency DNA type that represents the user's lifestyle type.

[0125] In the present invention, the propensity DNA type is a multidimensional propensity indicator vector summarized as a "higher lifestyle type that can be explained and controlled," and can be expressed as one or more of a plurality of DNA types, such as a practical family type, a minimal life type, a management liberation type, a home-focused type, and a home party-oriented type. The propensity DNA type determining unit (240) may be configured not only to select a single type, but also to determine the top two types together in a mixed ratio (e.g., practical family type 0.65 + management liberation type 0.35) to reflect the user's complex propensity.

[0126] In one embodiment, the propensity DNA type determining unit (240) may include the following clustering-based classification step.

[0127] (a) Normalize the user preference tag vector X = [x1, x2, ... , x15](T01~T15) to a predetermined range (e.g., 0~100 or 0~1) and correct for missing values ​​or outliers.

[0128] (b) Refer to the pre-constructed set of DNA centroid vectors C = {c1, c2, ... , ck}, where each centroid ck maps to one DNA type (Dk), and the centroids can be set based on historical user data or expert rules.

[0129] (c) Calculate the similarity or distance (e.g., cosine similarity, Euclidean distance) between the user vector X and each cluster center ci. For example, in the case of distance-based, calculate d(i) = ||X - ci|| and select the cluster with the minimum d(i) as the first-ranked DNA.

[0130] (d) In addition to the first-ranked DNA, if there are multiple clusters where similarity (or distance) satisfies a threshold, select the top two or more DNA types and calculate the mixing ratio of each DNA type in proportion to the similarity (e.g., p(i)=sim(i) / Σsim).

[0131] (e) Finally, the propensity DNA type determining unit (240) outputs a “DNA type identifier” and a “mixing ratio (optional)”, and the output is stored or transmitted so that it can be used as a common reference signal in a downstream module.

[0132] The output of the tendency DNA type determination unit (240) is linked to subsequent decision-making. For example, the process classification and option configuration unit (320) can determine the trade-offs (e.g., variable minimization type / balance type / performance enhancement type) of option A / B / C configurations by applying different process necessity weights according to the DNA type, even if the lifestyle tendency index value is the same. Additionally, the estimate calculation / update unit (330) can ensure that the budget "personalization criteria" operate consistently by applying different upper and lower limits of the basic ratio or weights for budget allocation by process according to the DNA type.

[0133] Additionally, if the user's process or material selection is changed and the propensity indicator is recalibrated (230), the propensity DNA type determining unit (240) can recalculate the DNA type using the recalibrated indicator vector as input and update it to the latest state.

[0134] FIG. 3 is a block diagram schematically showing the detailed configuration and interconnected flow of the remodeling decision-making output unit, the actual transaction price-based valuation unit, and the risk and basis provision unit according to the present invention.

[0135] Referring to FIG. 3, the remodeling decision-making output unit (300) includes a space priority calculation unit (310), a process classification and option configuration unit (320), an estimate calculation / update unit (330), and a life satisfaction index calculation unit (340).

[0136] The space priority calculation unit (310) is configured to calculate a "priority score" for each of multiple spaces, such as a kitchen, bathroom, living room, and bedroom, and to arrange the order of the spaces according to the score, so that the process configuration and budget allocation of the subsequent stages do not become arbitrary or result in simple trend recommendations.

[0137] That is, the space priority calculation unit (310) operates as a converter (tendency -> space priority) that causes the influence to first manifest in the "space" unit in a structure in which the lifestyle tendency indicator directly affects the actual process, materials, budget, and priority.

[0138] The space priority calculation unit (310) uses basic housing information (110), tendency / lifestyle information (120), and process selection information (130) collected from the user input collection unit (100) as inputs. Additionally, multiple lifestyle tendency indicators (e.g., tendency tags T01~T15 or expanded multidimensional indicators) calculated by the lifestyle tendency indicator calculation unit (200) are used as the primary basis data for calculating scores by space. In particular, the lifestyle tendency indicators are calculated according to the many-to-many mapping rule of the question-indicator mapping calculation unit (210) (a structure in which one answer contributes to multiple indicators in a weighted manner, or multiple answers contribute cumulatively to one indicator), and since the change of a specific indicator can be linked to and corrected by other indicators through the cross-influence information of the indicator cross-influence reflection unit (220), the indicators referenced by the space priority calculation unit (310) are not "simple survey sum values" but "cross-corrected tendency vectors."

[0139] The space priority calculation unit (310) can be configured to calculate a space priority score for each of the multiple spaces by reflecting a lifestyle tendency indicator and an addition / subtraction factor based on tendency / lifestyle information, starting from a reference value.

[0140] For example, the initial score of each space can be assigned the same reference value (e.g., 50 points), and then the final score can be derived by cumulatively applying three types of adjustment items: (i) adjustments based on lifestyle tendency indicators, (ii) adjustments based on inconvenience factors, and (iii) adjustments based on housing conditions. The reason for adopting the "reference value + adjustment" structure is that the comparison between spaces is intuitive (all spaces have the same starting point), and it is easy to explain in the later stages, using natural language evidence, which factors raised or lowered the score (explainability).

[0141] First, the addition or subtraction based on the lifestyle tendency indicator can be calculated by referring to "space-indicator weight information (mapping)" which indicates the correlation between the lifestyle tendency indicator (200) and the space (kitchen / bathroom / living room / bedroom, etc.). For example, if the tidiness / storage tendency is calculated to be high, additional points can be given to spaces with a high sense of storage (kitchen, entrance, bedroom storage area, etc.) to increase the priority of the space; and if the cleaning sensitivity or maintenance sensitivity is calculated to be high, additional points can be given to spaces where the burden of management is concentrated, such as contamination, moisture, mold, and detergent use (kitchen, bathroom). Additionally, if the spatial sense / openness tendency is calculated to be high, the score of the living room, which has a high sense of sight, furniture arrangement, and movement, can be added to increase the priority of the living room; and if the auditory sensitivity or sleep-related tendency is calculated to be high, the need for sound insulation, lighting, light blocking, and air quality improvement in the bedroom can be reflected to increase the priority of the bedroom.

[0142] At this time, when the cross-influence matrix of the indicator cross-influence reflection unit (220) is applied and a cross-change occurs such as “sleep problem response -> simultaneous increase in auditory sensitivity and independence preference,” the space priority calculation unit (310) can operate in a way that further raises the score of the bedroom or personal space by including the cross-correction result.

[0143] Next, the addition / subtraction based on inconvenience factors is a part that treats the "inconvenient space" or "inconvenient type" explicitly revealed by the user in the tendency / lifestyle information input section (120) or during the question-and-answer process as a strong signal in the priority calculation. Since one of the differentiating points of the present invention is a "priority algorithm in which the tendency score automatically raises specific inconvenient spaces (kitchen / bathroom / living room, etc.)," ​​the space priority calculation section (310) can grant bonus points (addition) or weighted upwards to the space where the user has indicated inconvenience, so that the space with the greatest perceived improvement can be placed at the top even with a limited budget.

[0144] For example, when an inconvenience such as "lack of storage" is input, before simply recommending the "carpentry / built-in" process, the score of a space with a high sense of storage, such as a kitchen, entrance, or bedroom storage area, is first added to raise it to a higher level, and the subsequent process classification and option configuration section (320) makes it easy to classify the processes related to that space (built-in wardrobe, pantry, kitchen storage, etc.) as essential or recommended. In addition, when an inconvenience such as "bathroom mold / odor / water stains" is input, the bathroom score is raised significantly, but for users whose cleaning sensitivity index is also calculated to be high, the addition range is increased even for the same inconvenience, so that the combined effect of "inconvenience × tendency" is reflected. This combined structure ultimately creates a personalized priority where "the priority of the same inconvenience is calculated differently for each user."

[0145] In addition, the addition / subtraction based on housing conditions is a part that reflects "physical and living constraints," such as area, year of construction, and family composition, collected from the housing basic information input section (110) into the priority score.

[0146] For example, in cases where the building is old (e.g., 20 years or more) and there is a high possibility of equipment deterioration, the score of wet spaces (bathroom, kitchen) where the risk of defects such as leaks, waterproofing, and plumbing is relatively high can be adjusted upward to encourage "risk avoidance over aesthetics." Conversely, in new or semi-new construction, if input such as "preserving what can be preserved and focusing on film / wallpapering" (110) is received, the wallpapering, lighting, and furniture arrangement of the living room / bedroom may be relatively more significant than structural changes or equipment work, so the score of common spaces and finishing-focused spaces can be adjusted in a direction that prioritizes raising the score.

[0147] In terms of family composition, the layout can be configured to increase scores for spaces where hygiene, ventilation, and safety (slip surfaces, corners, and uneven surfaces) are important (such as bathrooms, living rooms, and entrances) when infants and toddlers are present. For families with elderly parents, scores can be raised for spaces where safety is tangible, such as those with simplified circulation paths, door thresholds / levels, handrails, and non-slip features. Additionally, for families with pets, it is possible to adjust spatial priorities by adding scores for floors, living rooms, and hallways where dirt, scratches, and slipping directly impact living satisfaction.

[0148] After performing the above-mentioned addition / subtraction calculation, the space priority calculation unit (310) may include upper / lower limiting of scores or normalization processing to prevent the scores from becoming excessively skewed by space or fluctuating rapidly according to specific input changes.

[0149] For example, space scores can be limited to a range of 0 to 100, or the gap between upper and lower spaces can be smoothed (e.g., capping / smoothing) so that the gap between upper and lower spaces does not become excessively large, thereby stabilizing the subsequent option generation so that it does not produce an irrational result of being "excessively biased toward only one space." In addition, if there are multiple spaces with the same score or similar scores (e.g., difference is less than a threshold), consistency of alignment can be ensured by applying a tiebreak rule that first compares the user's selection process (130) or the existence of an inconvenience factor (120).

[0150] The output of the space priority calculation unit (310) may include at least "space priority scores by space identifier (e.g., kitchen / bathroom / living room / bedroom)" and "space alignment results," and depending on the embodiment, may include a "contribution log" or "foundation data" that records the major factors that contributed to the score calculation (e.g., which lifestyle tendency indicators influenced which direction, which inconvenience inputs generated a bonus, which residential conditions caused a correction).

[0151] The space priority calculated in this manner is directly used in the process classification and option configuration unit (320) and the estimate calculation / update unit (330). For example, when the process classification and option configuration unit (320) applies a classification rule to classify processes corresponding to the upper space into mandatory / recommended / optional, it can use the space priority score as a core criterion to arrange "core processes of the upper space as mandatory or recommended" and "optional processes of the lower space as optional." Additionally, when creating multiple remodeling options with different budget levels, material grades, or process scopes, the process configuration of options A / B / C can be configured differently so that the upper space is included first, and the estimate calculation / update unit (330) can complete an integrated structure in which "preference-based priority is reflected in the estimate number" by adjusting the budget allocation ratio in proportion to the space priority score and updating the amount per process.

[0152] The process classification and option configuration unit (320) is an “engine that translates space priority (310) and lifestyle tendency indicators (200) into mandatory / recommended / optional process classifications and 2 to 3 quotation options (e.g., minimum / recommended / premium or A / B / C)” and serves to determine process bundles and option parameters (material grade, range, inclusion / exclusion rules, etc.) in a form that can be calculated by the subsequent quotation calculation / update unit (330).

[0153] The process classification and option configuration unit (320) utilizes multiple input signals. Additionally, the process classification and option configuration unit (320) may refer to the DNA type (and mixing ratio) determined by the propensity DNA type determination unit (240) as an additional input, and by applying a predefined "process preference / avoidance weight table" or "option configuration policy (e.g., variable minimization / balance / performance enhancement)" for each DNA type, the configuration criteria for options A / B / C may be varied to suit the user type even under the same spatial priority score (310) and lifestyle propensity index (200) conditions.

[0154] First, the user input collection unit (100) collects basic residential information (110) (square footage, region, year of construction, etc.) and family composition and living conditions (120) to establish a basic premise of “what processes are structurally / livingly necessary in this house.”

[0155] Second, by referring to the tendency tags / indicators (e.g., cleaning tendency, tidiness, budget sensitivity, movement path, noise sensitivity, etc.) calculated by the lifestyle tendency indicator calculation unit (200) and the result of reflecting the cross-influence of indicators (220), weights are assigned so that the necessity of the process varies depending on the user even under the same house conditions.

[0156] Third, by referring to the space importance (or space priority) score (e.g., kitchen > bathroom > living room...) calculated by the space priority calculation unit (310), the process necessity is reflected in the calculation so that the priority of the process directly connected to the upper space increases and the priority of the process of the lower space decreases.

[0157] Fourth, the range of processes directly selected / deselected by the user in the process selection input section (130) can be reflected as a strong constraint so that "processes that the user intends to proceed with are necessarily managed in the option group (included or warned), and processes excluded by the user are, in principle, excluded from the options or alternative solutions are created." Additionally, according to the embodiment, the results of the complex risk assessment (510) and judgment assignment (520) of the risk and basis provision section (500) can be reinjected as process configuration constraints so that if a specific process combination is dangerous in terms of defects / maintenance / asset value, it is automatically excluded from the options (or replaced with a WARN / BLOCK-based alternative process).

[0158] The core of the process classification and option configuration unit (320) is to calculate a “process necessity score” on a process unit basis and classify it into mandatory / recommended / optional. To this end, the process classification and option configuration unit (320) first configures a set of candidate processes from a process master (e.g., demolition, wallpapering, flooring, electrical / lighting, kitchen, bathroom, tile, built-in wardrobe, sash, interior door, film, painting, etc.), and can filter the process range into partial construction / intensive construction / full remodeling based on user input (110, 130).

[0159] Subsequently, the necessity score for each process can be calculated by summing the following: (i) the contribution of the space priority score (310) of the connecting space, (ii) the contribution of the mapping weight between the lifestyle tendency index (200) and the process, (iii) the contribution of residential conditions (110, 120) such as the year of construction, area, and family composition, and (iv) the contribution of inconveniences / goals (120) complained about by the user. For example, if the kitchen priority is high, the tendency for storage / organization is high, and cooking and eating at home are important patterns due to the family composition, the necessity of kitchen processes and kitchen storage / built-in series processes increases. Conversely, if budget sensitivity is very high or a partial construction scope is selected, "high-variable processes (processes with many options and large fluctuations in estimates)" can be configured to be reduced from the necessity score and downgraded to recommended or optional.

[0160] In addition, in an embodiment of the present invention, the process classification and option configuration unit (320) may be configured to directly control the option configuration by receiving as input not only the PASS / WARN / BLOCK judgment result calculated by the judgment granting unit (520), but also constraint information generated in conjunction with the judgment. The constraint may include, for example, rules for (i) requiring the inclusion of a specific process, (ii) blocking the selection of a specific process or a specific combination, and (iii) replacing a specific process or material with an item that is functionally equivalent or has a lower risk.

[0161] Accordingly, the process classification and option configuration unit (320) can reconstruct the option structure itself by automatically excluding items with block constraints from options A / B / C or replacing them with alternative items, promoting items with require constraints to essential processes and including them in all options, and in the case of warnings, configuring option parameters in the form of conditional inclusion (e.g., allowed when additional reinforcement processes are included). At this time, it is important for the process classification and option configuration unit (320) to apply both "hard rules (mandatory rules)" and "soft rules (weighted rules)" together rather than dividing essential / recommended / optional based only on a simple score threshold.

[0162] A hard rule is a rule that elevates a specific process to a de facto mandatory level when the construction age or family composition (110) of the target residence satisfies specific conditions. For example, in a building (e.g., 20 years or older) where a bathroom process is included, if a process of waterproofing and equipment inspection is not bundled together, the risk of defects may increase. Therefore, the process classification and option configuration section (320) can be implemented in such a way that the bathroom-related basic process is classified as mandatory, or at least elevated above the recommended level, and is "dependently included" so that it is not omitted from the options.

[0163] On the other hand, soft rules are rules that flexibly raise and lower the necessity of processes according to the tendency indicator (200) or space priority (310). For example, if the cleaning sensitivity (T04) is high, elements with a high maintenance perception, such as bathroom tiles, grout, and ventilation, are reinforced in the upper grades of the options, and if the design priority tendency is high, processes for presentation, such as lighting, painting, and film, are raised to recommendations.

[0164] In addition, the process classification and option configuration unit (320) ensures logical consistency of the options by reflecting the "dependency relationship between processes (including preceding and succeeding / simultaneous)." For example, if there is no demolition process, construction of a specific finishing process may be impossible or quality may become unstable. Therefore, the process classification and option configuration unit (320) can automatically include necessary preceding processes (e.g., demolition, waste disposal, site cleanup) together with the selected process (130) (including mandatory or basic), or leave "process-basis linkage data" so that a warning can be given via the subsequent judgment / basis (520, 530) if the user wishes to exclude them.

[0165] In the same context, when changing bathroom tiles, connection with waterproofing / floor slope / drainage is required, and when changing kitchen sink / countertop, connection with electrical / water supply / drainage / hood piping is required, so the process classification and option configuration section (320) can manage process bundles as "function bundles (packages)" rather than as single processes.

[0166] The second axis of the process classification and option configuration section (320) is "option configuration." The process classification and option configuration section (320) can be configured to generate and provide multiple (typically 2 to 3) remodeling options, each having at least one different budget level, material grade, or process range. In this case, the options can be designed not simply as "three quotes with different prices," but as a structure that reflects user preferences and trade-offs.

[0167] For example, similar to the A / B / C options based on Interibity results, they can be differentiated by dividing them along axes of "strengthening standards (minimizing variables)," "maintaining standards (balance)," and "adding tangible impact (enhancing presentation)." Option A minimizes price fluctuations by reducing optional variables around essential processes, Option B includes a balanced mix of essentials and recommendations, and Option C aims to increase satisfaction by including tangible presentation processes (lighting, accent finishes, etc.) or additional options.

[0168] Alternatively, it is also possible to set different material grades and construction details for the same process group by using different option parameters such as Essential / Standard / Opus, material grades, and construction quality (standard / meticulous / superior). The important point is that the process classification and option configuration unit (320) defines not only the "included process list" but also "process-specific option parameters (material grade, range coefficient, selectability, upper / lower limit constraints)" for each option, so that the estimate calculation / update unit (330) can consistently calculate the amount for each option.

[0169] In the option configuration process, the process classification and option configuration section (320) can control that "essential processes are included in all options," the inclusion ratio of recommended processes increases as the option grade increases, and optional processes are exposed only in option C or premium grades or are reflected only when explicitly selected by the user.

[0170] For example, the minimum option (cost-effective type) is composed of essential processes plus some recommended processes and focuses on resolving the core of daily inconveniences, the recommended option (balanced type) includes sufficient recommended processes to raise the life satisfaction score, and the premium option can be designed to include optional processes or raise the material / construction level to aim for simultaneous increase in satisfaction and value. At this time, since the process configuration for each option directly affects the process range coefficient and scenario comparison of the subsequent real transaction price-based valuation, it is desirable for the process classification and option configuration section (320) to maintain the differences between options in an explainable form to make it possible to compare "how the difference in process range is reflected in the value indicator (HVI / ROI)."

[0171] Additionally, the process classification and option configuration unit (320) may be configured to support "dynamic options that are reconfigured according to changes in user selection." That is, when a user adds or deletes a process in the process selection input unit (130), or when the tendency indicator recalculation unit (230) recalculates the tendency value according to the selected process, the process classification and option configuration unit (320) re-performs the mandatory / recommended / optional classification by reflecting the recalculated lifestyle tendency indicator (200) and the changed process range (130), and regenerates option A / B / C or grade options.

[0172] As a result, the estimate calculation / update unit (330) can update the process configuration and amount in real-time or near-real-time while maintaining the same option system, and the risk and basis provision unit (500) can explain "what changed and why the option changed" in the form of a basis chain.

[0173] Finally, the process classification and option configuration unit (320) is also connected to the "before / after image 1:1 mapping by process" structure and the structure providing explainable grounds, which are emphasized in the present invention. That is, the process list determined by the process classification and option configuration unit (320) for each option can be used as reference data for process selection in a process that generates "before / after images of construction that reflect only the change elements corresponding to the selected process" by referring to process-image mapping information.

[0174] At the same time, if the reasons why each process is classified as mandatory / recommended / optional (space priority (310), propensity indicator (200), housing conditions (110), risk assessment (510), etc.) are structured, the subsequent basis generation unit (530) can provide an XAI output explaining "why this process is mandatory / why this process is included only in Option C" in natural language connected to user propensity.

[0175] The estimate calculation / update unit (330) is configured to generate key outputs (total construction cost and amounts per process / item) that are perceived on the actual service screen in the present invention through "Tendency -> Process -> Estimate," and functions as a "dynamic estimate engine" that automatically recalculates the estimate whenever an input changes, such as process selection input (130) or tendency indicator recalculation (230), so that it is maintained in the latest state.

[0176] The estimate calculation / update unit (330) basically receives as input basic residential information (110), such as the area, region, and year of construction of the target residence, collected from the user input collection unit (100), family composition and lifestyle information (120), multiple lifestyle tendency indicators (e.g., cleaning tendency, tidiness, budget tendency, movement pattern preference, etc.) calculated from the lifestyle tendency indicator calculation unit (200), and process configurations by option (including mandatory / recommended / optional) and option parameters (e.g., material grades such as Essential / Standard / Opus, process scope such as partial construction / full remodeling) determined from the process classification and option configuration unit (320).

[0177] The estimate calculation / update unit (330) can be configured to use these inputs to generate and provide estimate data including "total construction cost," "direct construction cost (total by process)," "indirect cost," "VAT," and "detailed breakdown by process (labor cost / material cost / total)" for each option. For example, as displayed on the user screen, amounts are broken down and presented by process unit, such as demolition work, wallpapering work, flooring, electrical / lighting, bathroom, kitchen, door replacement, site cleanup / waste, etc. Each process is further divided into labor cost and material cost to calculate the total, and subtotals, supervision / design costs (e.g., a certain percentage), and whether VAT is reflected can be managed together.

[0178] More specifically, the estimate calculation / update unit (330) may first be configured to calculate "basic costs based on area and region." This is a step of calculating a basic amount based on the area and region (110) of the target residence and setting a standard amount by reflecting the material grade or construction grade (320) specified in the options.

[0179] For example, the standard construction cost (basic amount) for each option can be derived by referring to a standard value corresponding to a combination of area, region, and grade, such as "32 pyeong × Gyeonggi-do × Standard grade." In this case, if the building age (110) is high and the need for foundation reinforcement work increases, a correction factor (e.g., construction addition) may be applied to the standard amount according to the results of the process configuration (320) or risk assessment (500) described later, or a specific foundation work may be automatically included, resulting in an increase in the total amount.

[0180] Next, the estimate calculation / update unit (330) may be configured to apply a process-specific distribution ratio (e.g., a predefined ratio such as kitchen 22%, carpentry 28%, bathroom 15%, flooring 12%, etc.) to distribute the above-mentioned basic amount by process. Here, the "process-specific distribution ratio" may be implemented as a reference value stored in the process master or estimate master, and the process classification and option configuration unit (320) determines the processes to be distributed according to the process range (part / whole) and the list of included processes selected by the process classification and option configuration unit (320), and ensures that distribution is performed consistently among the determined processes.

[0181] For example, if the number of processes is limited due to partial construction, the ratio allocated to non-selected processes can be redistributed to selected processes, or the total amount itself can be reduced by applying a process range coefficient (option parameter) and then performing allocation by process. At this stage, the estimate calculation / update unit (330) secures the "basic allocation amount per process (before reflecting tendencies)."

[0182] As the next step, the estimate calculation / update unit (330) personalizes the estimate by applying process-specific weights based on the lifestyle tendency index (200). At this time, the estimate calculation / update unit (330) may refer to the DNA type determined by the tendency DNA type determination unit (240) as a higher standard signal for budget allocation. For example, if the practical family type DNA is selected, the upper limit of the weight for durability and maintenance-centered processes (basic facilities, storage / movement flow improvement, etc.) is gradually increased, and if the minimal life type DNA is selected, the lower / upper limit of the weight is controlled so that the proportion of processes contributing to a sense of openness and order (reduction of unnecessary storage, lighting / finishing organization, etc.) does not increase excessively. This configuration ensures that the application of process-specific weights is performed stably based on the combination of "indicator (200) + DNA (240)".

[0183] This corresponds to the "structure in which lifestyle habits and family types are converted into scores and the scores directly influence the actual process, material, and budget allocation," which is a significant distinguishing feature of the present invention. Specifically, it can be configured to calculate the rate of increase or decrease for each process by referring to a propensity-process mapping (weight matrix) that indicates how each propensity indicator affects each process, and to reflect this in the basic allocated amount for each process.

[0184] For example, if the storage / organization tendency (T05) is calculated to be high, the weight of processes related to built-in wardrobes / carpentry / storage can be increased to raise the budget for those processes; if the cleaning sensitivity (T04) is high, the budget can be increased by improving the finishing quality or material level in processes with a high maintenance impact, such as floors / bathrooms / kitchens; and if the budget sensitivity is high, high-variable and high-cost options can be restricted to prevent the process cost from rising excessively.

[0185] At this time, the range of weight application may be limited to a maximum range of ±30% per process to prevent excessive fluctuation, and in cases where multiple tendencies act simultaneously, the final adjustment rate per process may be determined by applying a weighted sum or priority rule. Additionally, when the tendency value is corrected by the indicator cross-influence reflection unit (220) and the tendency indicator re-correction unit (230), a bidirectional structure is realized in which the weight per process changes even with the same process configuration, and the estimate is automatically recalculated.

[0186] Meanwhile, the estimate calculation / update unit (330) may be configured to reflect material costs by referencing cost or unit price data corresponding to the selected process and / or material, rather than calculating the "amount per process" as a simple ratio. That is, when a specific material grade (e.g., Essential / Standard / Opus) is selected by the process selection input unit (130) or option parameter (320), or when a specific material (e.g., countertop, tile, lighting, etc.) is selected, the estimate calculation / update unit (330) may be configured to determine material costs by querying unit price / cost data corresponding to the material (e.g., Material Rebot cost DB linkage or internal unit price master), and to calculate labor costs by applying standard man-hours or unit prices according to the process difficulty, scope, and area.

[0187] As a result, a structure becomes possible where labor and material costs are separated by process, as shown in the "Detailed Estimate" on the screen, and the material costs for the corresponding process are immediately updated when the selected materials change. Additionally, even within the same process, calculation rules can be established for each option so that the proportion of material costs increases when the material grade is upgraded, or labor costs (man-hours) increase when construction quality parameters, such as meticulous or premium construction, are enhanced.

[0188] In addition, the estimate calculation / update unit (330) may be configured to automatically calculate indirect costs in addition to direct construction costs (total by process) to calculate the final total. For example, indirect cost items such as industrial accident insurance premiums, employment insurance premiums, and corporate profit may be calculated at a predefined rate (e.g., industrial accident insurance premium 3.07%, employment insurance premium 1.5%, corporate profit 5%, etc.) and added to the direct construction costs, and if necessary, design / supervision costs (e.g., a certain percentage of direct construction costs or subtotals, 9% in the screen example, etc.) may be calculated as a separate item and reflected in the final total.

[0189] Next, the inclusion or exclusion of VAT can be selectively applied according to options or output formats, so that if "VAT separate" is required, the VAT is displayed separately, and if "final amount including VAT" is required, the total amount including VAT is calculated. In this way, the estimate calculation / update unit (330) contributes to transparent estimate output that prevents "blind estimates" by consistently providing a decomposition structure (direct costs / indirect costs / VAT, labor costs / material costs by process) in a form that is understandable to the user, rather than a simple sum.

[0190] Another key feature of the estimate calculation / update unit (330) is "update." In the present invention, the user can add / delete processes through the process selection input unit (130), and the tendency indicator is recalibrated (230) according to the selected process or material, and the process configuration and / or budget allocation of the remodeling option and the estimate are configured to be updated in conjunction with the recalibrated tendency indicator.

[0191] To this end, the estimate calculation / update unit (330) can be configured to recognize (i) a process selection change event (130), (ii) an option change event (e.g., Essential -> Standard, Option A -> 320), (iii) a propensity indicator recalculation event (230), (iv) a housing basic information change event (modification of area / year, etc.) (110), or (v) a process forced inclusion / exclusion event based on risk assessment (500) as update triggers, and when a trigger occurs, to re-acquire the latest input value and re-execute the estimate calculation pipeline (basic cost -> process allocation -> propensity weighting -> unit price reflection -> indirect cost / VAT). In particular, if only a part of the process is changed, rather than unconditionally recalculating the entire process, partial updates can be performed focusing on the changed process and the items linked thereto (e.g., preceding process, process range coefficient, indirect cost) so that they are immediately reflected on the user screen as a "real-time estimate."

[0192] In one embodiment for implementing a partial update, the estimate calculation / update unit (330) may maintain a dependency graph or an influence range table representing the dependency relationship between processes and parameters. For example, the amount per process (labor / material costs), process range coefficient, indirect cost item, whether VAT applies, whether preceding processes are included, etc., may be modeled as nodes / parameters, and the process or material / grade changed by the user may be mapped in advance to which parameters (influence range) are affected, and then configured to perform incremental recalculation only on the set of changed nodes and linked nodes when a change event occurs.

[0193] In addition, during partial updates, dependent total items such as the total amount, direct construction costs / indirect costs / VAT, and detailed totals by process can be updated by reflecting the increment (Δ) of the changed items. By configuring the option comparison screen to selectively update only the required values, computational efficiency compared to full recalculation and the responsiveness of real-time estimation implementation can be improved.

[0194] Furthermore, the estimate calculation / update unit (330) also performs the role of calculating multiple option estimates to enable comparison. When the process classification and option configuration unit (320) generates multiple options (e.g., Minimum / Recommended / Premium or Essential / Standard / Opus) with different budget levels, material grades, or process ranges, the estimate calculation / update unit (330) calculates the estimates for each option in parallel by applying only different parameters for each option within the same calculation rule system, and can increase the possibility of comparison by including in the output data what has changed between options and caused the amount to change (e.g., difference in material grade, difference in included process, difference in process range coefficient).

[0195] The total construction cost per option calculated in this way is linked to the input (total construction cost) of the subsequent real transaction price-based valuation unit (400), particularly the HVI / ROI calculation and scenario comparison unit (430), so that "how the estimated increase in house price and ROI change when this option is selected" can be calculated together for each option. In addition, since the risk and basis provision unit (500) can explain in a basis chain which propensity indicator raised which process cost or why a specific material is at risk of exceeding the budget in the estimate calculation, the estimate calculation / update unit (330) can be configured to store or transmit the results of applying weights per process and unit price reference results together as "logs / metadata in an explainable form."

[0196] The life satisfaction index calculation unit (340) is configured to calculate the satisfaction with housing improvement perceived by the user as a quantitative index and provide comparisons for each option. The life satisfaction index (LII) is an output distinct from the asset value indicator (HVI / ROI), and can use the user's lifestyle tendency indicator (200), tendency DNA type (240), space priority score (310), process configuration / scope (320), estimate / material grade (330), and defect / anxiety factors (e.g., risk signals related to defects and maintenance calculated by the complex risk assessment unit (510)) as inputs.

[0197] In one embodiment, the life satisfaction index (LII) can be calculated by normalizing it to a score in the range of 0 to 100, and, for example, can be calculated by a weighted sum formula as follows (weights are in the example).

[0198] Satisfaction Score (LII) = (Spatial Improvement × 0.4) + (Psychological Factors × 0.3) + ((100 - Defect / Anxiety Risk) × 0.3)

[0199] Here, "space improvement" can be calculated by combining the space importance (or space-specific priority) score calculated by the space priority calculation unit (310) and the improvement contribution of the process combination determined by the process classification and option configuration unit (320) to the space. For example, the space improvement score can be configured to increase as the core process for the higher priority space (the core process of the space important to the user among the kitchen / bathroom / living room / bedroom) is included.

[0200] “Psychological factors” can be calculated by reflecting the suitability of the option configuration with the propensity DNA type (240), style matching results (e.g., modern / minimal matching), sensory sensitivity (lighting / noise / air quality) and lifestyle routine type (home / night type, etc.). For example, if a combination of processes that improves sound insulation / lighting / work flow is included in the home-focused DNA, the psychological factors can be configured to increase.

[0201] “Defect / Anxiety Risk” is a value representing the degree to which anxiety felt by the user increases due to defects (leakage / condensation / cracks, etc.) or maintenance burdens after construction, and can be calculated, for example, by normalizing the risk signals related to DEFECT and MAINTENANCE calculated by the composite risk evaluation unit (510). By using the term (100 - Defect / Anxiety Risk) in the calculation formula, the effect is realized where satisfaction is positively contributed as the risk is lower, and satisfaction is reduced as the risk is higher.

[0202] The above formula is merely an example to illustrate the "existence of a quantitative calculation engine," and depending on the implementation, variations applying weights, element configurations (e.g., separating maintenance risk into a separate term), or non-linear functions (sharp decline above a threshold) are also possible.

[0203] The actual transaction price-based valuation unit (400) includes an actual transaction price data linkage unit (410), a market price difference calculation unit (420), and an HVI / ROI calculation and scenario comparison unit (430).

[0204] The actual transaction price data linkage unit (410) serves as the gateway for "external actual transaction data collection, normalization, and reliability assignment," which is the most fundamental aspect of the present invention. It is configured to generate a preprocessed transaction dataset by bringing in actual transaction price data from the Ministry of Land, Infrastructure and Transport (e.g., apartment transaction data) in a form that can be combined with user input information, so that the subsequent market price difference calculation unit (420) and HVI / ROI calculation and scenario comparison unit (430) can immediately use it for calculation.

[0205] Specifically, the actual transaction price data linkage unit (410) is configured to receive as input basic housing information (110) including regional information and basic information of the target housing collected from the user input collection unit (100), such as address / region, building name (complex name), exclusive area or size, and year of construction (or year of construction), convert it into a call parameter of the Ministry of Land, Infrastructure and Transport's actual transaction price API, and then query the source actual transaction price data corresponding to the parameter from the outside.

[0206] Here, "parameter conversion" refers to the administrative district / legal dong information entered by the user, converted into a legal dong code (LAWD CD Convert to ) and the inquiry period to contract year and month (DEAL YMD It is set in units of ), and may include a process of setting the exclusive area range (similar floor plan filter) as needed. For example, if a user inputs "Gwonseon-gu, Suwon-si, Gyeonggi-do, 32 pyeong (or exclusive area approximately 84㎡), 15th year," the actual transaction price data linkage unit (410) [sets] the LAWD corresponding to that area CD and recent N months' DEAL YMD It can be configured to generate a list (e.g., last 6 months, last 12 months, etc.) and collect actual transaction records for that period in stages.

[0207] The actual transaction price data linkage unit (410) may be configured to call the Ministry of Land, Infrastructure and Transport's actual transaction price API to receive a response in XML format and to parse it to extract transaction records. The transaction records extracted at this time may include items such as at least apartment name (complex name), transaction amount, year of construction, exclusive area, contract date (or contract year / month / day), floor, road name / legal district, etc., and if fields unnecessary for subsequent calculations are included, they may be removed, or conversely, derived variables necessary for subsequent calculations may be added.

[0208] For example, if the transaction amount is provided in the form of a "string (including commas)," it can be normalized into an integer amount (in units of ten thousand won, etc.), the contract date can be converted into a standard date format, and area range labels (e.g., 59㎡ range, 84㎡ range, etc.) can be assigned to determine whether it is a "similar floor plan" based on the exclusive area. Additionally, since the year of construction or the age of construction is directly used for "separating construction / repair groups" in the market price difference calculation unit (420), the actual transaction price data linkage unit (410) can be configured to calculate the age of construction based on the time of transaction (e.g., transaction year - year of construction) and include it in the record, or at least ensure that the year of construction field is secured without omission.

[0209] The actual transaction price data linkage unit (410) can be configured to perform consistency verification between user input and transaction records collected via API in order to increase "inquiry accuracy". For example, the same legal district code (LAWD CD Among the transactions retrieved by ), only records that match the target complex name (110) entered by the user are separated into a primary priority dataset, and if the complex name does not match completely or there is a variation in notation (spacing / abbreviation), the candidates can be expanded by applying similar string matching or a separate complex name normalization rule.

[0210] In addition, if the area / exclusive area (110) entered by the user and the exclusive area of ​​the transaction record are within a certain range (e.g., ±3㎡ or ±5㎡), it can be determined as a "similar area" and filtered to keep only transactions that can be compared within the same area. Through such consistency verification and filtering, the input quality is ensured so that the subsequent market price difference calculation unit (420) can calculate the average price and market gap under the premise of "similar area / similar area."

[0211] The actual transaction price data linkage unit (410) may be configured to calculate and provide "reliability information" based on the number of transactions, period coverage, and error status of the inquiry results for data quality and system stability.

[0212] For example, if the number of transactions collected in a specific region or unit size exceeds a certain threshold (e.g., 10 cases), the confidence level can be classified as "high"; if it is between 5 and 9 cases, as "medium"; and if it is less than 5 cases, as "low," and this can be output as metadata along with the transaction dataset. In addition, in cases where API call failures, response delays, or abnormal responses (XML parsing failures) occur, a retry policy (e.g., identical DEAL YMD It can be configured to secure a minimum number of comparable transactions by applying a fallback that re-requests a certain number of times, automatically extends the inquiry period (e.g., extends from the last 6 months to 12 months), or extends the range to adjacent regions / higher administrative districts.

[0213] In one embodiment for more systematically reflecting variations in the availability and quality of external data, the actual transaction price data linkage unit (410) may apply a three-stage fallback strategy. (Level 1) Use source transaction data collected from an external actual transaction price API first, (Level 2) If the sample at Level 1 is insufficient or repeated call failures occur, refer to internal standard data (e.g., standard transaction distribution by region / area interval or internally constructed DB) to correct the comparison group, and (Level 3) If a sufficient comparison group is not secured even with Level 1 and Level 2, use mock data or conservative estimates, and may be configured to specify the fact of such use in the output.

[0214] Additionally, the actual transaction price data linkage unit (410) may be configured to calculate and provide a data confidence score using input features such as the used fallback level (1-3), number of transactions, period coverage, degree of similar floor plan matching (area difference), whether the area is expanded, and the number of samples remaining after removing outliers. For example, the confidence score may be normalized to a range of 0 to 100, and the HVI / ROI calculation and scenario comparison unit (430) may be linked in a way that conservatively corrects the market gap term or the final value increase or outputs a warning message according to the confidence score.

[0215] However, as the precision of the comparison may decrease as the range is expanded, the actual transaction price data linkage unit (410) displays "whether the expansion is applied" and "reliability decrease due to expansion" as metadata so that it can be transparently notified at the subsequent stage (420, 430) and user output stage.

[0216] In addition, the actual transaction price data linkage unit (410) may be operated including a cache or storage to reduce performance degradation due to repeated inquiries and the burden of external API calls. For example (LAWD CD , DEAL YMD It can be configured to store recent inquiry results for a certain period using a combination of (region, period, area range) or a combination of (region, period, area range) as a key, and to return the stored results without calling the external API again when a request with the same conditions occurs again. Furthermore, even if the HVI / ROI is recalculated because the remodeling options provided by the system are changed (e.g., the total construction cost changes in the estimate calculation / update unit (330)), the actual transaction source data itself may not change significantly under the same region and period conditions, so the cache of the actual transaction price data linkage unit (410) serves as a means to increase the responsiveness of the valuation calculation.

[0217] The market price difference calculation unit (420) is configured to quantitatively calculate the average price difference, i.e., the market price difference (market gap), that is actually formed between "transactions with interior (repair) reflected" and "transactions without repair reflected" in the same area and similar floor plan, and provide this to the HVI / ROI calculation and scenario comparison unit (430) at the end, thereby reflecting as supporting data "how much the local market recognizes remodeling as price" rather than simply the scale of construction costs.

[0218] In other words, the market price difference calculation unit (420) receives the transaction dataset secured and normalized by the actual transaction price data linkage unit (410) as input, creates similar comparison groups, separates groups based on the building age criteria, and calculates the difference in representative prices between groups to derive the market gap.

[0219] Specifically, the market price difference calculation unit (420) determines a comparison target transaction group that satisfies the "similar area and similar floor area" conditions among the transaction records retrieved by the actual transaction price data linkage unit (410), by referring to basic conditions such as regional information (e.g., legal dong, city / county / district), exclusive area or pyeong, and year of construction (or year of construction) of the target residence entered from the residential basic information input unit (110) of the user input collection unit (100). Here, the similar area is the legal dong code (LAWD CD It can be determined based on the same administrative district or predetermined adjacent / extended rules, and similar floor plans can be implemented by keeping only transactions that satisfy a certain allowable range (e.g., ±3㎡ to ±5㎡) or area range (e.g., 84㎡ range) based on the exclusive area.

[0220] In addition, to reduce market distortions at the time of transaction, the contract year and month (DEAL YMDBased on ), an analysis period such as the most recent N months (e.g., 6 months or 12 months) can be set, and it can be configured to use only transactions within that period. The reason for matching "same area, similar size, and similar period" in this way is to control the comparison conditions as much as possible so that the Market Gap calculated by the market price difference calculation unit (420) can be interpreted as a difference attributable to the remodeling effect.

[0221] Subsequently, the market price difference calculation unit (420) may be configured to perform refinement and outlier mitigation processing to ensure the quality of the collected transaction records. For example, it may exclude records with missing transaction amounts, abnormal formats, or zero won, perform duplicate removal if the same transaction is included multiple times, and apply rules to truncate upper and lower extreme values ​​(e.g., trimming upper and lower percent) or exclude data in an abnormal range based on the unit price per area to prevent the average from being distorted by unusual transactions with excessively low or high amounts (quick sale, related party transaction, etc.). This refinement process contributes to increasing the explanatory power of the market gap by reducing the phenomenon where the group average price calculation described later is influenced by "a few transactions."

[0222] When refinement is complete, the market price difference calculation unit (420) classifies transaction data into at least a first group and a second group by applying a predetermined building age standard. In an embodiment of the present invention, based on a building age of 15 years, transactions of 15 years or more may be classified into a "non-repair (construction) group" and transactions of less than 15 years may be classified into a "repair or new construction group."

[0223] In other words, among transactions in the same area and similar floor plan, the structure separates and compares "prices occurring in relatively old (unrepaired) buildings" and "prices formed with new construction or repairs / improvements reflected." This separation criterion (e.g., 15 years) may be stored as a fixed value in the system or operated as a parameter adjustable according to market conditions and regional characteristics. Additionally, if only the year of construction exists in the transaction record, the system may be configured to calculate the age of construction based on the transaction date or the year and month of the transaction to determine whether it exceeds 15 years.

[0224] Next, the market price difference calculation unit (420) calculates the representative price of the transactions belonging to each group. The representative price may be a simple average transaction amount, but a method may also be applied in which the average or median value is calculated by converting it into a unit price per exclusive area (e.g., unit price per square meter or per pyeong) to account for cases where the floor area is not completely identical, and then converting it again to match the target floor area.

[0225] At this time, when the representative price of the first group (non-repair / construction) and the representative price of the second group (repair or new construction) are calculated, the market price difference calculation unit (420) calculates "representative price of the second group - representative price of the first group" as the market price difference (market gap). For example, if the average price of non-repair in the same area and same floor plan is 48 million won and the average price of repair / new construction is 53 million won, the market gap is calculated as 5 million won, which can be interpreted to mean that "the market in that area evaluates (the repair or new construction state) with an average premium of about 5 million won." This market gap is used as a core input for the HVI formula by the HVI / ROI calculation and scenario comparison unit (430) at the end, so that the 'market premium' is reflected in the value increase estimate separately from the construction cost.

[0226] In addition, the market price difference calculation unit (420) may be configured to apply a cap according to the regional tier to the calculated market gap in order to prevent overestimation and reflect differences in regional market structure. That is, even if calculated using the same calculation method, there may be cases where the market gap becomes abnormally large due to transaction fluctuations at a specific point in time or a lack of samples. Therefore, the region of the target residence is classified into Tier 1 / Tier 2 / Tier 3, and a market gap cap is assigned to each Tier so that if it exceeds a certain amount, it is corrected to the cap value.

[0227] For example, this method sets upper limits of 45 million won for Tier 1, 30 million won for Tier 2, and 20 million won for Tier 3, and if the calculated value exceeds these limits, the upper limits are fixed as the final Market Gap. Here, the Tier classification itself is based on the regional weighting (R) in the post-evaluation of value. i It can also be used as a “safety device to prevent the runaway market gap”, but in this configuration (420), it specifically functions as a “safety device to prevent the runaway market gap”.

[0228] The HVI / ROI calculation and scenario comparison unit (430) is configured to calculate a quantitative value of "how much the remodeling option is likely to be recovered as an increase in asset value in the actual market" within the actual transaction price-based valuation unit (400), and to provide the calculated result to the user by comparing it in parallel with multiple conservative / realistic / optimistic scenarios.

[0229] That is, the HVI / ROI calculation and scenario comparison unit (430) functions as a quantitative valuation engine that calculates the expected increase in asset value after construction (e.g., in units of ten thousand won) and the return on investment (ROI, %) by combining the "market premium based on actual transaction data (market price difference, Market Gap)" obtained through the actual transaction data linkage unit (410) and the market price difference calculation unit (420) with the "total construction cost by option" reflecting user tendencies and process selection results, rather than a simple sum of construction costs or recommendations based on intuition.

[0230] Specifically, the HVI / ROI calculation and scenario comparison unit (430) receives the total construction cost by option (final amount including direct construction costs, indirect costs, and VAT selectively reflected) from the estimate calculation / update unit (330) of the remodeling decision calculation unit (300), receives the area information, area / exclusive area, and year of construction of the target residence as input from the residential basic information input unit (110) of the user input collection unit (100), and can refer to the multidimensional lifestyle tendency index and tendency tags (e.g., T01~T15) calculated by the lifestyle tendency index calculation unit (200).

[0231] In addition, the process range determined in the process selection input section (130) and the process classification and option configuration section (320) (number of selected processes, whether core processes are included, whether partial / total remodeling is required, etc.) may be referenced and reflected in the calculation of the process range coefficient described later. Furthermore, the Market Gap calculated by the market price difference calculation section (420), its reliability (high / medium / low based on the number of transactions, etc.), and whether to apply a cap correction by region grade are received as inputs and used as key variables for the market premium term.

[0232] The HVI / ROI calculation and scenario comparison unit (430) is configured to perform calculations in a structure that separates and sums the construction cost contribution and the market gap contribution to calculate the "expected value increase" corresponding to the Home Value Index (HVI) based on the above inputs. In one embodiment, the HVI / ROI calculation and scenario comparison unit (430) may apply a calculation formula of the following form.

[0233] The expected value increase is, (i) "Total construction cost × Regional construction cost reflection rate (R rate The construction cost contribution expressed as ” and, (ii) "Market Gap × Style Match (S) × Age Factor (A) × Regional Weight (R i It can be configured to be calculated by summing the market gap contribution expressed as ) × process range coefficient (P)”. To explain this in sentences, the first term quantifies the possibility that construction costs will be passed on to the price by multiplying by the “average rate at which interior costs are reflected in the market price in that local market,” and the second term corrects the strength of the market premium actually realized by multiplying the “premium (market price difference) received by repair / new construction in the same area and similar floor plan” by “style suitability of the option, sensitivity to remodeling reflection based on age, price sensitivity of the local market, and level of process range (partial / whole)”.

[0234] Here, the regional construction cost reflection rate (R rate ) and local weights (R i ) can be configured to classify the target residential area into multiple predetermined tiers and apply differentially according to each tier. For example, Tier 1 (higher-tier area) is R rate Set the default value to 0.5, Tier 2 (major metropolitan areas) to 0.4, and Tier 3 (other areas) to 0.3, and the regional weight (R iSimilarly, Tier 1 can be set to 1.25, Tier 2 to 1.05, Tier 3 to 0.85, and so on, to reflect differences in price sensitivity by region. This coefficient system is intended to quantitatively reflect the fact that even for the same construction work, the "price reflection rate" accepted by the selling and trading markets varies depending on the region.

[0235] Additionally, the style match (S) is a coefficient that reflects the preference trends or popularity of the local market as well as the tendencies and style matching results (e.g., Modern / Minimal / Nordic, etc.), and can be set to a value in the range of 0.7 to 1.3. In one embodiment, style-specific weights such as Modern (1.15), Minimal (1.20), Nordic (1.10), Natural (1.05), French (1.00), Classic (0.95), Vintage (0.90), Industrial (0.85), and Oriental (0.80) may be predefined, and the value corresponding to the recommended style of the option may be adopted as S. This implements the assumption (market observation) that "even within the same process range, the greater the market preference for a style, the greater the possibility of realizing a price premium" as an internal engine variable.

[0236] The building age coefficient (A) is a coefficient designed to reflect the tendency for the effect of remodeling to increase with the age of the target residence, and can be applied in intervals such as 1.0 for 0 to 5 years, 1.05 for 6 to 10 years, 1.15 for 11 to 15 years, 1.30 for 16 to 25 years, and 1.45 for 26 years or more. In other words, it reflects in the form of a coefficient that even with the same interior cost, the "feeling of improvement and room for market reflection" may be greater the older the building is.

[0237] The process scope coefficient (P) is a coefficient that adjusts the reflection ratio of the Market Gap term according to the comprehensiveness of the process selected by the user (whether it is partial construction, medium-scale including core processes, or full remodeling), for example, it can be set to 1.0 for full remodeling (selection of 6 or more processes), 0.6 to 0.8 for partial construction (3 to 5 processes), and 0.3 to 0.5 for some processes (1 to 2 processes) (430).

[0238] At this time, a modified embodiment is also possible in which P is subdivided by weighting the inclusion of processes that have high price recognition in the market, such as kitchen / bathroom / wallpapering / flooring, in addition to the "number of processes." In that case, the results of the process selection input unit (130) and the process classification and option configuration unit (320) are directly used in the calculation of P. Through this, it is possible to quantify not just how many processes were performed, but whether it is a "process range that causes a market price premium."

[0239] Meanwhile, the HVI / ROI calculation and scenario comparison unit (430) may, in order to control the possibility that the Market Gap value provided by the market price difference calculation unit (420) may be distorted to be over / under due to a lack of transaction samples or distressed sales, (a) first use the Market Gap with a regional grade cap correction applied, and (b) additionally apply a compensation correction coefficient to the calculation of the Market Gap or the final value increase according to the transaction count-based reliability (high / medium / low).

[0240] For example, when confidence is low, the Market Gap term can be configured to mitigate overestimation by discounting it by a certain percentage (e.g., 0.85 or less), and when confidence is high, it can be configured to maintain the original value. This confidence-linking aligns with the design intent of presenting not only the "number of results" but also the "strength of the basis."

[0241] When the expected increase in value is calculated in this way, the HVI / ROI calculation and scenario comparison unit (430) is configured to calculate the return on investment (ROI) as a percentage of "expected increase in value ÷ total construction cost × 100". That is, ROI represents the "possibility of recovering the construction cost invested in the option in what percentage through the increase in asset value" as an intuitive ratio. For example, if the total construction cost is 30 million won and the expected increase in value is 24 million won, the ROI is calculated as 80%.

[0242] In addition, the HVI / ROI calculation and scenario comparison unit (430) can be configured to not only present "numbers (won / %)" to the user, but also calculate the HVI score (0~100 points) and grade (S / A / B / C / D) together to enable intuitive comparison.

[0243] In one embodiment, the HVI score may be calculated as a normalized value by combining ROI, the absolute size of the expected value increase, and the Market Gap reliability, and the grade may be configured to be determined according to predefined criteria, such as Grade S (overall score of 85 points or more and ROI of 100% or more), Grade A (70-84 points and ROI of 80-99%), Grade B (55-69 points and ROI of 60-79%), Grade C (40-54 points and ROI of 40-59%), and Grade D (less than 40 points or ROI of less than 40%). In this case, the grade is based on the premise that it is a "quantitative summary to aid in comparison and judgment" rather than an "investment solicitation," and the report provided to the user may be configured to include a disclaimer (e.g., an estimate based on actual transaction prices, subject to change depending on market conditions, and not an investment solicitation).

[0244] In particular, the key difference of the HVI / ROI calculation and scenario comparison unit (430) is that, rather than calculating only a single value for the same input, it calculates multiple scenarios (conservative / realistic / optimistic) with different coefficient settings in parallel and provides the results so that they can be compared on one screen (or one report).

[0245] In one embodiment, the conservative scenario is the regional construction cost reflection rate (R rate By lowering ) to the minimum value (Tier 1=0.3, Tier 2=0.2, Tier 3=0.15) and setting Style Match (S) to 0.85, assuming a distressed sale / downturn, the realistic scenario is R rate Assuming a standard trading environment with the default values ​​(Tier 1=0.5, Tier 2=0.4, Tier 3=0.3) and S set to 1.0, the optimistic scenario is R rate It can be configured to assume a bull market or positive news by raising to the maximum value (Tier 1=0.7, Tier 2=0.6, Tier 3=0.45) and setting S to 1.15. In this case, "Expected Value Increase," "ROI," "HVI Score," and "Rating" are calculated for each scenario, allowing the user to intuitively understand the range of how results differ depending on market condition assumptions, even for the same option.

[0246] Furthermore, the HVI / ROI calculation and scenario comparison unit (430) can be configured to repeat the above 3-step scenario calculation for each option for multiple remodeling options (e.g., Essential / Standard / Opus or cost-effective / balanced / premium types) so that "comparison between options" and "comparison between scenarios" can be performed simultaneously. That is, the user can simultaneously check (i) which of options A / B / C has a higher ROI within the same scenario, and (ii) how much the ROI fluctuates depending on conservative / realistic / optimistic within the same option, and this is a part that combines with the purpose of being an "engine that supports decision (selection)" (a data-based decision algorithm rather than a recommendation).

[0247] In addition, the HVI / ROI calculation and scenario comparison unit (430) may further include a function to calculate the residual value after a certain period of time, in addition to estimating the increase in value immediately after construction.

[0248] In one embodiment, the selected processes may be classified into infrastructure processes (piping, electrical, insulation, etc.) and aesthetic processes (wallpapering, flooring, lighting, etc.), and the infrastructure processes may be configured to have a depreciation rate of 5% per year and a lifespan of 20 years, and the aesthetic processes may be configured to have a depreciation rate of 20% per year and a lifespan of 5 years, and to calculate the residual value after 5 years and 10 years. The result of this residual value calculation can be used as a basis to quantitatively present the "value that remains over time (value protection)" from the perspective of actual residence as well as from the perspective of short-term sales, and can be configured to be included together in the final report.

[0249] To further specify the asset classification of the above processes, the processing unit may manage each process by assigning an "asset type" field (e.g., Infrastructure / Aesthetic) to the process master or process-attribute table. For instance, Infrastructure processes can be defined as processes of a functional basis nature, such as piping, electrical work, insulation, waterproofing, and equipment installation, while Aesthetic processes can be defined as processes centered on aesthetics and presentation, such as wallpapering, lighting, painting, and finishing materials (including film). Additionally, since some processes (e.g., bathroom tiles / flooring) may involve a mixture of functional elements (waterproofing / slope) and aesthetic elements, the type may be determined according to "main purpose" or "cost composition ratio" depending on the embodiment, or the system may be configured to assign multiple type tags and then break down and reflect costs by type.

[0250] In one embodiment, the residual value can be calculated by depreciating the process-specific costs by type-specific parameters (lifetime, depreciation rate, residual value reflection rate) and summing them. For example, the cost of each process i selected in the option is C i , process type g(i)∈{Infra, Aesthetic}, depreciation rate by type d g , lifespan by type L g , residual value reflection rate by type w gIf set as such, the residual value RV of process i after t years i(t) It can be produced in the following form (example).

[0251] RV i(t) = C i × w {g(i)} × max(0, 1 - d {g(i)} × (t / L {g(i)} ))

[0252] Alternatively, apply an exponential depreciation model to RV i(t) = C i Х w {g(i)} × (1 - d {g(i)}) t It can also be calculated in the form of RV. The final residual value is RV total(t) =Σ RV i(t) It can be calculated as.

[0253] In addition, the ROI can be extended to a recovery rate for period t by considering not only the expected increase in value immediately after construction (e.g., expected increase based on Market Gap) but also the above-mentioned residual value. For example, the extended ROI for period t can be calculated as follows (Example).

[0254] ROI(t) = (Expected Value Increase + RV total(t) ) ÷ Total Construction Cost × 100

[0255] At this time, the processing unit can calculate the total ROI as well as the sum of costs and the sum of residual values ​​for infrastructure and aesthetic processes in parallel, and provide the user with a breakdown of "which types of processes retain value over time." The above calculation formula and parameter values ​​(depreciation rate / lifespan / reflection rate) are examples and may be changed according to region, model year, market conditions, or process characteristics.

[0256] The risk and basis provision unit (500) includes a complex risk evaluation unit (510), a judgment granting unit (520), and a basis generation unit (530).

[0257] The complex risk evaluation unit (510) is a core evaluation engine that calculates quantitative and qualitative risk information that serves as raw material for judgment in three major axes (ASSET / MAINTENANCE / DEFECT) rather than directly "declaring" a judgment such as PASS / WARN / BLOCK (pass / caution / not recommended) (this is performed by the judgment granting unit (520)), and comprehensively calculates the risk level by including "complex risks" that are newly generated due to the combination of multiple processes and materials.

[0258] Specifically, the complex risk assessment unit (510) receives basic information of the target residence (e.g., region, area, year of construction, family composition) (110) and tendency / lifestyle information (e.g., cleaning tendency, storage tendency, daily routine, health factors, inconvenience factors, etc.) (120) as input from the user input collection unit (100), and receives the process range (e.g., kitchen / bathroom / wallpapering / flooring / sash, etc.) actually selected by the user and the selected material grade or option information through the process selection input unit (130).

[0259] Additionally, the complex risk assessment unit (510) is configured to reflect individual differences in the assessment, such as “maintenance risk may be greater for some users even with the same material,” by referring to the multidimensional lifestyle tendency indicators (e.g., cleaning sensitivity, tidiness tendency, budget sensitivity, sensory sensitivity, etc.) calculated by the lifestyle tendency indicator calculation unit (200).

[0260] Furthermore, the process classification and option configuration unit (320) and the estimate calculation / update unit (330) of the remodeling decision-making unit (300) receive the process configuration by option, process priority, and estimate by option (total construction cost, amount by process) as input, and can evaluate how risk is combined with cost, scope, and priority. In one embodiment, the composite risk evaluation unit (510) may refer to the results of the HVI / ROI calculation and scenario comparison unit (430) of the actual transaction price-based valuation unit (400) (e.g., option with low ROI, Market Gap reliability, etc.) to more precisely correct the "possibility of loss from an asset value perspective."

[0261] The composite risk assessment unit (510) is configured to calculate risks in at least three categories, including asset value risk (ASSET), maintenance risk (MAINTENANCE), and defect risk (DEFECT), based on the above inputs. Here, "asset value risk (ASSET)" refers to the possibility that the selected process, material, or style will act as a depreciation factor in future sales or market valuations. The risk is calculated by using input features such as non-standard finishes with low market preference, style bias based on excessive personal taste, the possibility of over-investment (so-called over-improvement) relative to surrounding market prices, and the omission of key processes preferred in the trading market (e.g., kitchen / bathroom / basic facilities).

[0262] At this time, ASSET risk can be configured to vary not simply based on "using high-quality materials lowers risk," but also based on the region, age, and size of the target residence, the scope of construction, and the alignment with actual transaction-based market response (market price difference, market gap). For example, if the total construction cost has increased significantly but the realization range of the interior premium (market gap) in the relevant region and similar size is limited or the reliability is low, the composite risk assessment unit (510) can reflect "reduced recoverability" as a factor increasing asset value risk, and this result is used as a basis data for the subsequent judgment assignment unit (520) to make a judgment in the direction of WARN or BLOCK.

[0263] “Maintenance risk” refers to how often and how difficult cleaning, maintenance, repair, and replacement occur during the post-construction residential phase, and is evaluated based on whether there is a conflict between the user’s lifestyle tendency indicator (200) and the maintenance characteristics of the materials / process (200, 510). In one embodiment, the composite risk evaluation unit (510) assigns weights to items where the maintenance burden increases according to inputs such as cleaning tendency or cleaning sensitivity (T04, etc.), tidiness tendency (T05, etc.), family composition (infants / pets / elderly parents, etc.) (110), and daily routine (weekend type / night type, etc.) (120).

[0264] For example, if a user who has high cleaning sensitivity (requires frequent cleanliness) or has low actual cleaning frequency and capacity selects a material that is sensitive to stains, moisture, and scratches or requires regular coating / sealing, the composite risk assessment unit (510) increases the maintenance risk score by reflecting "increased difficulty of maintenance" (510). Conversely, if a maintenance-friendly material (e.g., a finish that has high stain resistance and requires less separate coating) is selected, or if the user's disposition is a management type (high tidiness score, stable routine), the maintenance risk can be configured to be calculated as low even within the same process range.

[0265] “DEFECT” refers to the possibility of defects, such as leakage, cracks, peeling, condensation, mold, and electrical / plumbing troubles, occurring during the construction process or within a certain period after construction. It is evaluated by comprehensively considering the age of the subject residence, the deterioration of existing facilities, the inclusion or omission of selected processes, the sequence dependency between processes, and the physical and environmental suitability of specific materials.

[0266] For example, in a residential building (110) that has been built for a certain period of time or longer, if the bathroom process is selected but basic processes such as plumbing and waterproofing are omitted (130), or if the risk of leakage is estimated to be high based on existing condition information, the composite risk assessment unit (510) can be configured to significantly increase the DEFECT risk and to generate a risk signal including a time axis, such as "increased possibility of leakage within 5 years."

[0267] In addition, since the risk of defects can increase rapidly not only by the presence of a single process but also by the "combination of processes," the complex risk evaluation unit (510) can evaluate complex defect factors that appear when multiple processes and materials are combined (e.g., combination of finishing in wet spaces, possibility of cracking due to differences in thermal expansion coefficients, risk when electrical / lighting changes and structure drilling / demolition are combined, etc.) as a separate combination rule. Furthermore, according to the embodiment, if the estimation result of a metal structure (pipe / wire, etc.) within the wall calculated by the hardware sensor linkage unit is provided, the process combination in which drilling / demolition is planned in the corresponding risk zone can be reflected as a risk increase factor, and a risk feature of "work avoidance or additional safety measures required" can be generated.

[0268] The term "complex" in the complex risk assessment unit (510) implies that the final risk vector is calculated by reflecting the interaction between the axes and the combined effects between the processes, rather than simply listing the three axes above. That is, even with the same choice, the risk profile can differ, such as (i) a combination where the asset value risk (ASSET) is low but the maintenance risk (MAINTENANCE) is high (e.g., materials that are highly preferred by the market but difficult to manage), (ii) a combination where the maintenance risk is low but the defect risk (DEFECT) is high (e.g., bathroom processes where the exterior finishing is easy but old plumbing is left unattended), and (iii) a combination where all three axes rise simultaneously (e.g., non-standard materials + difficulty in maintenance + omission of foundation processes in old housing). The complex risk assessment unit (510) separates and calculates this in the form of "scores / grades by axis" and outputs a structured result so that the judgment assignment unit (520) can immediately use it for judgment.

[0269] In one embodiment, the composite risk assessment unit (510) can calculate the risk for each category by accumulating the “increase / decrease due to selected processes / materials” and the “increase / decrease due to lifestyle tendency indicators” to the “standard risk level (derived from the age, family composition, spatial characteristics, etc. of the target residence),” and the calculated score can be provided normalized to a range of 0 to 100.

[0270] At this time, the complex risk assessment unit (510) may be configured not only to provide a simple score, but also to output the rules and features that contributed to the score (e.g., "construction + bathroom + plumbing not replaced", "high cleaning sensitivity + materials vulnerable to contamination", "non-standard finish + regional preference mismatch") together as a "risk basis token" or "risk code," and this output is directly used by the subsequent basis generation unit (530) to construct a natural language basis (descriptive sentence) and a basis chain. Additionally, the complex risk assessment unit (510) may produce mitigation measures to reduce the risk (e.g., recommending the addition of specific processes, material substitution, changing the order of processes, avoiding risk zones) together in the form of a proposal, and this can also be utilized by the judgment granting unit (520) to design conditions that allow switching from WARN to PASS (conditional pass), etc.

[0271] The judgment assignment unit (520) is composed of a judgment layer that receives as input a risk level or risk signal for each of the three major risk categories (asset value risk (ASSET), maintenance risk (MAINTENANCE), and defect risk (DEFECT)) calculated by the complex risk evaluation unit (510), assigns at least one judgment among recommendation (PASS), caution (WARN), and non-recommendation (BLOCK) according to a predefined decision rule and threshold system, and structures the judgment into option, process, and material units and outputs it. That is, if the complex risk evaluation unit (510) calculates "how much risk is involved" as a multidimensional vector, the judgment assignment unit (520) compresses and aligns that vector into a three-stage decision signal of "so whether it can be selected now" so that it can be used immediately for user interface and subsequent basis generation.

[0272] Specifically, the judgment assignment unit (520) may be configured to adjust the sensitivity of the judgment according to user tendencies and household conditions even in the same risk situation by referring to basic information of the target residence (e.g., area, region, year of construction, family composition) (110), tendency / lifestyle information (120), and process and / or material selection information (130) collected from the user input collection unit (100), and also referring to multidimensional lifestyle tendency indicators (e.g., cleaning tendency, tidiness tendency, budget tendency, sensory sensitivity, etc.) (200, 230) calculated and corrected by the lifestyle tendency indicator calculation unit (200).

[0273] Additionally, the judgment assignment unit (520) can more precisely reflect judgment conditions such as "omission of essential processes" or "risk exposure relative to budget" by referring to the essential / recommended / optional process classification results calculated by the process classification and option configuration unit (320) of the remodeling decision calculation unit (300), and the construction cost per option and amount per process calculated by the estimate calculation / update unit (330). Furthermore, by referring to valuation results such as ROI, HVI score, and Market Gap reliability calculated by the HVI / ROI calculation and scenario comparison unit (430) of the actual transaction price-based valuation unit (400), it may be configured to reflect "reduced recoverability" in the asset value risk (ASSET) judgment in a WARN or BLOCK direction.

[0274] The judgment target of the judgment assignment unit (520) is not limited to one, and in one embodiment, it may be set to at least one of (i) the entire remodeling option (e.g., cost-effective / balanced / premium type), (ii) the process unit (e.g., kitchen, bathroom, wallpapering, window frames, etc.), and (iii) the material or option element unit (e.g., countertop material, tile type, lighting method, etc.). Accordingly, the judgment assignment unit (520) can simultaneously produce a "comprehensive option judgment" and a "detailed judgment by process / material" for the same option, and the user can quickly obtain a conclusion through the comprehensive judgment while simultaneously tracking the location of the risk (which process / material is the problem) through the detailed judgment.

[0275] In an embodiment of the present invention, the judgment granting unit (520) may not only display a judgment such as PASS / WARN / BLOCK to the user, but may also generate and output constraint information so that the judgment is substantially reflected in the process configuration. For example, the judgment granting unit (520) may generate a constraint including at least one of (i) mandatory inclusion of a specific process (Require), (ii) selective blocking of a specific process or combination (Block), (iii) substitution with an alternative process or alternative material (Replace), and / or (iv) conditional pass (relaxed to PASS when specific additional measures are included).

[0276] The above constraints can be transmitted to the process classification and option configuration unit (320) to be used for reconfiguring the process configuration and parameters of options A / B / C, and can be transmitted to the process selection input unit (130) to be reflected in the user terminal UI by disabling (blocking) specific options or maintaining the required process in an automatic selection state (including forced selection). Additionally, when a substitution is applied, the estimate calculation / update unit (330) can be configured to automatically update the amount based on the unit price and man-hours of the substituted process / material.

[0277] The judgment assignment unit (520) may be configured to first assign a “category judgment” to each of the asset value risk (ASSET), maintenance risk (MAINTENANCE), and defect occurrence risk (DEFECT) by breaking down the risk assessment results input from the complex risk assessment unit (510) by category.

[0278] For example, a category judgment can be calculated by assigning a WARN if the ASSET risk score or signal is above a predetermined first threshold, a BLOCK if it is above a second threshold, and a PASS if it is below the threshold. Maintenance risk and defect risk can also be judged as one of PASS / WARN / BLOCK (pass / caution / non-recommend) by applying a threshold or rule set suitable for the characteristics of each category. At this time, the judgment assignment unit (520) is configured not to stop at simple score comparison, but to apply "hard rules (mandatory non-recommendation rules)" and "soft rules (warning / conditional pass rules)" in parallel so that conditions with high criticality, such as safety and defects, are processed as BLOCK (non-recommendation) regardless of the score, and conditions with large individual differences, such as management difficulty, are processed as WARN (caution) or conditional PASS (conditional pass).

[0279] In one embodiment, the judgment granting unit (520) may include a mandatory non-recommendation rule that immediately grants a non-recommendation (BLOCK) when specific combination conditions are met. For example, if the construction age of the target residence is above a certain standard (e.g., older) (110), and while the bathroom process is selected, basic processes such as plumbing replacement and waterproofing are not selected (130), or if safety-related processes classified as mandatory in the process classification result are omitted (320), it may be configured to grant a BLOCK judgment to the bathroom process or the entire option, as it is considered that the risk of defect occurrence (DEFECT) increases rapidly.

[0280] In addition, if the user selects a drilling or demolition-related process (130) and a hazardous area (estimated location of pipes / wires / rebar) is detected by the hardware sensor linkage or the result of drawing reference (the referenced configuration may vary depending on the embodiment), the process including the work may be configured to give a BLOCK (not recommended) or at least WARN (caution) to the process to force avoidance of the hazardous area or expert verification.

[0281] On the other hand, the judgment assigning unit (520) may apply a conditional rule that leads to a "WARN" based on the user's lifestyle tendency indicator (200), even for the same process and material. For example, if a specific countertop material is selected in a kitchen process, and the material has characteristics that make maintenance difficult, and the user's cleaning tendency or cleaning sensitivity (e.g., T04) is lower than a predetermined standard and is presumed to be "insufficient maintenance capacity" (200), the judgment assigning unit (520) may assign a WARN judgment based on the MAINTENANCE risk.

[0282] At this time, the judgment granting unit (520) may be configured to output judgment transition conditions such as "replacement recommendation (e.g., replace with easy-to-maintain materials)" or "judgment relaxation upon additional measures (e.g., WARN -> (pass) possible if coating / sealing plan is included)" so that the subsequent basis generation unit (530) can create an explainable recommendation, rather than simply "do not do." In other words, WARN functions as a judgment stage that maintains the user's right to choose while providing a minimum guardrail to recognize and manage risks.

[0283] In addition, the judgment assignment unit (520) can calculate a "comprehensive judgment" by synthesizing the judgments by category (520). In one embodiment, the comprehensive judgment may be calculated by prioritizing the strictest judgment among the category judgments. For example, if any one of the asset value risk (ASSET), maintenance risk (MAINTENANCE), and defect risk (DEFECT) is BLOCK (not recommended), the comprehensive judgment may be BLOCK (not recommended); if there is no BLOCK (not recommended) and at least one is WARN (caution), the comprehensive judgment may be WARN (caution); and if all are PASS (pass), the comprehensive judgment may be PASS (pass).

[0284] This comprehensive method has the advantage of allowing users to immediately identify safety from option cards (e.g., cost-effective / balanced / premium), while also being able to break down and show "which axis the warning came from (asset / management / defect)" through detailed category judgment.

[0285] Furthermore, the judgment assignment unit (520) may be configured to generate tracking information for explainable AI (XAI). Specifically, the judgment assignment unit (520) may store or output "judgment metadata" comprising (i) identification information of the applied decision rule, (ii) input features used in the judgment (e.g., building age, whether a selected process is omitted, specific value range of a lifestyle tendency indicator), and (iii) risk level by category and a comprehensive judgment decision path (e.g., "DEFECT-BLOCK (defect occurrence - not recommended) priority rule triggered").

[0286] The basis generation unit (530) receives as input the comprehensive judgment and process / material unit judgment (e.g., WARN of the entire option, WARN of a specific material, BLOCK of a specific process, etc.) calculated by the judgment assignment unit (520), and also receives the evaluation result of the composite risk evaluation unit (510) regarding which risk category (asset value risk (ASSET), maintenance risk (MAINTENANCE), defect occurrence risk (DEFECT)) the judgment was caused by.

[0287] Additionally, the basis generation unit (530) refers to the basic housing information (110), tendency / lifestyle information (120), and process selection input (130) collected from the user input collection unit (100), and refers to the contribution information of “how much a particular response to a particular question contributed to a particular tendency indicator” held by the question-indicator mapping unit (210) of the lifestyle tendency indicator calculation unit (200), and refers to the indicator change path linked and corrected by the indicator cross-influence reflection unit (220) and the latest indicator value re-corrected according to the process selection by the tendency indicator re-correction unit (230), so that the explanation generated by the basis generation unit (530) is synchronized and updated with the “process / material combination currently selected by the user.”

[0288] For example, the result of the process classification and option configuration unit (320) classifying a specific process as “essential / recommended / optional” (320) can be used as a key basis in the basis generation unit (530) to explain “why it is essential (from the perspective of family composition / age / inconvenience / tendency)” or “why it is recommended (from the perspective of risk mitigation or satisfaction increase),” and the cost allocation by process (e.g., kitchen 25%, bathroom 20%, etc.) and the total of detailed items calculated by the estimate calculation / update unit (330) can be used as a basis to explain “where the cost is spent and why that item was reinforced / restricted by your tendencies.”

[0289] In addition, the Market Gap, HVI, ROI, and comparison results of conservative / realistic / optimistic scenarios calculated by the market price difference calculation unit (420) and the HVI / ROI calculation and scenario comparison unit (430) of the actual transaction price-based valuation unit (400) can be combined as a basis for explaining "how the construction project affects the asset value" on a numerical basis.

[0290] The basis generation unit (530) may be configured to provide, in order to increase the "convincingness" of the explanation, not just a simple summary sentence, but a "reason chain" in which the connection relationships serving as the basis for the judgment can be traced step by step. This reason chain may be stored and configured in a chain structure that follows, for example, (a) question response (answer selected by the user) (100) -> (b) propensity indicator contribution by question-indicator mapping (210) -> (c) indicator correction by cross-influence (220) -> (d) spatial priority score and process classification result (310, 320) -> (e) process / material recommendation and cost / estimation reflection (330) -> (f) risk assessment result (510) -> (g) ASS / WARN / BLOCK judgment (520).

[0291] The basis generation unit (530) can select the path with the greatest impact on the user in this chain structure (e.g., top contributing indicator, top risk factor, top cost item) and explain it in natural language, and if necessary, provide output data so that the user can review steps (a) to (c), i.e., the question-and-answer stage, in the form of a "Selection Reason Replay (Explanation Replay)".

[0292] The basis generation unit (530) can output the characteristics of the basis by separating and combining them into "physical basis" and "economic basis." The physical basis mainly corresponds to maintenance risk and defect risk. For example, if a specific material has characteristics that make maintenance difficult, such as stain, coating, or mold management, it can generate a WARN basis stating "This choice may increase the burden of management" by combining it with the user's cleaning tendency or management capacity. It can also generate a BLOCK or WARN basis stating "The risk of leakage / defect increases after construction" by combining it with the construction year, whether pipes have been replaced, and whether basic processes such as waterproofing or electrical work have been omitted.

[0293] The economic basis primarily corresponds to asset value risk (ASSET) and valuation results (HVI / ROI), and, for example, reflecting that even with the same construction cost, the recoverability varies depending on regional weighting, market gap, style match, building age coefficient, and process range coefficient, it can generate a sentence explaining “what the ROI of the option is in which scenario and what inputs formed that figure.” Additionally, the basis generation unit (530) can be configured to output a notice-like phrase stating “provision of judgment data, not investment solicitation,” which is intended to transparently convey to the user that the valuation result is an estimate based on actual transaction price data rather than a confirmed prediction.

[0294] Additionally, the basis generation unit (530) may be configured to display the fallback level and data confidence score calculated by the actual transaction price data linkage unit (410) together or include them in the basis chain when providing the valuation results. Through this, the user can verify which data level (external source / internal standard / simulated data) and sample characteristics the ROI / HVI figures were calculated based on, thereby improving the credibility and interpretability of the report.

[0295] The basis generation unit (530) can provide multiple output formats for the same basis to help the user understand. For example, on the option card screen, summarized basis such as "one-line summary" is provided, and on the detail screen, the basis can be broken down and described as "family-tailored analysis," "investment value diagnosis," "cautionary / risk warning," "alternative plan or additional process suggestion (masterstroke)," etc.

[0296] At this time, "family-tailored analysis" can be generated by combining the family composition and area / year information (110) of the user input collection unit (100) and the tendency indicator of the lifestyle tendency indicator calculation unit (200) to provide grounds explaining why the option is suitable or unsuitable for a specific family type (e.g., newlyweds / dual-income / infants, etc.), and "investment value diagnosis" can be generated by combining the ROI, HVI, and scenario comparison results (430) of the actual transaction price-based valuation unit (400) to provide comparative sentences such as "recovery rate based on realistic scenarios," "defense logic in conservative scenarios," and "upside in optimistic scenarios."

[0297] Additionally, "cautionary notes / risk warnings" can be generated in the form of recommendations that include measures to reduce risk through actual actions (e.g., adding missing basic processes, selecting alternative materials, changing the process sequence, etc.) by linking with the category-specific risks (ASSET / MAINTENANCE / DEFECT) of the composite risk assessment unit (510) and the WARN / BLOCK judgment of the judgment assignment unit (520).

[0298] The basis generation unit (530) can be configured to enhance the user's decision support function, particularly by explaining the "judgment" and "alternative" together. That is, when the judgment assignment unit (520) assigns a WARN or BLOCK, the basis generation unit (530) can generate a "conditional basis" that includes (i) why the judgment was made (cause: tendency indicator, housing conditions, process omission, material characteristics, etc.) (100, 200, 130), (ii) which risk category the problem is due to (ASSET / MAINTENANCE / DEFECT) (510), and (iii) what conditions are to mitigate the judgment (e.g., WARN -> PASS possible when additional processes are included, MAINTENANCE risk reduced when changing to alternative materials, etc.).

[0299] Through this, users can not only receive simple warnings but also verify which changes lead to which results (judgment / ROI / satisfaction), and this serves as an implementation of the technical concept of the "data-driven decision-making algorithm" of the present invention in terms of user experience.

[0300] Additionally, the basis generation unit (530) may be configured to store input values ​​used in generating the basis (propensity indicator values, selected processes, years, regions, etc.) (100, 200), applied rules or judgment paths (520), referenced valuation scenarios and key figures (430), scores or grades by risk category (510), etc., in the form of a “reason log” or at least output them together as identifiable metadata in order to ensure the reproducibility and traceability of the generated basis. This configuration enables the “why was said that way at that time” to be reconstructed in the same way when system improvements, responding to user inquiries, or requests for explanations regarding recommendation results occur in the future, and substantially supports the implementation of explainable AI (XAI).

[0301] The scope of protection of the present invention is not limited to the description and expression of the embodiments explicitly described above. Furthermore, it is added once again that the scope of protection of the present invention cannot be limited by obvious changes or substitutions in the technical field to which the present invention belongs. Explanation of the symbols

[0302] 100: User Input Collection Section 110: Basic Housing Information Input Section 120: Tendency / Lifestyle Information Input Section 130: Process Selection Input Section 200: Lifestyle Tendency Indicator Calculation Unit 210: Question-Indicator Mapping Calculation Unit 220: Indicator Cross-Impact Reflection Section 230: Propensity Indicator Re-correction Section 240: Propensity DNA Type Determinant 300: Remodeling Decision Output 310: Space Priority Calculation Unit 320: Process Classification and Option Configuration Unit 330: Estimate Calculation / Update Section 400: Actual Transaction Price-Based Valuation Department 410: Actual transaction price data linkage unit 420: Market price difference calculation unit 430: HVI / ROI Calculation and Scenario Comparison Section 500: Risk and Evidence Provision Department 510: Complex Risk Assessment Department 520: Judgment Assignment Department 530: Basis Generation Section

Claims

Claim 1 A residential remodeling decision support system comprising: an input unit that collects basic information, tendencies and lifestyle information, and process and / or material selection information of a target residence from a user; and a processing unit that calculates a plurality of lifestyle tendency indicators based on the input information collected from the input unit, and calculates the process configuration and / or budget allocation of at least one remodeling option according to the plurality of lifestyle tendency indicators to calculate an estimate for the remodeling option, wherein the processing unit refers to cross-influence information indicating correlation or cross-influence between the lifestyle tendency indicators and corrects the plurality of lifestyle tendency indicators so that other lifestyle tendency indicators are corrected in conjunction with changes in a specific lifestyle tendency indicator. Claim 2 A system according to claim 1, wherein the processing unit selects some of a plurality of tendency analysis questions based on basic information of the target residence and presents them to the user, and collects the user's response to the selected questions to calculate the plurality of lifestyle tendency indicators. Claim 3 A system according to claim 1, wherein the plurality of lifestyle tendency indicators are configured to be calculated based on a many-to-many mapping rule in which one question response contributes to a weighted contribution to the plurality of lifestyle tendency indicators or multiple question responses contribute cumulatively to a single lifestyle tendency indicator. Claim 4 delete Claim 5 In claim 1, the processing unit calculates (i) a survey-based lifestyle tendency indicator vector calculated based on survey responses included in the tendency and lifestyle information, and (ii) an action-based lifestyle tendency indicator vector calculated based on the history of actions regarding the selection, deselection, grade change, and / or option confirmation of the user's process and / or material selection information, respectively; calculates a final lifestyle tendency indicator vector by weightedly combining the survey-based lifestyle tendency indicator vector and the action-based lifestyle tendency indicator vector, while dynamically adjusting the weights according to the degree of accumulation of the action history or the recency of the survey responses; updates the action-based lifestyle tendency indicator vector and the final lifestyle tendency indicator vector in response to changes in the process and / or material selection information, and is configured so that the process configuration and / or budget allocation and quotation of the remodeling option are updated in conjunction with the updated final lifestyle tendency indicator vector; and is configured to receive additional changes in process and / or material selection information performed by the user based on the updated remodeling option and / or quotation to update the action-based lifestyle tendency indicator vector, thereby, according to the process and / or material selection A system configured to implement a bidirectional feedback loop in which propensity recalibration, option reconfiguration, and estimate updating are repeatedly performed. Claim 6 A system according to claim 1, wherein the processing unit calculates a space priority score for each of a plurality of spaces including at least some of a kitchen, bathroom, living room, and bedroom by reflecting addition / subtraction factors based on a lifestyle tendency index and tendency / lifestyle information starting from a reference value, and is configured such that the process configuration and / or budget allocation are calculated according to the space priority score. Claim 7 A system according to claim 6, wherein the processing unit applies a classification rule to classify a process into at least one of mandatory, recommended, or optional based on at least one of the space priority score and the construction year or family composition of the target residence, and the classification result is configured to be reflected in the process configuration of the remodeling option. Claim 8 A system according to claim 1, wherein the processing unit generates and provides multiple remodeling options in which at least one of different budget levels, material grades, or process ranges differs, and is configured to provide process configurations and estimates for each option together so that they can be compared. Claim 9 A system according to claim 1, wherein the processing unit calculates basic costs based on area and region, applies a distribution ratio for each process, and calculates an estimate by applying a weighting factor for each process based on a lifestyle tendency index, and is configured such that material costs are reflected by referring to cost or unit price data corresponding to the selected process and / or material. Claim 10 A system according to claim 1, wherein the processing unit calculates a similarity for a plurality of predefined interior style candidates based on the plurality of lifestyle tendency indicators to match at least one style, classifies lifestyle routine types based on lifestyle patterns included in the tendency and lifestyle information, and calculates and provides a scenario suitability score for each remodeling option based on the matched style and / or lifestyle routine type. Claim 11 A system according to claim 1, wherein the system estimates the possibility of the existence of a metal structure inside a wall using a magnetic field sensor or a direction sensor of a user terminal, applies a fallback strategy to estimate the location of the metal structure by referring to a drawing database when the sensor reliability is below a predetermined threshold, and is configured such that the estimation result is reflected in the selection of a process or process configuration related to drilling or demolition. Claim 12 A system according to claim 1, wherein the processing unit refers to process-image mapping information corresponding to process and / or material selection information to generate and provide pre-construction and post-construction images so that change elements corresponding to the selected process are reflected, and the post-construction image is configured so that change elements corresponding to the unselected process are not reflected. Claim 13 A system configured such that, in claim 1, the processing unit calculates a market price difference related to a target residence based on actual transaction price data, classifies transaction data of similar areas and similar floor plans into at least a first group and a second group according to a predetermined building age standard, calculates the market price difference based on the difference in average transaction prices of the first group and the second group, and calculates and provides a value index including an expected increase in asset value after construction and an investment recovery rate based on the market price difference and the estimate. Claim 14 A system configured such that, in calculating the value indicators according to Clause 13, the processing unit applies at least one of a regional grade or regional weight, a building age coefficient, a process range coefficient, and a style match coefficient, and calculates and provides for comparison a plurality of scenario-specific value indicators according to different coefficient settings. Claim 15 A system according to claim 1, wherein the processing unit evaluates a plurality of risks including at least one of asset value risk, maintenance risk, and defect occurrence risk based on the combination of processes and / or materials and the conditions of the target dwelling, assigns a judgment of one of recommendation (PASS), caution (WARN), or non-recommendation (BLOCK) according to the risk evaluation result, and generates and provides the basis for the judgment in natural language including the correlation with lifestyle tendency indicators and user input information, wherein the basis is configured to provide the user with a basis chain including the connection relationship between the question answer, lifestyle tendency indicator, and judgment result. Claim 16 A system configured such that, in the event that a partial change in process and / or material selection information occurs, the processing unit identifies process or estimation parameters linked by a predefined dependency relationship with the changed process and / or material, and selectively recalculates only for the identified process or estimation parameters to partially update the estimation and / or remodeling options. Claim 17 A system according to claim 15, wherein the processing unit, in conjunction with the above determination, generates at least one of (i) a first constraint that forces the inclusion of a specific process, (ii) a second constraint that blocks the selection of a specific process or a combination of specific processes, and (iii) a third constraint that replaces a specific process or material with an alternative process or material, and based on the above constraint, reconfigures the process configuration of the remodeling option or controls selectable items on the user terminal, and is configured such that the estimate is updated in conjunction with the reconfigured process configuration.

Images