Innovation and entrepreneurship guidance recommendation system and method based on knowledge graph and language model

Abstract

The invention relates to the technical field of innovation and entrepreneurship guidance recommendation, in particular to an innovation and entrepreneurship guidance recommendation system and method based on a knowledge graph and a language model. The system comprises a subjective evaluation atlas unit and an objective evaluation atlas unit. A subjective self-evaluation knowledge graph is established through the subjective evaluation graph unit and used for collecting self-evaluation of subjective persons, a relation network of the subjective persons is established through the objective evaluation graph unit, objective persons are determined, and objective evaluation content of the objective persons on subjective evaluation items is collected. The objective evaluation knowledge graph is established, the accuracy of subjective evaluation items is verified, the evaluation diversity of subjectors in the entrepreneurship guidance recommendation process is increased, the subjectors are helped to clearly know own advantages and defects, meanwhile, the objective evaluation knowledge graph is utilized to verify the accuracy of the subjective evaluation items from the side, and the subjective evaluation recommendation is more accurate. And the success rate of later matching of suitable entrepreneurship projects is increased.

Description

Technical Field

[0001] The present invention relates to the technical field of innovation and entrepreneurship guidance recommendation, and specifically, to an innovation and entrepreneurship guidance recommendation system and method based on knowledge graph and language model. Background Art

[0002] In the wave of innovation and entrepreneurship, college students, as a young force full of vitality and creativity, are gradually becoming important participants in the entrepreneurial field.

[0003] However, most college students have not been exposed to entrepreneurial projects during their college years, and are confused about the direction and content of entrepreneurship, and are unable to clearly recognize their own advantages and disadvantages, which greatly reduces the success rate of college students' entrepreneurship. Therefore, in order to solve the above problems, a college student innovation and entrepreneurship plan evaluation and analysis system with publication number CN109918482A discloses the following content: Business plan entry module, used to enter business plans in a fill-in-the-blank manner based on a preset template; The business plan dissection module dissects the business plan based on the content entered by the user in each space in the preset template; The business plan retrieval module mines the web page data related to each piece of content based on the dissected business plan through the web crawler module, and organizes and displays them in a table format; The business plan comparison module is used to compare the input business plan data with the corresponding comparison data, and output the corresponding similarity according to the comparison result; The visual analysis module realizes visual analysis of data based on the input business plan data through Tableau Desktop.

[0004] The effect is: it has built-in related web page data query function and data similarity comparison function, which greatly reduces the workload of staff.

[0005] However, during the process of entering business plans, the data entered are all subjective evaluations of the subject (college students). As mentioned in the previous article, most college students are confused about themselves and cannot clearly understand their own strengths and weaknesses. As a result, during the data entry process, most of the data are not representative and most of them are the visions of the subject. It is impossible to make an objective and fair self-evaluation, which leads to deviations in the subsequent data comparison, which greatly reduces the accuracy of business recommendations and matching.

[0006] In order to address the above problems, there is an urgent need for innovation and entrepreneurship guidance recommendation systems and methods based on knowledge graphs and language models. Summary of the invention

[0007] The purpose of the present invention is to provide an innovation and entrepreneurship guidance recommendation system and method based on knowledge graph and language model to solve the problems raised in the above background technology.

[0008] To achieve the above objectives, one of the objectives of the present invention is to provide an innovation and entrepreneurship guidance recommendation system based on knowledge graph and language model, including a knowledge graph expansion module, a large language model establishment module and an entrepreneurial project matching module, wherein; The knowledge graph expansion module includes a subjective evaluation graph unit and an objective evaluation graph unit. The subjective evaluation graph unit is used to collect self-evaluation information of the subject, and classify it according to the content of the self-evaluation information, obtain various subjective evaluation items, and establish a self-evaluation knowledge graph of the subject. The objective evaluation graph unit is used to establish a relationship network of the subject, determine each objective person, and collect the objective evaluation content of each objective person on the subjective evaluation item, establish an objective person evaluation knowledge graph, and verify the accuracy of the subjective evaluation item; The large language model building module is used to build a large language model, capture self-evaluation information and objective evaluation content, obtain content related to subjective evaluation items, and eliminate irrelevant information; The entrepreneurial project matching module combines the objective evaluation knowledge graph and the self-evaluation knowledge graph, and compares them with the objective evaluation knowledge graph and the self-evaluation knowledge graph of historical entrepreneurs to obtain the entrepreneurial trends of the subject and match the subject with a suitable entrepreneurial project.

[0009] As a further improvement of the present technical solution, the subjective evaluation items in the subjective evaluation map unit include source of funds, subjective expertise, family conditions, work style, school performance, personality orientation and interests and hobbies, and the objective persons in the objective evaluation map unit include teaching teachers, classmates, internship institutions and organizations.

[0010] As a further improvement of the technical solution, the method for establishing a subject self-evaluation knowledge graph in the subjective evaluation graph unit includes the following steps: Step 1: Determine the tendency of each subjective evaluation item, mark the subjective evaluation items proved by direct facts as direct subjective evaluation items, and mark the subjective evaluation items proved by language description as indirect subjective evaluation items; Step 2: Establish a scoring method for indirect subjective evaluation items, match the indirect subjective evaluation item type of the subject, and obtain the subjective evaluation score ; Step 3: Collect direct factual evidence of each direct subjective evaluation item, and bind and generate proof materials for the direct subjective evaluation items.

[0011] As a further improvement of the technical solution, the method for the objective evaluation graph unit to establish an objective person evaluation knowledge graph includes the following steps: The first step is to combine the scoring method of indirect subjective evaluation items to obtain the objective evaluation scores of each objective person for each indirect subjective evaluation item. ; Step 2: Obtain the objective evaluation scores of different objective persons for the same type of indirect subjective evaluation items , calculate the average objective evaluation score of the same category in to is the objective evaluation score of different objective persons for each indirect subjective evaluation item of the same type, and n is the number of objective persons; Step 3: Compare the average objective evaluation scores Subjective evaluation score , get the evaluation difference ; Step 4: Establish the evaluation difference threshold for the indirect subjective evaluation project , and make indirect subjective evaluation items elimination judgment; When evaluating the difference ≤ Poor evaluation threshold When Subjective evaluation score , take the two average values ​​as the final evaluation score; When evaluating the difference >Poor evaluation threshold When doing so, this type of indirect subjective evaluation item is excluded and does not serve as the basis for the subjective person's evaluation.

[0012] As a further improvement of the present technical solution, the knowledge graph expansion module also includes a reliability analysis unit, which obtains feedback information from historical entrepreneurs on subjective evaluation items, establishes objective reliability judgment rules, and determines the evaluation status of the objective person.

[0013] As a further improvement of the technical solution, the method for the reliability analysis unit to establish objective reliability evaluation rules includes the following steps: The first step is to collect the types of entrepreneurial projects of historical subjects and obtain the relevant objective persons of historical entrepreneurial projects; The second step is to combine the scoring method of indirect subjective evaluation items to obtain the historical evaluation scores of each related objective person for each indirect subjective evaluation item , calculate the average historical evaluation score ; The third step is to combine the historical evaluation score average , get the average historical evaluation score Corresponding objective evaluation scores , calculate the reliability difference , set the reliability threshold ; Step 4: Set the objective evaluation score and the unit decrement score; When the reliability difference ≤Reliability Threshold , then the objective person's evaluation score is maintained; When the reliability difference >Reliability threshold , the objective person's objective evaluation score will be reduced by the unit decreasing score until the objective person's objective evaluation score is reset to zero, and the objective person's evaluation power will be eliminated.

[0014] As a further improvement of the present technical solution, the large language model establishment module includes a keyword extraction unit and an evaluation tendency acquisition unit. The keyword extraction unit is used to obtain the evaluation content of the subjective and objective people for indirect subjective evaluation items, and establish a keyword evaluation database to obtain the evaluation tendency of each keyword match. The evaluation tendency acquisition unit combines the keyword evaluation database and the evaluation content to obtain the evaluation tendency of the subjective and objective people for different subjective evaluation items.

[0015] The second object of the present invention is to provide a method for using an innovation and entrepreneurship guidance recommendation system based on a knowledge graph and a language model, comprising the following method steps: S1. The subjective evaluation graph unit collects the self-evaluation information of the subject, classifies it according to the content of the self-evaluation information, obtains various subjective evaluation items, and establishes the subject's self-evaluation knowledge graph; S2. The objective evaluation graph unit establishes a relationship network of the subject, identifies each subject, and collects the objective evaluation content of each subject on the subjective evaluation project, establishes the objective evaluation knowledge graph, and verifies the accuracy of the subjective evaluation project; S3, a large language model building module builds a large language model to capture self-evaluation information and objective evaluation content, obtain content related to subjective evaluation items, and eliminate irrelevant information; S4. The entrepreneurial project matching module combines the objective evaluation knowledge graph and the self-evaluation knowledge graph, and compares them with the objective evaluation knowledge graph and the self-evaluation knowledge graph of historical entrepreneurs to obtain the subjective entrepreneurial trend and match the subjective with suitable entrepreneurial projects.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. In the innovation and entrepreneurship guidance recommendation system and method based on knowledge graph and language model, a subject self-evaluation knowledge graph is established through a subjective evaluation graph unit to collect the subject's self-evaluation of himself, and a relationship network of the subject is established through an objective evaluation graph unit to determine each objective subject, and collect the objective evaluation content of each objective subject on the subjective evaluation project, establish an objective subject evaluation knowledge graph, and verify the accuracy of the subjective evaluation project, which not only increases the evaluation diversity of the subject in the process of entrepreneurship guidance recommendation, helps the subject to more clearly understand his own strengths and weaknesses, but also uses the objective subject evaluation knowledge graph to verify the accuracy of the subjective evaluation project from the side, and increases the success rate of matching suitable entrepreneurial projects in the later stage.

[0017] 2. In the innovation and entrepreneurship guidance recommendation system and method based on knowledge graph and language model, the feedback information of historical entrepreneurs on subjective evaluation items is obtained through the reliability analysis unit, the reliability judgment rules of the objective person are established, the evaluation status of the objective person is determined, the reliability of the objective person's evaluation is improved, and the evaluation accuracy of the subjective evaluation items is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is the overall system structure block diagram of the present invention; Figure 2 It is a flowchart of the overall method of the present invention.

[0019] The meaning of each number in the figure is: 10. Knowledge graph expansion module; 110. Subjective evaluation graph unit; 120. Objective evaluation graph unit; 130. Reliability analysis unit; 20. Large language model building module; 210. Keyword extraction unit; 220. Evaluation tendency acquisition unit; 30. Entrepreneurship project matching module. DETAILED DESCRIPTION

[0020] The following will be combined with the accompanying drawings in the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] See also Figure 1 As shown, one of the purposes of the present invention is to provide an innovation and entrepreneurship guidance recommendation system based on knowledge graph and language model, including a knowledge graph expansion module 10, a large language model establishment module 20 and an entrepreneurial project matching module 30, wherein; The knowledge graph expansion module 10 includes a subjective evaluation graph unit 110 and an objective evaluation graph unit 120. The subjective evaluation graph unit 110 is used to collect self-evaluation information of the subject, and classify it according to the content of the self-evaluation information, obtain various subjective evaluation items, and establish a self-evaluation knowledge graph of the subject. The objective evaluation graph unit 120 is used to establish a relationship network of the subject, determine each objective person, and collect the objective evaluation content of each objective person on the subjective evaluation item, establish the objective person evaluation knowledge graph, and verify the accuracy of the subjective evaluation item. The large language model building module 20 is used to build a large language model, capture self-evaluation information and objective evaluation content, obtain content related to subjective evaluation items, and eliminate irrelevant information; The entrepreneurial project matching module 30 combines the objective evaluation knowledge graph and the self-evaluation knowledge graph, and compares them with the objective evaluation knowledge graph and the self-evaluation knowledge graph of historical entrepreneurs to obtain the entrepreneurial trends of the subject and match the subject with a suitable entrepreneurial project.

[0022] When used specifically, in the process of guiding and recommending innovation and entrepreneurship for college students, first, it is necessary to collect the self-evaluation information of the subject through the subjective evaluation map unit 110, and classify it according to the content of the self-evaluation information, obtain various subjective evaluation items, and establish a subjective self-evaluation knowledge map. This step is to collect the self-evaluation of the subject by the subject, which is highly subjective. Therefore, in order to further determine the accuracy of various subjective evaluation items, it is necessary to establish a relationship network of the subject through the objective evaluation map unit 120, determine each objective person, and collect the objective evaluation content of each objective person on the subjective evaluation item, establish an objective evaluation knowledge map, and verify the accuracy of the subjective evaluation item, that is, the relationship network of the subject, such as the evaluation of the classmates and the objective evaluation of the teacher, can be used as a reference for verifying the accuracy of the subjective evaluation item; Since in the process of establishing the subjective self-evaluation knowledge graph and the objective evaluation knowledge graph, it is necessary to collect the evaluation content of the subjective and each objective person for the subjective evaluation project, and the information therein contains relevant information and irrelevant information, if it is identified manually, the workload is too large. In order to reduce the workload and improve the efficiency of extracting relevant information, it is necessary to establish a large language model through the large language model establishment module 20, capture the self-evaluation information and the objective evaluation content, obtain the content related to the subjective evaluation project, and eliminate irrelevant information. For example, in the process of making recommendations, it is necessary for the subjective and each relevant objective person to conduct a questionnaire survey, and obtain the filled-in data on the questionnaire survey as a reference for the subjective evaluation project; Finally, the entrepreneurial project matching module 30 combines the objective evaluation knowledge graph and the self-evaluation knowledge graph, and compares them with the objective evaluation knowledge graph and the self-evaluation knowledge graph of historical entrepreneurs to obtain the subjective entrepreneurial trend and match the subjective with a suitable entrepreneurial project.

[0023] In addition, the subjective evaluation items in the subjective evaluation map unit 110 include funding sources, subjective strengths, family conditions, work style, school performance, personality orientation and interests and hobbies, and the objective subjects in the objective evaluation map unit 120 include teaching teachers, classmates, internship institutions and organizations.

[0024] Furthermore, the method for establishing the subject self-evaluation knowledge graph in the subjective evaluation graph unit 110 includes the following steps: Step 1: Determine the tendency of each subjective evaluation item, mark the subjective evaluation items proved by direct facts as direct subjective evaluation items, and mark the subjective evaluation items proved by language description as indirect subjective evaluation items; Step 2: Establish a scoring method for indirect subjective evaluation items, match the indirect subjective evaluation item type of the subject, and obtain the subjective evaluation score ; Step 3: Collect direct factual evidence of each direct subjective evaluation item, and bind and generate proof materials for the direct subjective evaluation items.

[0025] In specific use, in the process of establishing the subject's self-evaluation knowledge graph, due to the different types of subjective evaluation items, it is necessary to classify the subjective evaluation items. First, it is necessary to judge the tendency of each subjective evaluation item, mark the subjective evaluation items proved by direct facts as direct subjective evaluation items, and mark the subjective evaluation items proved by language description as indirect subjective evaluation items. For example, family conditions, this subjective evaluation item can be directly determined through the subject's industry, residence, parents' income and parents' occupations. As for the style of doing things, it can only be described by the subject in words, and it cannot be directly determined even if it is proved by facts. At this time, it is necessary to establish an indirect subjective evaluation project scoring method to match the subject's indirect subjective evaluation The present invention mainly formulates questionnaires for various questions and scores the answers of the subjects to various questions. For example, a questionnaire on the personality orientation of the subjects involves the subject's ability to deal with emergencies, including four options: A. Direct processing; B. Asking others to assist in processing; C. Choosing to escape; D. No expression. Each option has a corresponding score, and finally the evaluation score of the indirect subjective evaluation project is obtained. For the direct subjective evaluation project, it is necessary to obtain the corresponding direct factual proof and bind the proof materials of the direct subjective evaluation project, such as school grades. The direct factual proof includes the test scores of various subjects, certificates obtained during school, and the results of various competitions.

[0026] Furthermore, the method for the objective evaluation graph unit 120 to establish the objective person evaluation knowledge graph includes the following steps: The first step is to combine the scoring method of indirect subjective evaluation items to obtain the objective evaluation scores of each objective person for each indirect subjective evaluation item. ; Step 2: Obtain the objective evaluation scores of different objective persons for the same type of indirect subjective evaluation items , calculate the average objective evaluation score of the same category ,in to is the objective evaluation score of different objective persons for each indirect subjective evaluation item of the same type, and n is the number of objective persons; Step 3: Compare the average objective evaluation scores Subjective evaluation score , get the evaluation difference ; Step 4: Establish the evaluation difference threshold for the indirect subjective evaluation project , and make indirect subjective evaluation items elimination judgment; When evaluating the difference ≤ Poor evaluation threshold When Subjective evaluation score , take the two average values ​​as the final evaluation score; When evaluating the difference >Poor evaluation threshold When doing so, this type of indirect subjective evaluation item is excluded and does not serve as the basis for the subjective person's evaluation.

[0027] In specific use, since the objective person has different perspectives when evaluating the subjective person, the corresponding evaluation accuracy is higher. Therefore, in the process of establishing the objective person evaluation knowledge graph, it is necessary to combine the indirect subjective evaluation project scoring method to obtain the objective evaluation score of each objective person for each indirect subjective evaluation project. , that is, each objective person conducts a secondary evaluation according to the corresponding indirect subjective evaluation item scoring method to obtain the objective evaluation score ,It is worth explaining that for direct subjective evaluation items, ,since there is direct factual proof, there is no need for secondary proof by an ,objective person; Complete objective evaluation score After obtaining the work, it is necessary to calculate the average objective evaluation score of the same category , by averaging the objective evaluation scores Subjective evaluation score Compare and obtain the evaluation difference , and when the evaluation difference ≤ Poor evaluation threshold , it means that the subjective evaluation is confirmed by the objective evaluation, and the average objective evaluation score is retained. Subjective evaluation score , take the two average values ​​as the final evaluation score; When evaluating the difference >Poor evaluation threshold When , it indicates that there is a large difference between the subjective evaluation and the objective evaluation at this time, and the validity of the subjective evaluation and the objective evaluation cannot be determined. This type of indirect subjective evaluation project needs to be eliminated and not used as the evaluation basis of the subject.

[0028] Specifically, the knowledge graph expansion module 10 also includes a reliability analysis unit 130, which obtains feedback information from historical entrepreneurs on subjective evaluation items, establishes objective reliability judgment rules, and determines the evaluation status of the objective person.

[0029] In addition, the method for the reliability analysis unit 130 to establish objective reliability evaluation rules includes the following steps: The first step is to collect the types of entrepreneurial projects of historical subjects and obtain the relevant objective persons of historical entrepreneurial projects; The second step is to combine the scoring method of indirect subjective evaluation items to obtain the historical evaluation scores of each related objective person for each indirect subjective evaluation item , calculate the average historical evaluation score ; The third step is to combine the historical evaluation score average , get the average historical evaluation score Corresponding objective evaluation scores , calculate the reliability difference , set the reliability threshold ; Step 4: Set the objective evaluation score and the unit decrement score; When the reliability difference ≤Reliability Threshold , then the objective person's evaluation score is maintained; When the reliability difference >Reliability threshold , the objective person's objective evaluation score will be reduced by the unit decreasing score until the objective person's objective evaluation score is reset to zero, and the objective person's evaluation power will be eliminated.

[0030] In specific use, since some subjective evaluation items are not fully covered during the evaluation of subjective evaluation items by each objective person, that is, the reliability of each objective person is different, in order to improve the reliability of the objective person's evaluation and the evaluation accuracy of the subjective evaluation items, it is necessary to obtain the feedback information of historical entrepreneurs on the subjective evaluation items through the reliability analysis unit 130, establish objective person reliability judgment rules, and determine the evaluation status of the objective person. First, it is necessary to collect the entrepreneurial project types of historical subjective persons, obtain the related objective persons of the historical entrepreneurial projects, that is, the entrepreneurs evaluated by the objective person, and the corresponding related objective persons include their company members and partners, etc., and combine the scoring method of indirect subjective evaluation items to obtain the historical evaluation scores of each related objective person for each indirect subjective evaluation item. , calculate the average historical evaluation score , still using questionnaires to obtain historical evaluations of indirect subjective evaluation projects, combined with the average historical evaluation scores , get the average historical evaluation score Corresponding objective evaluation scores , calculate the reliability difference , set the reliability threshold , the larger the difference, the lower the reliability of the historical evaluation of the objective person for the indirect subjective evaluation item, and vice versa. Finally, the objective person evaluation score and the unit decreasing score are set, that is, the initial objective person evaluation score of each objective person is the same; When the reliability difference ≤Reliability Threshold , then the objective person's evaluation score is maintained, indicating that this evaluation is reliable; When the reliability difference >Reliability threshold , the objective person's objective evaluation score will be reduced by the unit decreasing score until the objective person's objective evaluation score is reset to zero, and the objective person's evaluation power will be eliminated.

[0031] Furthermore, the large language model establishment module 20 includes a keyword extraction unit 210 and an evaluation tendency acquisition unit 220. The keyword extraction unit 210 is used to obtain the evaluation content of the subjective and objective persons for the indirect subjective evaluation items, and establish a keyword evaluation database to obtain the evaluation tendency of each keyword match. The evaluation tendency acquisition unit 220 combines the keyword evaluation database and the evaluation content to obtain the evaluation tendency of the subjective and objective persons for different subjective evaluation items.

[0032] In specific use, during the subjective evaluation project, due to the different language description methods of different groups of people, different sentences are used for the same tendency conditions. For example, to evaluate the subjective person's strong ability to do things, the evaluation sentence of an objective person is "high efficiency", while for another objective person, when it comes to the subjective evaluation project, the evaluation sentence he uses is "efficiency in doing things". The two have the same meaning, but the evaluation sentences used are different. Therefore, it is necessary to obtain the evaluation content of the subjective and objective persons for the indirect subjective evaluation project through the keyword extraction unit 210, and establish a keyword evaluation database to obtain the evaluation tendency of each keyword match. When the same type of keywords appear, it means that the evaluation score of the subjective evaluation project is the same. Finally, the evaluation tendency acquisition unit 220 combines the keyword evaluation database and the evaluation content to obtain the evaluation tendency of the subjective and objective persons for different subjective evaluation projects.

[0033] Depend on Figure 2 As shown, the second object of the present invention is to provide a method for using an innovation and entrepreneurship guidance recommendation system based on a knowledge graph and a language model, comprising the following method steps: S1. The subjective evaluation graph unit 110 collects the self-evaluation information of the subject, and classifies it according to the content of the self-evaluation information, obtains various subjective evaluation items, and establishes the subject's self-evaluation knowledge graph; S2, the objective evaluation graph unit 120 establishes a relationship network of the subject, identifies each subject, and collects the objective evaluation content of each subject on the subjective evaluation project, establishes the objective evaluation knowledge graph, and verifies the accuracy of the subjective evaluation project; S3, the large language model building module 20 builds a large language model to capture self-evaluation information and objective evaluation content, obtains content related to subjective evaluation items, and removes irrelevant information; S4. The entrepreneurial project matching module 30 combines the objective evaluation knowledge graph and the self-evaluation knowledge graph, and compares them with the objective evaluation knowledge graph and the self-evaluation knowledge graph of historical entrepreneurs to obtain the entrepreneurial trend of the subject and match the subject with a suitable entrepreneurial project.

[0034] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. An innovation and entrepreneurship guidance recommendation system based on knowledge graph and language model, characterized by: It includes a knowledge graph expansion module (10), a large language model building module (20) and an entrepreneurial project matching module (30), wherein; The knowledge graph expansion module (10) comprises a subjective evaluation graph unit (110) and an objective evaluation graph unit (120), wherein the subjective evaluation graph unit (110) is used to collect self-evaluation information of a subject, and classify the self-evaluation information according to the content, obtain various subjective evaluation items, and establish a subject self-evaluation knowledge graph, and the objective evaluation graph unit (120) is used to establish a relationship network of the subject, identify each objective subject, and collect the objective evaluation content of each objective subject on the subjective evaluation item, establish the objective subject evaluation knowledge graph, and verify the accuracy of the subjective evaluation item; The large language model building module (20) is used to build a large language model, capture self-evaluation information and objective evaluation content, obtain content related to subjective evaluation items, and eliminate irrelevant information; The entrepreneurial project matching module (30) combines the objective evaluation knowledge graph and the self-evaluation knowledge graph, and compares them with the objective evaluation knowledge graph and the self-evaluation knowledge graph of historical entrepreneurs to obtain the entrepreneurial trend of the subject and match a suitable entrepreneurial project for the subject.

2. The innovation and entrepreneurship guidance recommendation system based on knowledge graph and language model according to claim 1 is characterized by: The subjective evaluation items in the subjective evaluation map unit (110) include funding sources, subjective strengths, family conditions, work style, academic performance, personality orientation, and interests and hobbies, and the objective persons in the objective evaluation map unit (120) include teaching teachers, classmates, internship institutions, and organizations.

3. The innovation and entrepreneurship guidance recommendation system based on knowledge graph and language model according to claim 2 is characterized by: The method for establishing a subject self-evaluation knowledge graph in the subjective evaluation graph unit (110) comprises the following steps: Step 1: Determine the tendency of each subjective evaluation item, mark the subjective evaluation items proved by direct facts as direct subjective evaluation items, and mark the subjective evaluation items proved by language description as indirect subjective evaluation items; Step 2: Establish a scoring method for indirect subjective evaluation items, match the indirect subjective evaluation item type of the subject, and obtain the subjective evaluation score ; Step 3: Collect direct factual evidence of each direct subjective evaluation item, and bind and generate proof materials for the direct subjective evaluation items.

4. The innovation and entrepreneurship guidance recommendation system based on knowledge graph and language model according to claim 3 is characterized by: The method for the objective evaluation graph unit (120) to establish an objective person evaluation knowledge graph comprises the following steps: The first step is to combine the scoring method of indirect subjective evaluation items to obtain the objective evaluation scores of each objective person for each indirect subjective evaluation item. ; Step 2: Obtain the objective evaluation scores of different objective persons for the same type of indirect subjective evaluation items , calculate the average objective evaluation score of the same category ,in to is the objective evaluation score of different objective persons for each indirect subjective evaluation item of the same type, and n is the number of objective persons; Step 3: Compare the average objective evaluation scores Subjective evaluation score , get the evaluation difference ; Step 4: Establish the evaluation difference threshold for the indirect subjective evaluation project , and make indirect subjective evaluation items elimination judgment; When evaluating the difference ≤ Poor evaluation threshold When Subjective evaluation score , take the two average values ​​as the final evaluation score; When evaluating the difference >Poor evaluation threshold When doing so, this type of indirect subjective evaluation item is excluded and is not used as the basis for the subjective person's evaluation.

5. The innovation and entrepreneurship guidance recommendation system based on knowledge graph and language model according to claim 1 is characterized by: The knowledge graph expansion module (10) also includes a reliability analysis unit (130), which obtains feedback information from historical entrepreneurs on subjective evaluation items, establishes objective reliability evaluation rules, and determines the evaluation status of the objective person.

6. The innovation and entrepreneurship guidance recommendation system based on knowledge graph and language model according to claim 5 is characterized by: The method for the reliability analysis unit (130) to establish objective reliability evaluation rules comprises the following steps: The first step is to collect the types of entrepreneurial projects of historical subjects and obtain the relevant objective persons of historical entrepreneurial projects; The second step is to combine the scoring method of indirect subjective evaluation items to obtain the historical evaluation scores of each related objective person for each indirect subjective evaluation item , calculate the average historical evaluation score ; The third step is to combine the historical evaluation score average , get the average value of the historical evaluation score Corresponding objective evaluation scores , calculate the reliability difference , set the reliability threshold ; Step 4: Set the objective evaluation score and the unit decrement score; When the reliability difference ≤Reliability Threshold , then the objective person's evaluation score is maintained; When the reliability difference >Reliability threshold , the objective person's objective evaluation score will be reduced by the unit decreasing score until the objective person's objective evaluation score is reset to zero, and the objective person's evaluation rights will be eliminated.

7. The innovation and entrepreneurship guidance recommendation system based on knowledge graph and language model according to claim 1 is characterized by: The large language model building module (20) comprises a keyword extraction unit (210) and an evaluation tendency acquisition unit (220). The keyword extraction unit (210) is used to obtain evaluation contents of a subject and an objector for an indirect subjective evaluation item, and to establish a keyword evaluation database to obtain evaluation tendencies of each keyword match. The evaluation tendency acquisition unit (220) combines the keyword evaluation database and the evaluation contents to obtain evaluation tendencies of the subject and the objector for different subjective evaluation items.

8. A method for using the innovation and entrepreneurship guidance recommendation system based on knowledge graph and language model as claimed in claim 1, characterized in that: The method comprises the following steps: S1. The subjective evaluation graph unit (110) collects the self-evaluation information of the subject, and classifies it according to the content of the self-evaluation information, obtains various subjective evaluation items, and establishes the subject's self-evaluation knowledge graph; S2, the objective evaluation graph unit (120) establishes a relationship network of the subject, identifies each subject, collects the objective evaluation content of each subject on the subjective evaluation item, establishes the objective evaluation knowledge graph, and verifies the accuracy of the subjective evaluation item; S3, a large language model building module (20) builds a large language model to capture self-evaluation information and objective evaluation content, obtain content related to subjective evaluation items, and eliminate irrelevant information; S4. The entrepreneurial project matching module (30) combines the objective evaluation knowledge graph and the self-evaluation knowledge graph, and compares them with the objective evaluation knowledge graph and the self-evaluation knowledge graph of historical entrepreneurs to obtain the subjective entrepreneurial trend and match the subjective with a suitable entrepreneurial project.

Images