An entrepreneurship guidance system and method based on big data

Abstract

The application relates to the technical field of big data and discloses an entrepreneurship guidance system and method based on big data, which comprises collecting multi-source data information of a current entrepreneurship project of a target entrepreneurship subject, wherein the multi-source data information comprises structured data and unstructured data, the structured data comprises financial data, operation data and market data, and the unstructured data comprises an entrepreneurship plan and team vision research; semantic feature extraction is performed on the entrepreneurship plan to obtain project vision evaluation, and analysis is performed on the team vision research to obtain team vision evaluation. When the semantic feature extraction is performed on the entrepreneurship plan, a pre-constructed project vision keyword library is introduced, linear scaling mapping is performed on the basis of score extreme values of historical samples of similar industries, project vision, a highly subjective element, is objectively quantified under a unified scale, and the deviation caused by text expression styles and lengths of different industries is eliminated.

Description

Technical Field

[0001] This invention relates to the field of big data technology, and in particular to a startup guidance system and method based on big data. Background Technology

[0002] Currently, entrepreneurship guidance primarily relies on the personal experience and subjective judgment of mentors, investors, or consultants, providing guidance through qualitative analysis of business plans, team composition, and market prospects. This highly expert-dependent guidance model not only suffers from inconsistent evaluation standards and strong subjectivity, but also struggles to achieve large-scale, standardized evaluation and guidance for entrepreneurial projects. While big data technology has been widely applied in finance and marketing, in the specific field of entrepreneurship guidance, there is a lack of mature technical solutions for systematically collecting and utilizing multi-source data (including structured data such as financial, operational, and market data, as well as unstructured data such as business plans and team visions) to support entrepreneurial decision-making. Existing big data entrepreneurship service platforms mainly focus on information display and resource matching, failing to achieve a complete intelligent guidance loop from data collection and semantic analysis to quantitative evaluation and strategy matching.

[0003] Specifically, existing technologies have significant shortcomings in the following key areas: First, for the highly subjective element of the project vision in a business plan, there is a lack of effective quantitative evaluation methods, making it difficult to transform textual descriptions into comparable numerical indicators; second, for assessing the alignment of vision among members of the entrepreneurial team, traditional methods often rely on questionnaires, which are easily influenced by societal expectation biases and cannot objectively reflect the true degree of consensus within the team. These technological bottlenecks make it difficult for existing entrepreneurial guidance to achieve precise, intelligent, and quantifiable evaluation and guidance. Summary of the Invention

[0004] This invention provides a big data-based entrepreneurship guidance system and method to solve existing technical problems, thereby addressing the issues mentioned in the background section.

[0005] To address the aforementioned technical problems, according to one aspect of the present invention, more specifically, a big data-based entrepreneurship guidance method, comprising the following steps: Step S1: Collect multi-source data information of the target entrepreneurial entity's current entrepreneurial project. The multi-source data information includes structured data and unstructured data. The structured data includes financial data, operating data, and market data. The unstructured data includes business plans and team vision surveys. Step S2: Extract semantic features from the business plan to obtain a project vision assessment, and analyze the team vision survey to obtain a team vision assessment; Step S3: Convert the structured data into a quantitative feature vector, and evaluate the profitability based on the project vision assessment and team vision assessment. Step S4: Analyze the matching degree between the target entrepreneurial entity and each guidance strategy in the preset entrepreneurial guidance strategy library based on the quantitative feature vector; Step S5: Based on the profitability status assessed in Step S3 and the matching guidance strategies in Step S4, guidance can be provided to the target entrepreneurial entities for their current entrepreneurial projects.

[0006] Furthermore, in step S2, the specific steps for extracting semantic features from the business plan to obtain a project vision assessment are as follows: Step 1: Extract paragraphs or sentences describing the project vision from the target entrepreneur's business plan as raw text material for analyzing the project vision; Step 2: Use a pre-built project vision keyword library (containing words such as "change", "benefit", "breakthrough", "responsibility", "mission" etc.) to match the text and record the location and frequency of each keyword; Step 3: Encode the context of each project vision keyword and assess the importance of the keyword in its specific context; Step 4: Multiply the frequency of each keyword by its contextual contribution, and sum the results of all keywords to obtain a comprehensive raw score; Step 5: Collect a large number of historical entrepreneurial samples from similar or related industries, calculate their original project vision scores in the same way, and statistically determine the highest and lowest historical scores. Step 6: Compare the original score of the current target subject with the historical range, and map it to a value between 0 and 1 by linear scaling. This is the project vision assessment.

[0007] Furthermore, in step S2, the specific steps for analyzing the team vision survey to obtain a team vision assessment are as follows: Step 1: Obtain anonymous written statements from each member of the startup team regarding the project's long-term goals. The statements should focus on "what we hope to develop the project into" or "what problem we ultimately want to solve," ensuring that the texts accurately reflect each individual's core expectations for the team's future. Step 2: Convert each vision description text into a fixed-length numerical vector. This numerical vector can abstractly represent the semantic meaning of the text, so that the more similar the content of the description, the closer the direction of their corresponding vectors in space. Step 3: Compare the directions of the vision vectors of any two members to obtain the degree of similarity in their directions; Step 4: For all possible pairwise combinations within the team, complete the directional similarity assessment obtained in Step 3, and obtain the average directional similarity score. This average score serves as a quantitative indicator of the team's overall vision consistency.

[0008] Furthermore, in step S3, the specific steps for converting structured data into quantized feature vectors are as follows: Step S301: Extract multiple measurable raw indicators from the financial system, business ledgers, and market research reports; Step S302: Based on the data distribution of a large number of similar entrepreneurial entities in the historical database, or by referring to the industry's recognized standard values, set a worst value (lower limit) and an optimal value (upper limit) for each indicator. Step S303: For each indicator value of the current entrepreneurial entity, first compare it with the corresponding worst value (lower limit) and best value (upper limit), and map each original indicator value to between 0 and 1; Step S304: Arrange each value between 0 and 1 obtained after the above mapping into an ordered array according to a preset order. This array is the quantitative feature vector of the entrepreneurial entity.

[0009] Furthermore, in step S303, if the value of the indicator is lower than the worst value (lower limit), it is directly assigned the value of 0; if the value of the indicator is higher than the best value (upper limit), it is directly assigned the value of 1. If the value of the indicator is between the worst (lower limit) and the best (upper limit) value, it is proportionally converted into a decimal between 0 and 1 according to the relative position of the indicator value between the worst (lower limit) and the best (upper limit) value.

[0010] Furthermore, based on the relationship between project vision assessment, team vision assessment, and profitability status in historical entrepreneurial samples, a fusion indicator model can be constructed and applied, specifically as follows: Step 1: Collect a large number of historical entrepreneurial samples in similar or related industries, and obtain project vision assessments, team vision assessments, and corresponding profitability status of the historical entrepreneurial samples; Step 2: By controlling variables, obtain the relationship between project vision assessment and profitability, and the relationship between team vision assessment and profitability. Step 3: Based on the relationship determined in Step 2, fit and construct a fusion index model with project vision assessment and team vision assessment as inputs and profitability status as output. Step 4: Input the project vision assessment and team vision assessment obtained in step S2 into the fusion indicator model, and output the corresponding profitability status through the fusion indicator model to provide guidance for the target entrepreneurial entity to the current entrepreneurial project.

[0011] Furthermore, the profitability status refers to a normalized indicator that reflects whether profitable samples exist and the degree of profitability of profitable samples in historical entrepreneurial samples of the same or similar industries.

[0012] Furthermore, the fusion index model refers to a quantitative evaluation model that uses project vision assessment and team vision assessment as input variables and profitability status as the output variable. This model collects a large number of historical entrepreneurial samples from similar or related industries to obtain the project vision assessment, team vision assessment, and corresponding profitability status for each historical sample. It then uses a controlled variable method to determine the mapping relationships between project vision assessment and profitability status, and between team vision assessment and profitability status. Based on this, the two mapping relationships are comprehensively fitted to construct a fusion index model that can output a predicted profitability status value based on the two input vision assessment values. The profitability status output by this model is a normalized numerical indicator that can simultaneously reflect whether profitable samples exist in the historical samples and the degree of profitability of profitable samples.

[0013] A big data-based entrepreneurship guidance system includes: The data acquisition module is used to collect multi-source data information of the target entrepreneurial entity's current entrepreneurial project. The multi-source data information includes structured data and unstructured data. The structured data includes financial data, operational data, and market data. The unstructured data includes business plans and team vision surveys. The semantic analysis module is used to extract semantic features from the business plan to obtain a project vision assessment, and to analyze the team vision survey to obtain a team vision assessment. The feature transformation module is used to convert the structured data into quantized feature vectors; The profitability assessment module is used to assess profitability based on the project vision assessment and the team vision assessment. The matching analysis module is used to analyze the matching degree between the target entrepreneurial entity and each guidance strategy in the preset entrepreneurial guidance strategy library based on the quantitative feature vector; The guidance output module is used to provide guidance information on the current entrepreneurial project to the target entrepreneurial entity based on the profitability status assessed by the profitability assessment module and the guidance strategy matched by the matching analysis module.

[0014] This invention provides a big data-based entrepreneurship guidance system and method. Compared with existing technologies, the advantages of this method are: 1. This invention introduces a pre-built project vision keyword library when extracting semantic features from business plans, and encodes the context of the keywords to distinguish between "common mentions" and "strong expressions". Then, it combines the extreme scores of historical samples from similar industries for linear scaling mapping, so that the highly subjective element of project vision can be objectively quantified under a unified scale. This eliminates the bias caused by different text expression styles and lengths between different industries, and provides reliable basic data for subsequent evaluation.

[0015] 2. When assessing the consistency of a team's vision, this invention requires each member to anonymously write a statement of long-term goals and convert it into a semantic numerical vector. By calculating the cosine similarity between all pairs of members and taking the average, it effectively avoids the interference of social expectation bias on the authenticity of data in traditional questionnaire surveys. It can truly and objectively reflect the actual degree of consensus within the team on the project direction, thereby helping entrepreneurial entities to discover team disagreements early and take targeted adjustments.

[0016] 3. This invention sets worst and best values ​​for structured indicators such as finance, operations, and market based on historical databases or industry-recognized standard values. Through piecewise linear mapping, all original indicators are uniformly normalized to the 0 to 1 range and then arranged into a quantitative feature vector in a preset order. This makes the operating data from different enterprises, different development stages, and even different industries horizontally comparable, creating a standardized input space for subsequent strategy matching.

[0017] 4. The fusion index model constructed in this invention takes project vision assessment and team vision assessment as inputs and normalized profitability status, which simultaneously reflects "profitability" and "profitability level," as output. By using the control variable method, the relationship between each vision assessment and profitability status is fitted, and finally a nonlinear mapping function is formed. This model can quantitatively predict the profitability expectations of current entrepreneurial entities based on historical sample patterns, providing a clear numerical reference for entrepreneurial decisions. Attached Figure Description

[0018] Figure 1 This is a flowchart of Embodiment 1 of the present invention; Figure 2 This is a diagram showing the relationship between project vision assessment and team vision assessment in this invention. Figure 3 This is a flowchart of Embodiment 2 of the present invention. Detailed Implementation

[0019] To make the technical solution of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] Example 1

[0021] like Figure 1As shown, according to one aspect of the present invention, a big data-based entrepreneurship guidance method is provided, including collecting multi-source data information of the current entrepreneurial project of the target entrepreneurial entity. The multi-source data information includes structured data and unstructured data. The structured data includes financial data, operational data, and market data. The unstructured data includes business plans and team vision surveys. Semantic feature extraction is performed on business plans to obtain project vision assessment, and team vision surveys are analyzed to obtain team vision assessment. In this embodiment, the specific steps for extracting semantic features from a business plan to obtain a project vision assessment are as follows: Step 1: Extract paragraphs or sentences from the target entrepreneur's business plan, project roadshow transcript, in-depth interview records with the founder, etc., that describe the "significance of the project", "social value", "long-term vision" and "core problem to be solved", as raw text materials for analyzing the characteristics of the sense of mission.

[0022] Step 2: Use a pre-built project vision keyword library (containing words such as "change", "benefit", "breakthrough", "responsibility", "mission" etc.) to match the text and record the location and frequency of each keyword.

[0023] Step 3: Encode the context of each project vision keyword and assess the importance of the keyword in its specific context. For example, a word is more significant if it appears in the center sentence of a paragraph or is repeatedly emphasized by the founder. This step distinguishes between "ordinary mentions" and "strong expressions."

[0024] Step 4: Multiply the frequency of each keyword by its contextual contribution, and sum the results of all keywords to obtain a comprehensive raw score; the higher the score, the stronger the sense of mission conveyed by the text as a whole, and the more prominent the entrepreneurial entity's identification with and willingness to drive the project.

[0025] Step 5: Collect a large number of historical entrepreneurial samples from similar or related industries, calculate their original project vision scores in the same way, and statistically determine the highest and lowest historical scores. Step 6: Compare the original score of the current target subject with the historical range, and map it to a value between 0 and 1 by linear scaling. This is the project vision assessment.

[0026] Therefore, extract keywords related to "mission" (such as "change," "solution," "vision," and "mission") from the business plan or founder's statement, and statistically analyze their frequency and contextual strength (e.g., using TF-IDF values ​​or attention weights) to obtain the raw mission strength index 's'. Set an empirical maximum value. (This can be obtained from historical data statistics) and mapped accordingly, specifically: ; in and The sense of mission intensity is statistically determined based on a large number of historical samples, ensuring that the output falls within [0,1].

[0027] In this embodiment, the specific steps for analyzing the team vision survey to obtain a team vision assessment are as follows: Step 1: Obtain anonymous written statements from each member of the startup team regarding the project's long-term goals. The statements should focus on "what we hope to develop the project into" or "what problem we ultimately want to solve," ensuring that the texts accurately reflect each individual's core expectations for the team's future. Step 2: Convert each vision description text into a fixed-length numerical vector. This numerical vector can abstractly represent the semantic meaning of the text, so that the more similar the content of the description, the closer the direction of their corresponding vectors in space. Step 3: Compare the directions of the vision vectors of any two members to obtain the degree of similarity in their directions; Step 4: For all possible pairwise combinations within the team, complete the directional similarity assessment obtained in Step 3, and obtain the average directional similarity score. This average score serves as a quantitative indicator of the team's overall vision consistency.

[0028] Therefore, the semantic embedding vectors of the text describing the visions of each team member are... ( (where N is the number of team members), calculate the cosine similarity between each pair of teams and map it to [0,1]. The specific calculation is as follows: ; in ; Since the original cosine similarity value ranges from [-1,1], it is mapped to [0,1] by (cosine+1) / 2, which ultimately makes g also fall within [0,1].

[0029] The "vision" element, which is the most subjective element in entrepreneurship guidance, has been systematically and objectively quantified.

[0030] In terms of project vision assessment, this method does not simply count the number of times keywords appear. Instead, it performs text matching through a pre-built project vision keyword library, and then encodes the context of each keyword to assess its contextual importance. For example, it distinguishes between "ordinary mentions" and "strong expressions." The frequency of occurrence is then multiplied by the contextual contribution and summed to obtain a comprehensive raw score. Finally, by collecting a large number of historical samples from the same industry, the extreme values ​​of the scores are statistically analyzed, and the raw scores are mapped to the 0 to 1 range in a linear scaling manner, thereby eliminating the influence of differences in text expression styles between different industries.

[0031] In assessing team vision consistency, this method requires each member to anonymously write a statement of the project's long-term goals. Each vision description is then converted into a fixed-length semantic numerical vector, making semantically similar descriptions converge in the vector space. The cosine similarity of all pairwise combinations of members is then calculated, and the average is taken as a quantitative indicator of consistency. This approach effectively avoids the interference of social expectation bias on the authenticity of data in traditional questionnaire surveys and can objectively reflect the actual degree of consensus within the team on the project direction.

[0032] Example 2

[0033] like Figure 2 As shown, according to one aspect of the present invention, a big data-based entrepreneurship guidance method is provided, including converting structured data into quantitative feature vectors and assessing profitability based on project vision assessment and team vision assessment. In this embodiment, the specific steps for converting structured data into quantized feature vectors are as follows: Step S301: Extract multiple measurable raw indicators from the financial system, business ledgers, and market research reports; Step S302: Based on the data distribution of a large number of similar entrepreneurial entities in the historical database, or by referring to the industry's recognized standard values, set a worst value (lower limit) and an optimal value (upper limit) for each indicator. Step S303: For each indicator value of the current entrepreneurial entity, first compare it with the corresponding worst-case (lower limit) and best-case (upper limit) values. If the indicator value is lower than the worst-case (lower limit), assign it a value of 0 directly; if the indicator value is higher than the best-case (upper limit), assign it a value of 1 directly; if the indicator value is between the worst-case (lower limit) and best-case (upper limit), convert it proportionally to a decimal between 0 and 1 according to its relative position between the worst-case (lower limit) and best-case (upper limit). Then map each original indicator value to a value between 0 and 1. Step S304: Arrange each value between 0 and 1 obtained after the above mapping into an ordered array according to a preset order. This array is the quantitative feature vector of the entrepreneurial entity.

[0034] In this embodiment, based on the relationship between project vision assessment, team vision assessment, and profitability status in historical entrepreneurial samples, a fusion indicator model can be constructed and applied as follows: Step 1: Collect a large number of historical entrepreneurial samples in similar or related industries, and obtain project vision assessments, team vision assessments, and corresponding profitability status of the historical entrepreneurial samples; Among them, profitability status refers to a normalized indicator used to quantify the expected profitability level of a target entrepreneurial entity relative to a historical profit and loss benchmark within the historical entrepreneurial sample space of similar or related industries; this indicator also reflects whether there are profitable samples in the historical sample and the degree of profitability of profitable samples. Its acquisition method is as follows: 1) Collect a large number of historical entrepreneurial samples from similar or related industries, and extract the profitability index value of each historical entrepreneurial sample. Profitability indicators are any one of net profit margin, return on total assets, or return on equity, and are real numbers. Positive values ​​indicate profitability, while negative values ​​or zero indicate no profit or being at the break-even point.

[0035] 2) Calculate the maximum value of the profitability indicator among all historical entrepreneurial samples, and denote it as... and ensure ;by As the break-even point.

[0036] 3) For the first For each historical entrepreneurial example, calculate its corresponding profitability using the following formula. :

[0037] in, .

[0038] 4) This yields a set of profitability status labels for all historical entrepreneurial samples; when When the value is zero, it indicates that the sample is not profitable or is in a loss-making state, and its profit status is assigned a value of 0; when the value is zero, it indicates that the sample is not profitable or is in a loss-making state. When this occurs, it indicates that the sample is profitable, and Follow It increases linearly with the increase of , when hour This allows for a unified expression of both "whether a profit has been made" and "the degree of profitability" within the same normalized indicator.

[0039] In the model application phase, the project vision assessment and team vision assessment of the target entrepreneurial entity obtained in step 2 are input into the fusion index model. Based on the mapping relationship fitted from historical samples, the fusion index model outputs the profit status prediction value of the target entrepreneurial entity. This prediction value is also between 0 and 1, which is used to guide the target entrepreneurial entity to judge and make decisions on the profit expectations of the current entrepreneurial project.

[0040] Step 2: By controlling variables, obtain the relationship between project vision assessment and profitability, and the relationship between team vision assessment and profitability. Then, the obtained profitability statuses are arranged in ascending order using an arithmetic progression, and the corresponding project vision assessment and team vision assessment are obtained. Next, the relationship between profitability status and project vision assessment is obtained by controlling the variable of team vision assessment (e.g., all team vision assessments are set to 0.45, generally taking the most frequent value). The relationship between profitability status and project vision assessment is shown in Table 1. Table 1. Profitability Status and Project Vision Assessment

[0041] Therefore, the mathematical expression generated by fitting the data in Table 1 is: ; In the above formula, Represents the mathematical expression generated by the fitting; where , All are used to ensure the formula Parameters that approximate the profit state. Furthermore, based on the information in Table 1, it can be determined that when... , When, formula The profitability status is similar to that in Table 1.

[0042] Then, by controlling for the project vision assessment variable (e.g., all project vision assessment values ​​are 0.45), the relationship between profitability and team vision assessment was obtained. The relationship between profitability and team vision assessment is shown in Table 2. Table 2 Profitability Status and Team Vision Assessment

[0043] Therefore, the mathematical expression generated by fitting the data in Table 2 is: ; In the above formula, Represents the mathematical expression generated by the fitting; where Used to ensure formula Parameters that approximate the profit state. Furthermore, based on the information in Table 2, it can be determined that when... When, formula The profitability status is similar to that in Table 2.

[0044] Similarly, based on the mathematical expressions fitted from Tables 1 and 2, we can further determine the relationship between project vision assessment, team vision assessment, and profitability. Therefore: ; In the above formula, This indicates the predicted profitability of the target entrepreneurial entity. Used to ensure formula Parameters that approximate a profitable state. Among them, based on... Figure 2 The information in the middle can be determined when hour, Figure 2 Formulas in It is approaching a profitable state.

[0045] Step 3: Based on the relationship determined in Step 2, fit and construct a fusion index model with project vision assessment and team vision assessment as inputs and profitability status as output. Step 4: Input the project vision assessment and team vision assessment obtained in step S2 into the fusion indicator model, and output the corresponding profitability status through the fusion indicator model to provide guidance for the target entrepreneurial entity to the current entrepreneurial project.

[0046] Finally, based on quantitative feature vector analysis, the matching degree between the target entrepreneurial entity and each guidance strategy in the pre-set entrepreneurial guidance strategy library is analyzed; and based on the profitability and guidance strategy, the target entrepreneurial entity can be provided with guidance for the current entrepreneurial project.

[0047] In terms of quantifying structured data, this method extracts multiple measurable raw indicators from financial systems, business ledgers, and market research reports. Based on historical databases or industry standard values, it sets the worst and best values ​​for each indicator. Indicators below the worst value are assigned a value of 0, those above the best value are assigned a value of 1, and those in between are converted to decimals between 0 and 1 proportionally according to their relative positions. Finally, they are arranged into a quantified feature vector in a preset order. This normalization processing method based on industry benchmarks makes business data across enterprises and stages comparable.

[0048] In terms of profitability assessment, this method innovatively proposes a fusion index model: First, a large number of historical entrepreneurial samples from the same industry are collected, and any one of net profit margin, return on total assets, or return on net assets is used as a profitability indicator. The maximum value is calculated and 0 is used as the profit and loss dividing point, mapping the profitability indicator to a profitability status value between 0 and 1. Then, the relationship between project vision assessment and profitability status, and between team vision assessment and profitability status are fitted separately by the control variable method. Finally, a fusion index model is constructed with the two vision assessments as input and profitability status as output. This design allows the dual information of "whether it is profitable" and "the degree of profitability" to be uniformly expressed in the same normalized index, providing a quantitative basis for the accurate matching of subsequent entrepreneurial guidance strategies.

[0049] Example 3

[0050] like Figure 1 As shown, according to one aspect of the present invention, a big data-based entrepreneurship guidance system is provided, comprising: The data acquisition module is used to collect multi-source data information on the current entrepreneurial project of the target entrepreneurial entity. The multi-source data information includes structured data and unstructured data. The structured data includes financial data, operational data, and market data, while the unstructured data includes business plans and team vision surveys. The semantic analysis module is used to extract semantic features from business plans to obtain project vision assessment, and to analyze team vision surveys to obtain team vision assessment. The feature transformation module is used to convert structured data into quantized feature vectors; The profitability assessment module is used to evaluate profitability based on project vision assessment and team vision assessment. The matching analysis module is used to analyze the matching degree between the target entrepreneurial entity and each guidance strategy in the preset entrepreneurial guidance strategy library based on quantitative feature vectors. The guidance output module is used to provide guidance information on the current entrepreneurial project to the target entrepreneurial entity based on the profitability status assessed by the profitability assessment module and the guidance strategies matched by the matching analysis module.

[0051] The system achieves end-to-end automation from data acquisition to guidance output through modular design. It comprises six core components: data acquisition module, semantic analysis module, feature conversion module, profit assessment module, matching analysis module, and guidance output module. The semantic analysis module is responsible for the calculation tasks of project vision assessment and team vision assessment in Example 1. The feature conversion module and the profit assessment module work together to complete the quantitative feature vector generation and fusion index model prediction in Example 2. The matching analysis module performs matching degree analysis based on the quantitative feature vector and the preset entrepreneurship guidance strategy library.

[0052] The system's technological advancement lies in its organic integration of various technologies such as natural language processing, semantic vectorization, industry benchmark normalization, and fusion model prediction, forming a complete closed loop from mixed input of unstructured text and structured data to output of personalized guidance strategies. This effectively solves the pain points of traditional entrepreneurship guidance, such as subjective evaluation standards and generalized guidance strategies.

[0053] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A startup guidance method based on big data, characterized in that, Includes the following steps: Step S1: Collect multi-source data information of the target entrepreneurial entity's current entrepreneurial project. The multi-source data information includes structured data and unstructured data. The structured data includes financial data, operating data, and market data. The unstructured data includes business plans and team vision surveys. Step S2: Extract semantic features from the business plan to obtain a project vision assessment, and analyze the team vision survey to obtain a team vision assessment; Step S3: Convert the structured data into a quantitative feature vector, and evaluate the profitability based on the project vision assessment and team vision assessment. Step S4: Analyze the matching degree between the target entrepreneurial entity and each guidance strategy in the preset entrepreneurial guidance strategy library based on the quantitative feature vector; Step S5: Based on the profitability status assessed in Step S3 and the matching guidance strategies in Step S4, guidance can be provided to the target entrepreneurial entities for their current entrepreneurial projects.

2. The entrepreneurship guidance system and method based on big data according to claim 1, characterized in that: In step S2, the specific steps for extracting semantic features from the business plan to obtain a project vision assessment are as follows: Step 1: Extract paragraphs or sentences describing the project vision from the target entrepreneur's business plan as raw text material for analyzing the project vision; Step 2: Use a pre-built project vision keyword library to match the text and record the location and frequency of each keyword. Step 3: Encode the context of each project vision keyword and assess the importance of the keyword in its specific context; Step 4: Multiply the frequency of each keyword by its contextual contribution, and sum the results of all keywords to obtain a comprehensive raw score; Step 5: Collect a large number of historical entrepreneurial samples from similar or related industries, calculate their original project vision scores in the same way, and statistically determine the highest and lowest historical scores. Step 6: Compare the original score of the current target subject with the historical range, and map it to a value between 0 and 1 by linear scaling. This is the project vision assessment.

3. The entrepreneurship guidance system and method based on big data according to claim 1, characterized in that: In step S2, the specific steps for analyzing the team vision survey to obtain a team vision assessment are as follows: Step 1: Obtain anonymous written statements from each member of the startup team regarding the project's long-term goals; Step 2: Convert each vision description text into a fixed-length numerical vector, which can abstractly represent the semantic meaning of the text; Step 3: Compare the directions of the vision vectors of any two members to obtain the degree of similarity in their directions; Step 4: For all possible pairwise combinations within the team, complete the directional similarity assessment obtained in Step 3, and obtain the average directional similarity score. This average score serves as a quantitative indicator of the team's overall vision consistency.

4. The entrepreneurship guidance system and method based on big data according to claim 1, characterized in that: In step S3, the specific steps for converting structured data into quantized feature vectors are as follows: Step S301: Extract multiple measurable raw indicators from the financial system, business ledgers, and market research reports; Step S302: Based on the data distribution of a large number of similar entrepreneurial entities in the historical database, or by referring to the industry-recognized standard values, set a worst value and an optimal value for each indicator. Step S303: For each indicator value of the current entrepreneurial entity, first compare it with the corresponding worst and best values, and map each original indicator value to a range of 0 to 1; Step S304: Arrange each value between 0 and 1 obtained after the above mapping into an ordered array according to a preset order. This array is the quantitative feature vector of the entrepreneurial entity.

5. The entrepreneurship guidance system and method based on big data according to claim 4, characterized in that: In step S303, if the value of the indicator is lower than the worst value, it is directly assigned the value of 0; if the value of the indicator is higher than the best value, it is directly assigned the value of 1. If the value of the indicator is between the worst and best values, it is proportionally converted into a decimal between 0 and 1 according to the relative position of the indicator value between the worst and best values.

6. The entrepreneurship guidance system and method based on big data according to claim 1, characterized in that: Based on the relationship between project vision assessment, team vision assessment, and profitability in historical entrepreneurial samples, a fusion indicator model can be constructed and applied as follows: Step 1: Collect a large number of historical entrepreneurial samples in similar or related industries, and obtain project vision assessments, team vision assessments, and corresponding profitability status of the historical entrepreneurial samples; Step 2: By controlling variables, obtain the relationship between project vision assessment and profitability, and the relationship between team vision assessment and profitability. Step 3: Based on the relationship determined in Step 2, fit and construct a fusion index model with project vision assessment and team vision assessment as inputs and profitability status as output. Step 4: Input the project vision assessment and team vision assessment obtained in step S2 into the fusion indicator model, and output the corresponding profitability status through the fusion indicator model to provide guidance for the target entrepreneurial entity to the current entrepreneurial project.

7. The entrepreneurship guidance system and method based on big data according to claim 6, characterized in that: The profitability status refers to a normalized indicator that reflects whether there are profitable samples and the degree of profitability of profitable samples in historical entrepreneurial samples of the same or similar industries.

8. The entrepreneurship guidance system and method based on big data according to claim 6, characterized in that: The fusion index model collects a large number of historical entrepreneurial samples from similar or related industries to obtain project vision assessment, team vision assessment, and corresponding profitability status for each historical sample. It then uses the control variable method to determine the mapping relationship between project vision assessment and profitability status, and between team vision assessment and profitability status.

9. A big data-based entrepreneurship guidance system, characterized in that, The entrepreneurship guidance system based on big data, applicable to any one of claims 1-8, comprises: a data acquisition module, a semantic analysis module, a feature conversion module, a profit assessment module, a matching analysis module, and a guidance output module.

Images