Information processing method, information processing device, and computer program

Abstract

This invention provides an information processing method, an information processing device, and a computer program that can suggest suitable construction companies to users considering building a custom-built home. [Solution] The information processing method involves obtaining survey information about housing entered by the user, and having a computer execute a process that uses a language model that references the obtained survey information and characteristic information about the characteristics of multiple construction companies stored in a memory unit to output a predetermined number of recommended information for each construction company.

Description

Technical Field

[0001] The present invention relates to an information processing method, an information processing apparatus, and a computer program.

Background Art

[0002] When considering an ordered house, a user has various desires or conditions. Generally, price, design, floor plan, performance, location, after-sales service, etc. are the main items to be considered. By organizing these desires or conditions and setting priorities, it becomes possible to build a more satisfactory ordered house.

[0003] Patent Document 1 discloses a my-home consultant system including means for inputting various conditions corresponding to various needs desired by a user and latest technical information, means for storing the input information, means for editing the input information, and means for searching these information.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, in the ordered house industry, due to an excessive amount of information on construction companies, users cannot select a construction company suitable for themselves and spend a great deal of time and effort on comparative study of construction companies. In addition, there is a risk of selecting a construction company that does not match the user's values or desired lifestyle, or users lacking knowledge necessary for house construction have an anxiety that they cannot select an appropriate construction company.

[0006] This invention has been made in view of the above circumstances, and aims to provide an information processing method, an information processing device, and a computer program that can suggest an appropriate construction company to a user considering building a custom-built home. [Means for solving the problem]

[0007] This invention includes several means for solving the above-mentioned problems, but to give one example, the information processing method involves having a computer perform a process that acquires questionnaire information about housing entered by the user, and uses a language model that references the acquired questionnaire information and characteristic information about the characteristics of multiple construction companies stored in a memory unit to output a predetermined number of recommended information for each construction company. [Effects of the Invention]

[0008] According to the present invention, it is possible to suggest a suitable construction company to a user who is considering building a custom-built home. [Brief explanation of the drawing]

[0009] [Figure 1] This figure shows an example of the configuration of the information processing device according to this embodiment. [Figure 2] This figure shows an example of survey information. [Figure 3] This figure shows an example of characteristic information for a construction company. [Figure 4] This figure shows an example of a method for generating proposals and reports using an information processing device. [Figure 5] This figure shows an example of prompts for generating user-facing suggestions. [Figure 6] This figure shows an example of a systematization logic for "classification axes." [Figure 7] This figure shows an example of the scoring logic for "housing construction intention." [Figure 8] This figure shows an example of a proposed policy based on a residential construction intention score. [Figure 9] This figure shows an example of the logic used to calculate a compatibility score between the homeowner and the house builder. [Figure 10] This is a diagram showing an example of a basic proposal report (proposal for users). [Figure 11] This is a diagram showing an example of a thank-you email and proposal notification email (proposal for users) addressed to the user. [Figure 12] This is a diagram showing an example of a prompt for generating a personal matching report. [Figure 13] This is a diagram showing an example of a personal matching report (detailed version). [Figure 14] This is a diagram showing an example of a matching dashboard. [Figure 15] This is a diagram showing an example of a prompt for generating a customer insight report. [Figure 16] This is a diagram showing an example of a customer insight report. [Figure 17] This is a diagram showing an example of a prompt for generating a general market insight report. [Figure 18] This is a diagram showing an example of a prompt for generating a personalized market insight report. [Figure 19] This is a diagram showing an example of the procedure for report creation processing by an information processing device.

Embodiments for Carrying Out the Invention

[0010] Hereinafter, embodiments of the present invention will be described based on the drawings. FIG. 1 is a diagram showing an example of the configuration of an information processing device 50 according to this embodiment. The information processing device 50 can be connected to a large language model (LLM: Large Language Models) 100 via a communication network 1. The information processing device 50 can access a questionnaire information DB 61, a construction company DB 62, a user-oriented proposal DB 63, a personal matching report DB 64, and a customer insight report DB 65. The information processing device 50 can be configured by a computer or the like.

[0011] The large-scale language model 100 is a machine learning model that utilizes natural language processing technology to learn language patterns from existing extensive text data and can generate or understand text such as articles and conversations. The large-scale language model 100, for example, divides text data into small chunks, and by performing embedding processing for each divided chunk, generates a vector representation (embedding) of each chunk. Note that the embedding processing includes those obtained by quantifying the types of chunks and those obtained by quantifying the positional relationships of the chunks. As a result, each chunk is represented as a point in a high-dimensional space, and in the high-dimensional space, chunks with similar meanings are arranged close to each other, and chunks with different meanings are arranged far apart. The large-scale language model 100 can numerically capture the semantic relevance and similarity between chunks and can understand the semantics, that is, the meaning, of text data. Also, the large-scale language model 100 has an attention mechanism and can obtain a numerical representation that expresses the complex structure of an article by appropriately adjusting the magnitudes of the linear combination coefficients when linearly combining the numerical representations of each chunk calculated by the embedding processing.

[0012] The large-scale language model 100 is a model that has completed pre-training. For example, Gemini can be used. Specifically, it adopts an approach such as in-context learning that interprets and integrates in real time the input (prompt) from the user and the information of the construction company DB62 held as internal knowledge to generate an optimal answer. That is, the large-scale language model 100 treats the information of the construction company DB62 as a "textbook to be referred to on the spot" and executes an inference process of generating an optimal answer for the user's questionnaire information. Note that the large-scale language model 100 is not limited to Gemini, and models composed of deep neural networks such as GPT-4, GPT-3.5, BERT, LaMDA, PaLM, and LLaMA may also be used.

[0013] The information processing device 50 includes a control unit 51 that controls the entire device, a communication unit 52, a memory 53, a display unit 54, an operation unit 55, an interface unit 56, and a storage unit 57.

[0014] The control unit 51 may be configured by incorporating a required number of CPUs (Central Processing Units), MPUs (Micro-Processing Units), GPUs (Graphics Processing Units), etc. Alternatively, the control unit 51 may be configured by combining DSPs (Digital Signal Processors), FPGAs (Field-Programmable Gate Arrays), etc.

[0015] The communication unit 52 is equipped with a communication module and can communicate with the large-scale language model 100 and external devices.

[0016] The display unit 54 can be made of a liquid crystal panel or an organic EL (Electro-Luminescence) display, etc. An external display device may be provided instead of the display unit 54.

[0017] The operation unit 55 is composed of a touch panel or the like, and allows the user to operate icons displayed on the display unit 54, move and manipulate the cursor, input text, etc. The operation unit 55 may also be a mouse or keyboard. An external terminal device (for example, a computer) may be provided instead of the display unit 54 and the operation unit 55. In this case, the information processing device 50 may be composed of a server or the like.

[0018] The interface unit 56 provides interface functions for accessing the survey information DB61, the construction company DB62, the user proposal DB63, the personal matching report DB64, and the customer insight report DB65.

[0019] The storage unit 57 can be made up of semiconductor memory or a hard disk, and stores a computer program 58 (program product) and necessary information.

[0020] The computer program 58 can be read by a recording medium (e.g., an optically readable disc storage medium such as a CD-ROM) M using a recording medium reading unit (not shown) and stored in a storage unit 57. The computer program 58 may also be read by a recording medium such as a storage device (semiconductor memory such as an SSD (Solid State Drive)) connected by a standard for connecting to a computer (e.g., USB (Universal Serial Bus) or other standards) and stored in a storage unit 57. Alternatively, the computer program 58 may be downloaded from an external device via a communication unit 52 and stored in a storage unit 57.

[0021] The memory 53 can be composed of semiconductor memory such as SRAM (Static Random Access Memory), DRAM (Dynamic Random Access Memory), or flash memory. The computer program 58 is loaded into the memory 53, and the control unit 51 can execute the computer program 58. The control unit 51 can execute the processing defined in the computer program 58. In other words, the processing performed by the control unit 51 is also the processing performed by the computer program 58.

[0022] The information processing device 50 is operated, for example, by a company that conducts custom home matching services. To realize a service that supports users (customers, homeowners) in building their homes, it can provide a matching platform that matches users with construction companies that can meet the users' aspirations for building their homes, thereby improving user satisfaction. The information processing device 50 is used by a person in charge at a company that conducts custom home matching services (for example, a person in charge of the sales department, hereinafter also referred to as "person in charge"). Details of the matching platform will be described later.

[0023] The survey information DB61 is a database that stores user survey information.

[0024] The Construction Company DB62 is a database that stores characteristic information about construction companies.

[0025] The user-facing suggestion DB63 is a database that stores user-facing suggestions obtained by the information processing device 50 from the large-scale language model 100.

[0026] The Personal Matching Report DB64 is a database that stores the personal matching reports obtained by the information processing device 50 from the large-scale language model 100.

[0027] The Customer Insight Report DB65 is a database that stores customer insight reports obtained by the information processing device 50 from the large-scale language model 100.

[0028] Figure 2 shows an example of survey information. Survey information is information stored in the survey information DB61. Survey information is a compilation of responses entered by users to questionnaires provided to them in advance. As shown in Figure 2, survey information includes items such as age, household income, family structure, desired area for home purchase, land ownership, home purchase budget, items important in home building, concerns about home building, requests for consultation, building intentions, and free-response items.

[0029] The survey items include: age "OO years old", household income "OXO million yen", family structure "couple + 1 child", desired housing area "Setagaya Ward, Tokyo", land ownership "no land", housing purchase budget "△△△△ million yen", items considered important in home building "1st: earthquake resistance, 2nd: insulation and airtightness, 3rd: design", concerns about home building "whether my wishes can be fulfilled within budget and whether the construction company is trustworthy", desire for consultation "I would definitely like to consult", intention to build "I would like to start within one year", and free comment "My child has allergies, so I am also interested in natural materials". Note that the survey information is just an example and is not limited to the example in Figure 2. The content of the survey information may be changed depending on the content or type of report generated by the information processing device 50, or it may be changed when the information processing device 50 generates a report again after it has been generated, or it may be changed according to the user.

[0030] Figure 3 shows an example of characteristic information for a construction company. Characteristic information for construction companies (hereinafter also referred to as "characteristic information") is information stored in the Construction Company DB62. Characteristic information is detailed information that shows the characteristics of a construction company. As shown in Figure 3, characteristic information includes items such as company name, head office location, category, business scale, main areas of expertise, estimated price per square foot, technology and design characteristics, market evaluation, strategic segments, and B2B opportunities.

[0031] The details of the item are as follows: Company name is "AAA Co., Ltd.", Head office location is "Bunkyo-ku, Tokyo...", Category is "Regional Powerhouse, Design Niche Specialist", Business scale is "Sales of OO0 billion yen, number of houses constructed per year is approximately XXX", Main area of ​​expertise is "Small housing, cost performance", Estimated price per tsubo is "OO yen to OX yen", Technical and design characteristics are "Wooden frame construction, semi-custom style, Space maximization technology on small plots (skylights, skeleton staircases, etc.), Models compliant with HEAT20 standards available", Market evaluation is "Highly rated for design ability on difficult site conditions, Many praise the attentive service of the staff, Concerns: After-sales service response, variability in construction quality", Strategic segment is "Design Niche Specialist", B2B opportunity is "Mass supply, space-saving housing". Note that the characteristic information is an example and is not limited to the example in Figure 3.

[0032] Next, we will describe the proposals and reports provided by the matching platform using the information processing device 50.

[0033] Figure 4 shows an example of how the information processing device 50 generates proposals and reports. Figure 4(A) shows an example of how to generate proposals for users. Proposals for users include, for example, a basic proposal report and a thank-you email / proposal notification email to the user. The basic proposal report includes a list of a predetermined number of recommended construction companies (e.g., 5 companies), a compatibility score between each construction company and the user, reasons for matching (reasons for recommendation), attractive points (strengths of the construction company), etc.

[0034] As shown in Figure 4(A), when the person in charge inputs "questionnaire information," "construction company characteristic information," and "prompts for generating user-oriented proposals" from the operation unit 55, the information processing device 50 inputs (provides) the inputted "questionnaire information," "construction company characteristic information," and "prompts for generating user-oriented proposals" to the large-scale language model 100. Based on the inputted "questionnaire information," "construction company characteristic information," and "prompts for generating user-oriented proposals," the large-scale language model 100 generates user-oriented proposals and inputs the generated user-oriented proposals to the information processing device 50. As a result, the information processing device 50 can obtain the user-oriented proposals generated by the large-scale language model 100.

[0035] As described above, the control unit 51 acquires survey information about housing entered by the user, and uses a large-scale language model 100 (language model) that references the acquired survey information and characteristic information about the characteristics of multiple construction companies stored in the construction company DB 62 to output a predetermined number of recommended information (suggestions for the user) for each construction company.

[0036] This allows us to suggest suitable construction companies to users considering building a custom home.

[0037] Figure 4(B) shows an example of how to generate a personal matching report. The personal matching report is a rich report that rewrites the "User Suggestions" shown in Figure 4(A) in a way that is easier for users to understand and view, and is equivalent to a detailed version of the "User Suggestions."

[0038] As shown in Figure 4(B), when the person in charge inputs "questionnaire information," "user suggestions," and "prompts for generating a personal matching report" from the operation unit 55, the information processing device 50 inputs the inputted "questionnaire information," "user suggestions," and "prompts for generating a personal matching report" into the large-scale language model 100. Based on the inputted "questionnaire information," "user suggestions," and "prompts for generating a personal matching report," the large-scale language model 100 generates a personal matching report and inputs the generated personal matching report into the information processing device 50. As a result, the information processing device 50 can obtain the personal matching report generated by the large-scale language model 100.

[0039] Figure 4(C) shows an example of how to generate a customer insight report. The customer insight report is a report strategically restructured for use by sales representatives of recommended construction companies that have matched the user, as shown in the "User Proposal" in Figure 4(A), for initial appointments with the user. The customer insight report is a sales strategy sheet that summarizes the user's needs, priorities, and potential concerns, and can be used to optimize the initial approach with the user.

[0040] As shown in Figure 4(C), when the person in charge inputs "questionnaire information," "characteristics information of the matching construction company," and "prompt for generating a customer insight report" from the operation unit 55, the information processing device 50 inputs the inputted "questionnaire information," "characteristics information of the matching construction company," and "prompt for generating a customer insight report" into the large-scale language model 100. Based on the inputted "questionnaire information," "characteristics information of the matching construction company," and "prompt for generating a customer insight report," the large-scale language model 100 generates a customer insight report and inputs the generated customer insight report into the information processing device 50. As a result, the information processing device 50 can obtain the customer insight report generated by the large-scale language model 100.

[0041] Next, we will explain the details of generating user recommendations, personalized matching reports, and customer insight reports. First, we will explain the details of generating user recommendations.

[0042] Figure 5 shows an example of a prompt for generating user suggestions. The elements that make up a prompt (also called a "system prompt") may include, for example, the role, output format, supplementary definitions, and execution instructions.

[0043] Your role might be something like this: "You are a specialist advisor who selects the five best house builders (construction companies) for a customer considering building a custom home. Based on customer survey information, match the characteristics of each house builder and propose five companies that are estimated to have a high success rate. Refer to the file of construction company characteristics information as part of your knowledge base."

[0044] The output format can include, for example, the following as selection results: 1. Classification based on the user's basic needs, 2. Five recommended house builders, and 3. Creation of an email message to the user regarding the selection results.

[0045] Supplementary definitions can, for example, define the systematization of classification axes, the logic for scoring housing construction intentions, and the logic for scoring compatibility between homeowners (users) and house builders.

[0046] An example of an execution instruction might be, "Create suggestions for the user's survey information based on the 'role,' 'output format,' and 'supplementary definition' defined above."

[0047] Please note that the prompts for generating user suggestions are just examples and are not limited to the items and their contents shown in Figure 5.

[0048] Next, we will explain, in order, the systematization logic of the classification axes defined in the supplementary definition, the calculation logic for the home building intention score, and the calculation logic for the compatibility score between the homeowner and the house builder.

[0049] Figure 6 shows an example of the systematization logic for "classification axes." The systematization logic for classification axes is a logic for classifying types of "ideal homes" in order to clarify the direction of the "ideal home" that users vaguely have in mind. In this embodiment, five classification axes, "cost," "performance," "design," "support," and "location," are combined to verbalize and classify types of ideal homes.

[0050] The cost axis determines whether you prefer a "low-cost," "mid-range," or "high-grade" price range based on your housing purchase budget and household income.

[0051] The performance axis is based on "points that are important in home building," and classifies people into categories such as "performance-oriented type" or "insulation / airtightness-focused type" based on whether performance-related items (e.g., earthquake resistance, insulation, etc.) are ranked highly, or whether they have a high level of interest in ZEH (Zero Energy House).

[0052] The design axis involves, for example, linking the desired design (e.g., simple modern, Scandinavian style) to the architectural firm's area of ​​expertise.

[0053] The support axis categorizes the needs for support systems in the construction process, such as whether one prioritizes "land search" or "warranty."

[0054] Location is used to classify properties based on factors such as the desired area for purchasing a home, transportation convenience, surrounding environment, necessary facilities, safety, and future asset value.

[0055] By combining these classification axes, for example, users who want to keep their budget down but don't want to compromise on insulation and earthquake resistance can be classified as "low-cost + performance-oriented." On the other hand, users who are particular about design and have some leeway in their budget can be classified as "design-oriented mid-range."

[0056] By utilizing classification axes, it is possible to roughly narrow down the pool of potential construction companies to propose from among many. This can serve as an important initial filter to prevent mismatches, such as proposing only high-grade construction companies to a user who desires low cost.

[0057] Figure 7 shows an example of the scoring logic for "housing construction intention." The scoring logic for housing construction intention objectively identifies and scores the user's level of enthusiasm (also referred to as "enthusiasm") for housing construction. Based on the housing construction intention score, the way recommended construction companies are proposed or the selection policy for recommended construction companies can be adjusted.

[0058] The calculation logic assigns points to the following items and evaluates based on the total points. If the item is "80% or higher response rate for survey information items," the score is increased by +1 point, and the reason for the score is "Serious information entry."

[0059] If the entry for "Other requests regarding home building" contains 20 or more characters, the score is increased by +3 points, and the reason for the score is "Specific requests / ideas are present."

[0060] If the item is "I would like to discuss this," the score is +3 points, and the reason for the score is "I am seeking specific action."

[0061] If the item is "I want to move in as soon as possible," the score is +3 points, and the reason for the score is "I have specific requests and they are a high priority."

[0062] If the item is "I want to live there within six months," the score is +2 points, and the reason for the score is "There are specific requests and the priority is somewhat high."

[0063] If the item is "I want to build it within one year," the score is +1 point, and the reason for the score is "The plan is specific."

[0064] In the example in Figure 7, the score for residential construction intention is the sum of all individual scores, resulting in 10 points. Note that the calculation logic for the residential construction intention score is just one example and is not limited to the example in Figure 7; items or scores can be changed, deleted, or added as appropriate.

[0065] Figure 8 shows an example of a proposal strategy based on a residential construction intention score. Scores can be classified into, for example, high scores (7-10 points), medium scores (4-6 points), and low scores (0-3 points). For example, a high score might be described as "Judging that the user has a high level of enthusiasm for construction. We will directly propose the construction companies that best match the user's needs, in order of highest matching rate." A medium score might be described as "Judging that the user has a moderate level of enthusiasm for construction. Considering the possibility that they are still in the comparison and consideration stage, we will propose five companies that offer a balanced selection, including a wide range of price points or design flexibility." A low score might be described as "Judging that the user has a low level of enthusiasm for construction. We will consider this to be the initial stage of information gathering and select construction companies that are skilled at providing careful follow-up, such as those that will listen attentively to the user's pace without rushing them." Note that the score classification method and the content of the proposal strategies are examples and are not limited to the example in Figure 8.

[0066] As described above, the control unit 51 can output recommendation information by identifying the user's level of enthusiasm for building based on the user's questionnaire information and providing the large-scale language model 100 with proposed building company policies corresponding to the identified level of enthusiasm.

[0067] Figure 9 shows an example of the logic for calculating the compatibility score between a client and a house builder. The compatibility score between the client (user) and the house builder (construction company) is the most important score, evaluated on a scale of 100 points, to see how well the user's "specific requests" and the construction company's "strengths" match. It is a relative evaluation among the five recommended construction companies. The compatibility score allows for a multifaceted analysis of the user's values ​​and the characteristics of the construction company, and the degree of matching can be visualized numerically.

[0068] The evaluation items and their corresponding point values ​​are as follows: "Consistency with budget" is 40 points, "Compatibility with the most important item" is 30 points, "Compatibility with the second most important item" is 15 points, "Compatibility with the third most important item" is 10 points, and "Compatibility with other requests" is 5 points.

[0069] If the evaluation item is "consistency with budget," the scoring method is as follows: "Estimate a reasonable mortgage repayment limit based on the user's annual income and age. Subtract the average land price in the planned construction area and general miscellaneous expenses (for example, 10% of construction costs) from the estimated repayment limit to calculate the "maximum budget that can be allocated to construction." Compare the maximum budget with the estimated cost calculated from the average per-square-foot cost of each construction company. The higher the feasibility within the budget, the higher the compatibility score. This evaluation item is the most important filter."

[0070] The control unit 51 can derive budget-related compatibility scores for multiple construction companies by providing the large-scale language model 100 with questionnaire information including the user's age, annual income, desired construction area, and whether or not they own land, characteristic information including the price per tsubo (unit of area), and a command to calculate a compatibility score related to budget.

[0071] When the evaluation item is "Compatibility with the most important item," the scoring method is as follows: "We evaluate whether the item that the user answered as 'most important' (for example, earthquake resistance, design, cost performance, etc.) matches the greatest strength of the construction company. For example, if the most important item is 'insulation,' the compatibility score of a construction company that boasts the highest level of performance in the industry or a construction company that provides high-performance housing according to Nordic standards will be significantly higher."

[0072] If the evaluation item is "Compatibility with the second most important item," the scoring method is to "evaluate the degree of match between the second most important item and the strengths of the construction company." If the evaluation item is "Compatibility with the third most important item," the scoring method is to "evaluate the degree of match between the third most important item and the strengths of the construction company."

[0073] The control unit 51 can derive compatibility scores for each of the important items of multiple construction companies by providing the large-scale language model 100 with questionnaire information including prioritized important items, characteristic information of construction companies, and a command to calculate compatibility scores for each important item of multiple construction companies for important items up to a predetermined priority.

[0074] When the evaluation item is "Compatibility with other requests," the scoring method is as follows: "• Identify potential needs that can be gleaned from the questionnaire information other than the important items (for example, 'I'm concerned about whole-house air conditioning' in the free-response section, or 'I'd also like help with finding land' in the options) and evaluate them as bonus points. • Evaluate the degree of match between potential needs and the strengths of the construction company." Note that the example in Figure 9 is just one example, and the evaluation items, scoring ratios, and scoring method are not limited to the example in Figure 9.

[0075] As described above, the control unit 51 can extract recommended construction companies based on an overall compatibility score identified by the budget compatibility scores for multiple construction companies and the priority compatibility scores for each important item included in the questionnaire information of multiple construction companies.

[0076] Figure 10 shows an example of a basic proposal report (a proposal for users). The basic proposal report lists five recommended construction companies (BBB Co., Ltd., CCC Co., Ltd., DDD Co., Ltd., EEE Co., Ltd., and FFF Co., Ltd.).

[0077] BBB Co., Ltd. has a compatibility score of 92 points, and the reason for the match is "They are a truly reliable local expert rooted in your desired area, Setagaya Ward, and you can proceed with the process from land searching to direct communication with an architect." Their attractive points are "All staff members are architects and provide consistent support from design to site management. Their absolute reliability is appealing, as they have had zero financial troubles after completion since their founding."

[0078] CCC Corporation has a compatibility score of 90 points, and the reason for matching is that it is an extremely rational choice that meets the request for "a house that does not compromise on performance within budget" with the best cost performance. Its attractive points are: "It boasts overwhelming performance with seismic resistance level 3 and insulation level 6 (HEAT20 G2.5 level) as standard specifications. A generous warranty system, including a 20-year ground warranty, ensures peace of mind for the future."

[0079] DDD Corporation received a compatibility score of 95 points, with the reason for the match being, "It is the best choice when you want to combine the outstanding design quality created by 'wood professionals' with the comprehensive capabilities and peace of mind that only a large company can offer." Its attractive points are, "• Their unique BF construction method allows for the design of large openings that combine the warmth of wood with earthquake resistance. • They have a comprehensive support system that takes future maintenance costs into consideration, including a long-term warranty program of over 60 years."

[0080] EEE Co., Ltd. has a compatibility score of 85 points, and the reason for the match is "The value of healthy, comfortable, and high-performance housing using Tokyo wood 'TOKYO WOOD' will surely resonate deeply with the discerning OOO customer." Its attractive points are "We practice high-quality and environmentally conscious home building using locally sourced Tama timber. All our homes have the highest seismic resistance rating of 3, and we achieve high levels of comfort with our external insulation method."

[0081] FFF Co., Ltd. has a compatibility score of 82 points, and the reason for matching is "If your budget allows, they will realize your dream of pursuing design to the ultimate degree in a unique way in collaboration with an architect." The attractive points are "• Their design proposals, which are described as "spaces like hotels and art museums" by architects, set them apart from others. • They create only 30 buildings per year, boasting a quality that is truly worthy of being called a "work of art."

[0082] As described above, the control unit 51 can extract recommended construction companies by providing the user's survey information and the characteristics information of the construction companies to the large-scale language model 100.

[0083] Please note that the basic proposal report is just one example and is not limited to the example shown in Figure 10.

[0084] Figure 11 shows an example of a thank-you email and proposal notification email (user-oriented proposal) sent to a user. As shown in Figure 11, the thank-you email and proposal notification email sent to the user is an email that introduces recommended construction companies that are a good match for the user, "Mr. / Ms. OOO," and includes text expressing gratitude, text selecting recommended construction companies, and text regarding meetings with construction companies, all related to the basic proposal report. The "House Building Chart" is included in the personal matching report. Note that the content of the thank-you email and proposal notification email sent to the user is just an example and is not limited to the example in Figure 11.

[0085] The control unit 51 provides the large-scale language model 100 with an output command for each extracted construction company, including an overall compatibility score, reasons for matching with the construction company, and recommended information including the attractiveness of the construction company. This allows the control unit 51 to output recommended information (e.g., a basic proposal report) including the overall compatibility score, reasons for matching with the construction company, and attractiveness of the construction company for each extracted construction company to the user's terminal device.

[0086] Next, we will explain the details of generating a personal matching report.

[0087] Figure 12 shows an example of prompts for generating a personal matching report. The items that make up the prompts can include, for example, instructions, a previously generated list of 5 recommended companies, and the output format.

[0088] The instructions might read something like this: "You are a 'Custom Home Matching Advisor.' The keyword 'Personal Matching Report (Detailed Version)' has been entered. Based on the content of the previously generated 'Basic Proposal Report' and the following questionnaire information, structure the information and generate a designed report that can be output."

[0089] The list of five recommended companies generated earlier includes, for example, 1. BBB Co., Ltd., 2. CCC Co., Ltd., 3. DDD Co., Ltd., 4. EEE Co., Ltd., and 5. FFF Co., Ltd.

[0090] The output format can be a "Personal Matching Report (Detailed Version)," which can include 1. "Customer's Home Building Profile," 2. "Construction Company Profile," and 3. "Next Steps," etc.

[0091] The "Customer's Home Building Profile" can include elements such as "Home Building Theme," "Basic Needs," "Area and Budget," "Values ​​that are Important," "Potential Interests," and a "Matching Dashboard." The "Matching Dashboard" can include, for example, an "Overall Compatibility Score," a chart showing the degree of match with the customer's priorities (e.g., seismic resistance, insulation, design, support, cost, etc.) (e.g., a radar chart), and "Reasons for Matching."

[0092] The "Construction Company Profile" can include, for example, "Reasons for Matching," "Breakdown of Compatibility Score," "Advantages of Choosing This Company," and "Points to Note."

[0093] The "next step" can include information regarding the email invitation for the initial consultation. Note that the prompts for generating the personal matching report are examples only and are not limited to the example in Figure 12.

[0094] As described above, the control unit 51 can output a detailed report of the extracted construction companies (personal matching report) by giving a command to the large-scale language model 100 to output detailed information including user questionnaire information, recommended information of the extracted construction companies, an overall compatibility score of the extracted construction companies, a graph of the compatibility score between the user's important items included in the questionnaire information and the extracted construction companies, and diagnostic information of the extracted construction companies.

[0095] Figure 13 shows an example of a personal matching report. The items in the personal matching report generated by following the prompts shown in Figure 12 include, for example, "Theme for building your home," "Basic needs," "Area," "Budget," "Values ​​you prioritize," "Potential interests," "Matching dashboard," "Building company profile," and "Next steps."

[0096] The "theme for building the house" could be, for example, "A high-grade house in Setagaya Ward, Tokyo, that combines performance and design."

[0097] "Basic needs" can be defined, for example, as "high-grade + balance of performance and design."

[0098] The "area" can be, for example, a prefecture such as "Tokyo".

[0099] The "budget" can be expressed as, for example, "XXXX million yen to △△△△ million yen."

[0100] "Values ​​that are prioritized" can be defined as, for example, "1. Safety (earthquake resistance), 2. Comfort (insulation), 3. Design."

[0101] "Potential interests" could be, for example, "natural materials."

[0102] The "matching dashboard" could, for example, display "detailed data on the five companies in the basic proposal report."

[0103] The "Construction Company Profile" can be, for example, "Profiles for the 5 Companies in the Basic Proposal Report." For the "Profiles for the 5 Companies in the Basic Proposal Report," the reason for matching could be, for example, "Regarding 'earthquake resistance,' which is of utmost importance to the customer, the strength against disasters provided by ALC concrete is unmatched by other companies. The long-term support system of over 60 years also matches the needs of customers who value reliability." The "Compatibility Score Breakdown" could be, for example, "-Earthquake Resistance: ★★★★★, -Insulation: ★★★★☆, -Design: ★★★★☆." The "Advantages of Choosing This Company" could be, for example, "Overwhelming peace of mind against earthquakes and fires, and maintenance of asset value over the long term." The "Points to Note" could be, for example, "Does the heavy design suit your taste? You may feel design constraints compared to wooden houses." The "Construction Company Profile" is generated in the same format for each of the 5 companies in the Basic Proposal Report.

[0104] The "next step" could be something like, "You will receive an email from the company you selected with information about your initial online consultation. Having a memo outlining your vision for your home and your requirements will help the conversation go more smoothly."

[0105] The personal matching report shown in Figure 13 is an example, and the items and specific examples are not limited to those shown in Figure 13.

[0106] Figure 14 shows an example of a matching dashboard. As shown in Figure 14, the matching dashboard includes, for example, the "overall compatibility score" of the recommended construction company, "deciding factors for matching," "reasons for matching," and a "radar chart that shows the degree of matching with each of the five classification axes, including performance, design, support, cost, and location." The reasons for matching show the reasons for matching with the customer's "ideal home" from the perspective of the classification axes. The radar chart is just an example and is not limited to this; other charts such as bar graphs, line graphs, histograms, or parallel coordinate plots may also be used.

[0107] Next, we will explain the details of generating customer insight reports.

[0108] Figure 15 shows an example of prompts for generating a customer insights report. The elements that make up the prompt can include, for example, instructions, matching company information, and output format.

[0109] The "instructions" could be something like this: "You are a 'Custom Home Matching Advisor.' The keyword 'Customer Insight Report' has been entered. Based on the following survey information and the company information of the matching partner, 'DDD,' please generate a 'Customer Insight Report (Initial Appointment Strategy Sheet)' to give to their sales representative."

[0110] "Matching partner company information" could, for example, include company name: {DDD}, and strengths: "commitment to wood, design flexibility with Big Frame (BF) construction method, sense of luxury, outstanding design, and strong brand power among affluent customers."

[0111] The "Output Format" can be set as a "Customer Insight Report (Initial Appointment Strategy Sheet)" and can include: 1. "Customer Profile", 2. "Personal Insight Analysis", 3. "Potential Needs", 4. "Initial Appointment Strategy", 5. "Recommended Action Plan and Next Steps", 6. "Anticipated Q&A", and 7. "Competitor Countermeasures".

[0112] The "customer profile" can include, for example, "basic information," "enthusiasm for building a home," "points that are important to them in building a home," and "concerns they have during the consideration process."

[0113] The control unit 51 can output a user report (customer insight report) for the first construction company by giving commands to the large-scale language model 100 to output user survey information, recommended information from extracted construction companies, profile information including the user's level of enthusiasm for construction, a recommended approach for recommending the first construction company (matching company), anticipated Q&A, and insight information including competitive countermeasures for extracted construction companies other than the first construction company.

[0114] The prompt shown in Figure 15 is just one example, and the items and specific examples of items are not limited to the example in Figure 15.

[0115] Figure 16 shows an example of a customer insights report. The items in the customer insights report generated according to the prompts shown in Figure 15 include, for example, "Customer Profile," "Personal Insight Analysis," "Uncovering Hidden Customer Needs," "Initial Appointment Strategy," "Recommended Action Plan," "Next Steps," "Anticipated Q&A," and "Competitor Strategy."

[0116] The "customer profile" includes, for example, "basic information," "level of enthusiasm for building," "points that are important in building a house," and "concerns during consideration." The "basic information" includes, for example, the customer's name, age, current address, desired building area, dissatisfaction with their current home (e.g., noise and sound leakage, poor ventilation, concerns about being overlooked from outside), and an outline of their home building plan (e.g., current status of consideration, land conditions, household income, etc.). The "level of enthusiasm for building" could be, for example, 9 points (out of 10), broken down as follows: "3 points for wanting to build within one year, 3 points for high rate of voluntary responses, 1 point for providing free-form text, and 2 points for wanting face-to-face consultation." The "points that are important in building a house" could be, for example, "1st place: location, 2nd place: housing performance (earthquake resistance, durability, sound insulation), 3rd place: exterior and interior design." "Concerns during consideration" can include, for example, "1. Choosing the right plot of land, 2. Neighbor disputes or changes in the environment, and 3. The gap between the image of the completed house and reality."

[0117] "Personal Insight Analysis" analyzes customer types, which can be categorized as, for example, "Cautious and Analytical," "Introspective and Privacy-Oriented," or "Independent and Plan-Oriented." The "Cautious and Analytical" type prioritizes functional value (location, performance, etc.) and logically examines information to avoid mistakes. They prefer objective facts based on data over sensory appeals. The "Introspective and Privacy-Oriented" type is sensitive to noise or outside glances, has a strong desire to make their home a truly relaxing "personal sanctuary," and values ​​privacy and quietness. The "Independent and Plan-Oriented" type dislikes being led by others, has a clear intention to proceed with things systematically at their own pace, and dislikes being rushed.

[0118] "Uncovering the hidden needs of our customers" can be expressed as, for example, 1. "The ultimate personal space," 2. "An investment in lasting peace of mind," and 3. "An equal partnership." "The ultimate personal space" is a need where, for example, the customer is not simply looking for a "house," but rather a "castle of their own" completely protected from the outside world, a place of absolute privacy where they can immerse themselves in hobbies and contemplation without being disturbed by anyone. "An investment in lasting peace of mind" is a need where, for example, the customer views building a house as an "investment" in a peaceful life for years to come, and, given their concerns about "environmental changes," they prioritize long-term durability and security through warranties over short-term comfort. "An equal partnership" is a need where, for example, the customer seeks a sales representative from the construction company to act as a "professional advisor" who deeply understands and respects their values, and they desire a partner who will work with them on the house-building project on an equal footing.

[0119] The "initial appointment strategy" can be defined as the "core of the proposal," for example, "We will build a castle just for you, using the best technology." Furthermore, the "matching strategy" can be described from three perspectives: "performance," "support," and "values." "Performance matching" involves logically explaining, using specific figures, how the construction company's (our company's) methods and materials achieve the "sound insulation" and "durability" that the customer values. "Support matching" involves clearly communicating that the biggest concern, "land selection," can be addressed by presenting the construction company's (our company's) real estate network or construction track record, providing a one-stop solution. "Value matching" involves clearly demonstrating respect for the customer's value of "wanting to proceed at their own pace," establishing a position as a "trustworthy partner."

[0120] The "Recommended Action Plan" presents recommended actions for each phase, such as building trust (initial stages), empathizing with the problem (listening), proposing solutions (proposals), and laying the groundwork for the next steps (closing).

[0121] The "next steps" could, for example, respect the customer's lead-driven process and include things like "submission of materials for consideration," "meetings at the customer's pace," and "a clear follow-up plan."

[0122] The "anticipated Q&A" can be structured as follows: Q: "To what extent can we focus on design within the budget?" A: "Provide a direction for the answer."

[0123] "Competitive measures" could, for example, involve identifying assumed competitors as BBB and EEE, and defining differentiation points as "emphasizing overwhelming expertise as a 'wood professional' and the story behind the timber supplied from company-owned forests." Note that the content of the customer insight report is just an example and is not limited to the example in Figure 16.

[0124] Next, we will discuss the creation of a universal market analysis report for companies engaged in custom home matching services to analyze and formulate which construction companies they should focus on partnering with in the future. This market analysis report is a "general-purpose market insight report" for B2B business strategy planning.

[0125] Figure 17 shows an example of prompts for generating a general-purpose market insight report. The items that make up the prompts can include, for example, instructions and report structure. The "General-Purpose Market Insight Report" is intended to analyze which markets (areas) and with what characteristics it would be most effective for a custom home matching business to partner with. The intended readers of the "General-Purpose Market Insight Report" are management, business development personnel, and marketing personnel of companies engaged in custom home matching.

[0126] The "instructions" could be something like, "You are a B2B strategy consultant for the housing market. Integrate and analyze the provided databases of 'List of Leading Construction Companies in the Tokyo Metropolitan Area' and 'List of House Builders Nationwide,' and create a 'Strategic Analysis Report on the Custom-Built Housing Market in the Tokyo Metropolitan Area' with the following structure."

[0127] The "report structure" can include, for example, 1. an executive summary, 2. an analysis of market characteristics, 3. strategic segmentation of companies, and 4. a list of recommended targets.

[0128] The "Executive Summary" can include market overview and key trends (e.g., rising prices, shift towards performance), key analysis results and insights. The "Market Characteristics Analysis" could be, for example, a comparative analysis of the characteristics of the markets in Tokyo, Kanagawa, Saitama, and Chiba prefectures. The "Strategic Segmentation of Companies" could be, for example, "Classify all companies in the database into the following four strategic segments, and list the characteristics of each segment and representative companies: A. High Performance / Technology-Driven, B. Design / Niche Specialized, C. Regional Power Builder / Cost-Effective, D. Trust / Tradition-Focused," and "Recommendations for effective B2B engagement strategies for each segment." The "Recommended Target List" could be, for example, "Based on the analysis, select five target companies with high priority for business partnerships and clearly describe the reasons."

[0129] The prompt shown in Figure 17 is an example, and the items and their contents are not limited to the example in Figure 17.

[0130] Next, I will explain the "Personalized Market Insight Report," which supplements proposals for specific users. The "Personalized Market Insight Report" provides supplementary explanations for the five recommended construction companies selected in the user proposal, explaining why those five companies are the best fit for that user, including comparisons with the market environment and competitors, in order to enhance the user's understanding and acceptance of the proposal. The target audience for the "Personalized Market Insight Report" is specific users who are considering building a house.

[0131] Figure 18 shows an example of prompts for generating personalized market insight reports. The items that make up the prompts may include, for example, instructions and report structure.

[0132] The "instructions" could be something like, "You are an excellent housing consultant. Based on the provided database, please create a 'Personalized Market Insights Report' with the following structure as supplementary material to support 'Mr. / Ms. OO's' home building project. The report should avoid technical jargon and be written in simple language that is easy for Mr. / Ms. OO, who is considering building a home, to understand."

[0133] The report structure could include: 1. Market trends in Tokyo, where Mr. / Ms. OO will be building their home; 2. Positioning analysis of the five proposed companies in Tokyo; and 3. Why these five companies are recommended for Mr. / Ms. OO.

[0134] "Market Trends in Tokyo, Where Mr. / Ms. OO is Building His House" briefly explains recent trends in custom-built homes in Tokyo. "Position Analysis of the Five Proposed Companies in Tokyo" explains each company's position (strengths, expertise) in the Tokyo housing market, comparing it with competitors. "Why These Five Companies are Recommended for Mr. / Ms. OO" connects Mr. / Ms. OO's requests with the market positions of the five companies mentioned above, leading to the conclusion that "these five companies are therefore the most suitable partners for Mr. / Ms. OO's ideal home building project."

[0135] The prompt shown in Figure 18 is an example, and the items and their contents are not limited to the example in Figure 18.

[0136] Figure 19 shows an example of the procedure for report creation processing by the information processing device 50. The control unit 51 acquires user questionnaire information (S11) and acquires characteristic information of the construction company (S12). The control unit 51 inputs the user questionnaire information, characteristic information of the construction company, and prompts for generating user-oriented proposals, prompts for generating a personal matching report, and prompts for generating a customer insight report into the large-scale language model 100 (S13).

[0137] The control unit 51 obtains user suggestions, a personal matching report, and a customer insight report from the large-scale language model 100 (S14). The control unit 51 outputs the obtained reports (S15) and terminates the process.

[0138] According to this embodiment, it is possible to suggest a suitable construction company to a user who is considering building a custom-built home.

[0139] Furthermore, this embodiment can resolve users' (homeowners') concerns about building a house, such as not knowing which house builder (construction company) is reliable, not knowing which house builder is right for them, and not wanting to regret their home building decision.

[0140] Furthermore, this embodiment allows for matching users with construction companies, enabling users to build a home they won't regret. Specifically, the AI ​​(language model) narrows down the most suitable construction company candidates for the user, reducing the burden of selecting a construction company; it provides a neutral diagnosis and objective compatibility score by AI; and it fosters a sense of anticipation for building a home.

[0141] Furthermore, for construction companies, it enables them to deliver proposals for home construction that incorporate their strengths to clients with a high conversion rate. Specifically, by selecting and introducing only highly motivated (serious) users, it reduces wasted sales effort and enables efficient user acquisition. By providing an initial appointment strategy based on personality and emotional analysis in addition to basic user information, it is possible to improve the conversion rate with proposals that accurately grasp the user's needs. Comparative analysis with competitors allows for the clarification of the company's strengths and areas for improvement. By having AI (language model) navigate the content of proposals or conversations, it is possible to reduce the reliance on individual salespeople and improve overall performance.

[0142] (Note 1) The information processing method involves having a computer perform a process that involves acquiring survey information about housing entered by the user, and using a language model that references the acquired survey information and characteristic information about the characteristics of multiple construction companies stored in the memory unit, to output a predetermined number of recommended information for each construction company.

[0143] (Note 2) The information processing method involves providing the aforementioned questionnaire information and characteristic information to the language model, as described in Note 1, and having a computer perform the process of extracting recommended construction companies.

[0144] (Note 3) The information processing method involves, in Note 1 or Note 2, determining the user's level of enthusiasm for construction based on the questionnaire information, and then providing the language model with the construction company's proposed policy corresponding to the determined level of enthusiasm, thereby causing the computer to perform the process of outputting recommended information.

[0145] (Note 4) The information processing method involves having a computer perform a process to extract recommended construction companies based on the overall compatibility score identified by the budget compatibility score for the multiple construction companies and the compatibility score for each priority item included in the questionnaire information of the multiple construction companies, as specified in any one of Notes 1 to 3.

[0146] (Note 5) The information processing method involves providing the language model with a command to calculate a compatibility score related to budget for the multiple construction companies, based on any one of Notes 1 to 4, including questionnaire information such as age, annual income, desired construction area and whether or not land is owned, characteristic information including price per tsubo, and budget-related compatibility scores, thereby causing the computer to perform the process of deriving a budget-related compatibility score for the multiple construction companies.

[0147] (Note 6) The information processing method involves providing the language model with a command to calculate the compatibility score for each of the multiple construction companies for each of the priorities, based on any one of Notes 1 to 5, including questionnaire information containing priority items, the characteristic information, and the priorities up to a predetermined priority, thereby causing the computer to perform a process to derive the compatibility score for each of the multiple construction companies for each of the priorities.

[0148] (Note 7) The information processing method involves providing the language model with an output command for each of the extracted construction companies, including the overall compatibility score, the reason for matching with the construction company, and the attractiveness of the construction company, thereby causing the computer to perform a process that outputs the overall compatibility score, the reason for matching with the construction company, and the attractiveness of the construction company to the user's terminal device.

[0149] (Note 8) The information processing method involves, in any one of Notes 1 to 7, giving the language model a command to output detailed information including the questionnaire information, the recommended information of the extracted construction companies, the overall compatibility score of the extracted construction companies, a graph of the compatibility score between the user's important items included in the questionnaire information and the extracted construction companies, and diagnostic information of the extracted construction companies, thereby causing the computer to perform a process to output a detailed report of the extracted construction companies.

[0150] (Note 9) The information processing method involves providing the language model with a command in any one of Notes 1 to 8 to output the survey information, the recommended information of the extracted construction companies, profile information including the user's level of enthusiasm for construction, a recommended approach for recommending the extracted first construction company, anticipated Q&A, and insight information including competitive measures against extracted construction companies other than the first construction company, thereby causing the computer to perform a process to output a user report for the first construction company.

[0151] (Note 10) The information processing device includes a control unit, which acquires survey information about housing entered by the user, and outputs a predetermined number of recommended information for each construction company, using a language model that references the acquired survey information and characteristic information about the characteristics of multiple construction companies stored in the memory unit.

[0152] (Note 11) The computer program obtains survey information about housing entered by the user, and uses a language model that references the obtained survey information and characteristic information about multiple construction companies stored in the memory unit to output a predetermined number of recommended information for each construction company.

[0153] The matters described in each embodiment can be combined with each other. Furthermore, the independent and dependent claims described in the claims can be combined with each other in any combination, regardless of the form of reference. In addition, the claims use a form in which claims referencing two or more other claims (multi-claim form), but are not limited to this. A form in which multi-claims referencing at least one multi-claim (multi-multi-claim) may also be used. [Explanation of Symbols]

[0154] 1. Communication Network 50 Information Processing Devices 51 Control Unit 52 Communications Department 53 memory 54 Display section 55 Operation section 56 Interface section 57 Memory section 58 Computer Programs 61 Survey Information Database 62 Construction Company Database 63 User-oriented suggestion database 64 Personal Matching Report Database 65 Customer Insight Report Database 100 Large-Scale Language Models

Claims

1. We obtain survey information about housing entered by users. Based on the acquired survey information and a language model that references characteristic information about multiple construction companies stored in the memory unit, recommended information is output for each of the predetermined number of selected construction companies. Furthermore, by providing the aforementioned survey information and the aforementioned recommendation information to the language model, a report which is a detailed version of the recommendation information is output. Based on the aforementioned survey information, the degree of enthusiasm for the user's intention to build is identified. By providing the language model with the proposed policies of the construction company corresponding to the identified level of enthusiasm, recommended information is output. An information processing method that involves having a computer perform a task.

2. Obtain survey information about housing entered by the user, Based on the acquired survey information and a language model that references characteristic information about multiple construction companies stored in the memory unit, recommended information is output for each of the predetermined number of selected construction companies. Furthermore, by providing the aforementioned survey information and the aforementioned recommendation information to the language model, a report which is a detailed version of the recommendation information is output. Based on the budget compatibility scores for the aforementioned multiple construction companies, and the overall compatibility score identified from the priority-based compatibility scores for each important item included in the aforementioned questionnaire information of the aforementioned multiple construction companies, a recommended construction company is extracted. An information processing method that involves having a computer perform a task.

3. Obtain survey information about housing entered by the user, Based on the acquired survey information and a language model that references characteristic information about multiple construction companies stored in the memory unit, recommended information is output for each of the predetermined number of selected construction companies. Furthermore, by providing the aforementioned survey information and the aforementioned recommendation information to the language model, a report which is a detailed version of the recommendation information is output. By providing the language model with a command to output the aforementioned survey information, recommended information from extracted construction companies, profile information including the user's level of enthusiasm for construction, a recommended approach for recommending the extracted first construction company, a Q&A of what the first construction company should answer in response to anticipated questions from the user, and insight information including competitive strategies for other extracted construction companies besides the first construction company, a user report for the first construction company is output. An information processing method that involves having a computer perform a task.

4. By providing the aforementioned survey information and characteristic information to the language model, recommended construction companies are extracted. The information processing method according to claim 1, which causes a computer to perform the processing.

5. By providing the language model with questionnaire information including age, annual income, desired construction area, and land ownership status, characteristic information including price per square meter, and instructions for calculating a compatibility score related to budget, a compatibility score related to budget for the multiple construction companies is derived. The information processing method according to claim 2, which causes a computer to perform the processing.

6. By providing the language model with questionnaire information including prioritized items, characteristic information, and a command to calculate compatibility scores for each of the multiple construction companies for each of the priorities up to a predetermined priority, the compatibility scores for each of the multiple construction companies for each of the priorities are derived. The information processing method according to claim 2, which causes a computer to perform the processing.

7. For each selected construction company, the language model is given an output command that includes the overall compatibility score, the reason for matching with the construction company, and the attractiveness of the construction company. This command outputs the overall compatibility score, the reason for matching with the construction company, and the attractiveness of the construction company to the user's terminal device. The information processing method according to claim 2, which causes a computer to perform the processing.

8. By providing the language model with a command to output detailed information including the aforementioned survey information, recommended information of the extracted construction companies, the overall compatibility score of the extracted construction companies, a graph of the compatibility score between the user's important items included in the survey information and the extracted construction companies, and diagnostic information of the extracted construction companies, a detailed report of the extracted construction companies is output. An information processing method according to any one of claims 1 to 4, which causes a computer to perform the processing.

9. Equipped with a control unit, The control unit, We obtain survey information about housing entered by users. Based on the acquired survey information and a language model that references characteristic information about multiple construction companies stored in the memory unit, recommended information is output for each of the predetermined number of selected construction companies. Furthermore, by providing the aforementioned survey information and the aforementioned recommendation information to the language model, a report which is a detailed version of the recommendation information is output. Based on the aforementioned survey information, the degree of enthusiasm for the user's intention to build is identified. By providing the language model with the proposed policies of the construction company corresponding to the identified level of enthusiasm, recommended information is output. Information processing device.

10. comprising a control unit, The control unit, We obtain survey information about housing entered by users. Based on the acquired survey information and a language model that references characteristic information about multiple construction companies stored in the memory unit, recommended information is output for each of the predetermined number of selected construction companies. Furthermore, by providing the aforementioned survey information and the aforementioned recommendation information to the language model, a report which is a detailed version of the recommendation information is output. Based on the budget compatibility scores for the aforementioned multiple construction companies, and the overall compatibility score identified from the priority-based compatibility scores for each important item included in the aforementioned questionnaire information of the aforementioned multiple construction companies, a recommended construction company is extracted. Information processing device.

11. comprising a control unit, The control unit, We obtain survey information about housing entered by users. Based on the acquired survey information and a language model that references characteristic information about multiple construction companies stored in the memory unit, recommended information is output for each of the predetermined number of selected construction companies. Furthermore, by providing the aforementioned survey information and the aforementioned recommendation information to the language model, a report which is a detailed version of the recommendation information is output. By providing the language model with a command to output the aforementioned survey information, recommended information from extracted construction companies, profile information including the user's level of enthusiasm for construction, a recommended approach for recommending the extracted first construction company, a Q&A of what the first construction company should answer in response to anticipated questions from the user, and insight information including competitive strategies for other extracted construction companies besides the first construction company, a user report for the first construction company is output. Information processing device.

12. We obtain survey information about housing entered by users. Based on the acquired survey information and a language model that references characteristic information about multiple construction companies stored in the memory unit, recommended information is output for each of the predetermined number of selected construction companies. Furthermore, by providing the aforementioned survey information and the aforementioned recommendation information to the language model, a report which is a detailed version of the recommendation information is output. Based on the aforementioned survey information, the degree of enthusiasm for the user's intention to build is identified. By providing the language model with the proposed policies of the construction company corresponding to the identified level of enthusiasm, recommended information is output. A computer program that instructs a computer to perform a process.

13. Obtain survey information about housing entered by the user, Based on the acquired survey information and a language model that references characteristic information about multiple construction companies stored in the memory unit, recommended information is output for each of the predetermined number of selected construction companies. Furthermore, by providing the aforementioned survey information and the aforementioned recommendation information to the language model, a report which is a detailed version of the recommendation information is output. Based on the budget compatibility scores for the aforementioned multiple construction companies, and the overall compatibility score identified from the priority-based compatibility scores for each important item included in the aforementioned questionnaire information of the aforementioned multiple construction companies, a recommended construction company is extracted. A computer program that instructs a computer to perform a process.

14. Obtain survey information about housing entered by the user, Based on the acquired survey information and a language model that references characteristic information about multiple construction companies stored in the memory unit, recommended information is output for each of the predetermined number of selected construction companies. Furthermore, by providing the aforementioned survey information and the aforementioned recommendation information to the language model, a report which is a detailed version of the recommendation information is output. By providing the language model with a command to output the aforementioned survey information, recommended information from extracted construction companies, profile information including the user's level of enthusiasm for construction, a recommended approach for recommending the extracted first construction company, a Q&A of what the first construction company should answer in response to anticipated questions from the user, and insight information including competitive strategies for other extracted construction companies besides the first construction company, a user report for the first construction company is output. A computer program that instructs a computer to perform a process.

Images