Enterprise department and wound management platform based on artificial intelligence

Abstract

The invention discloses an enterprise department and innovation management platform based on artificial intelligence, relates to the technical field of artificial intelligence, and solves the problem that because department and innovation policies are complex and have a cooperative and complex relationship with each other, the enterprise is not easy to operate when facing related department and innovation policies. The real intention of the medical innovation policy cannot be accurately and concisely understood; the enterprise department and wound management platform comprises an enterprise management module, a policy collection module, a matching diagnosis module, a service query module, a department and wound shopping mall module and a data security module. The enterprise information is analyzed in detail through the department and wound inference model, and the department and wound policies are interpreted in detail, so that the department and wound policies corresponding to the enterprise development stages are matched for each enterprise, and a plurality of department and wound projects are recommended and a plurality of department and wound services are provided for each enterprise according to the department and wound policies; the difficulty of manually interpreting the medical creative policies in the prior art is reduced; and the accuracy of manually recommending the department and creative policies for the enterprise is improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an enterprise science and technology innovation management platform based on artificial intelligence. Background Art

[0002] In the current era of innovation and technological development, technological innovation is becoming increasingly important to enterprises. At the same time, due to the vigorous promotion of the "double innovation" policy, a large number of entrepreneurs and enterprises have been attracted to the innovation field. However, it is difficult to start and develop in a new field, and understanding and using policies and resources is a huge challenge.

[0003] At this time, the enterprise science and technology innovation management platform based on artificial intelligence can provide assistance to a large number of enterprises seeking innovation and solve the needs and challenges they face. Among them, enterprise policy matching is one of the important functions provided by the platform. Enterprise policy matching refers to analyzing the information of enterprises through artificial intelligence technology and providing customized services to enterprises in terms of policies and resources. Enterprise policy matching can help enterprises understand policies and intelligently screen policies that are beneficial to enterprises.

[0004] Patent No. CN2023109726501 discloses a method and server for managing scientific and technological innovation of an enterprise. In the above method, the project information of the target project sent by the user terminal is received, and it is determined that there is a project policy matching the project information in the policy project management system, and a policy set display instruction is sent to the user terminal to make the user terminal display the policy set, and the target policy is determined from the policy set. The corresponding application template is determined according to the target policy, and the text conversion instruction sent by the user terminal is received. The project information is input into the application template according to the set format, and the successful generation instruction corresponding to the project application is sent to the user terminal so that the user terminal displays the successful generation information. Through the above method, you can directly see which policies match the target project, and directly click the text conversion control to obtain the application corresponding to the policy, which saves the user's time to find the policy, and also saves the user's time to input the project information into the policy application template, thereby improving the efficiency of the user's policy application.

[0005] Patent No. CN2022102439200 discloses a full-chain and full-process intelligent science and technology service data processing method and system, which belongs to the field of information mining and information management technology. Through intelligent data mining means, the needs of science and technology enterprises are diagnosed, and the science and technology mall is provided to provide online management of the whole process of science and technology enterprise demand services, which is convenient for subsequent process monitoring and subsequent data management. Starting from the needs of enterprises, a science and technology mall service model is provided, and modular management is carried out at all stages of enterprise services. The service contracts generated can be entered into the contract management module of the system, and the system can automatically generate service orders according to the service content of the contract to achieve full-process and full-chain service tracking. The data processing and system of the above invention can be provided to investment institutions, service agencies with investment functions, incubators, park operating agencies, and government science and technology service departments for intelligent and visualized full-chain and full-process science and technology data management.

[0006] However, the above patents have different methods for enterprise scientific and technological innovation management, but because the policies are very complex, there are also synergies and complex relationships between them. When facing relevant policies, it is sometimes difficult for enterprises to accurately understand the true intention of the policy and the impact of this policy on the enterprise. Summary of the invention

[0007] The purpose of the present invention is to provide an enterprise science and technology innovation management platform based on artificial intelligence, which can use science and technology innovation reasoning models through artificial intelligence technology to conduct detailed analysis of enterprise information and interpret science and technology innovation policies in detail, so as to match each enterprise with the corresponding science and technology innovation policies at each stage of enterprise development, and recommend a number of science and technology innovation projects and provide a number of science and technology innovation services for each enterprise according to the science and technology innovation policies; so as to reduce the difficulty of existing manual interpretation of science and technology innovation policies; and improve the accuracy of manual recommendation of science and technology innovation policies for enterprises.

[0008] The present invention utilizes the following technical solutions:

[0009] An enterprise science and technology innovation management platform based on artificial intelligence, including an enterprise management module, a policy collection module, a matching diagnosis module, a service query module, a science and technology innovation mall module and a data security module; wherein,

[0010] The enterprise management module is used to add, delete and modify the information of settled enterprises and obtain the final version of enterprise information; and to classify settled enterprises according to enterprise types and obtain several enterprise profile tables;

[0011] The policy collection module is used to obtain all science and technology innovation policies from the State Council policy document library, the government official website and the national laws and regulations database, and classify and grade them to obtain a time-series science and technology innovation policy list;

[0012] A matching diagnosis module is used to match the corresponding science and technology innovation policies in the time series science and technology innovation policy list according to the enterprise profile table by using the covariate-propensity value algorithm or the science and technology innovation reasoning model, and recommend a number of science and technology innovation projects and provide a number of science and technology innovation services for the settled enterprises according to the science and technology innovation policies;

[0013] The service query module is used by real customers to manage the technology innovation service contracts and query the service status of the technology innovation projects of the settled enterprises;

[0014] The Science and Technology Innovation Mall module is used by potential customers to browse the Science and Technology Innovation policy section, Science and Technology Innovation service section and successful cases of enterprise services;

[0015] The data security module is used to store and encrypt the information of settled enterprises and scientific and technological innovation service contracts, so as to provide subsequent service tracking and data backtracking.

[0016] Preferably, enterprise information includes enterprise basic information, enterprise financing data, enterprise operation data, enterprise intellectual property data, enterprise label data, enterprise competitive product data and enterprise risk information; enterprise basic information includes enterprise annual report and industrial and commercial information; industrial and commercial information includes unified social credit code, organization code, registered address and legal representative;

[0017] The enterprise management module is used for managers of settled enterprises to register enterprise accounts; after completing the enterprise account registration, the enterprise management module has a built-in electronic template for basic enterprise information. After the manager completes the filling of the basic enterprise information and confirms the automatically filled information, the enterprise management module uses the unified social credit code and organization code in the industrial and commercial information and the optical character extraction algorithm to automatically fill in the electronic templates of enterprise financing data, enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitor data and enterprise risk information in turn according to the enterprise information queried in the enterprise information query tool to obtain the initial enterprise information template; the enterprise management module transmits each initial enterprise information template to the manager for inspection and confirmation in turn: if the information data in the initial enterprise information template is wrong and / or missing, the manager directly modifies the erroneous information data and / or missing information data until the initial enterprise information template is completely correct to obtain the final version of the enterprise information; if there is no error and / or omission in the information data in the initial enterprise information template, the initial enterprise information template is used as the final version of the enterprise information;

[0018] The enterprise management module divides the enterprise information according to the final version and the enterprise category to obtain several enterprise industry lists; enterprise categories include agriculture, industry and service industries; then according to the ownership form of the enterprise, all enterprise industry lists are divided into several enterprise ownership lists; ownership forms include state-owned, private, joint-stock and collective; finally, according to the registered address, according to province, city, district, county and township, all enterprise ownership lists are divided into several enterprise profile tables.

[0019] Preferably, the policy collection module automatically obtains all sci-tech innovation policies and policy application templates through the external interface opened by the official websites of government departments at all levels; governments at all levels include national, provincial, municipal and county levels; the policy application templates and sci-tech innovation policies are then associated using a data association algorithm to obtain a number of aggregated policy lists; and all aggregated policy lists are divided into a number of sci-tech innovation-specific policy lists according to the government level; and all sci-tech innovation-specific policy lists are divided into a number of sci-tech innovation-specific policy lists according to the policy type; policy types include tax reduction and exemption policies, venture capital preferential policies, preferential policies for attracting and cultivating talents, preferential policies for research and experimental development, preferential policies for transformation of results and preferential policies for the entire industrial chain; and all sci-tech innovation-specific policy lists are sorted by year according to the policy release time to obtain a chronological sci-tech innovation policy list.

[0020] Preferably, the matching diagnosis module first uses the enterprise evaluation layer of the scientific and technological innovation reasoning model to evaluate the scientific research capabilities of the settled enterprises according to the enterprise profile table to obtain a scientific research evaluation table; then uses the policy analysis layer of the scientific and technological innovation reasoning model to extract the key information of the scientific and technological innovation policy according to the time-series scientific and technological innovation policy list to obtain a policy information key table; then uses the fusion evaluation layer of the scientific and technological innovation reasoning model to match the scientific and technological innovation support policy for the settled enterprises according to the scientific research evaluation table and the policy information key table; finally, uses the scientific and technological innovation recommendation layer of the scientific and technological innovation reasoning model to recommend a number of scientific and technological innovation projects to the settled enterprises according to the scientific and technological innovation support policy, and provides a number of scientific and technological innovation services.

[0021] Preferably, the enterprise assessment layer first uses information analysis algorithms to analyze the enterprise profile table of the settled enterprises and obtain the annual report and legal representative of the enterprise;

[0022]

[0023] Among them, T represents the enterprise profile table, ∧ represents the logical AND operation, t represents the year, i represents the year collection sequence number, n represents the total number of year collections, α represents the enterprise annual report, represents the Kroger product operation, j represents the order number of the legal person drawn, m represents the total number of legal persons drawn, and β represents the legal representative;

[0024] Then, the balance sheet, income statement, cash flow statement, total operating income, main business income, total tax and R&D investment are extracted from the annual reports of the company for n consecutive years, and the proportion of R&D investment in each year is calculated respectively. The line chart of R&D proportion is drawn and analyzed to obtain the R&D analysis table:

[0025] If the broken line shows an upward trend, it means that the R&D investment of the current settled enterprises is increasing year by year, and the R&D capabilities are strengthening year by year; if the broken line shows a stable trend, it means that the R&D investment of the current settled enterprises remains stable, and the R&D capabilities are improving steadily; if the broken line shows a downward trend, it means that the R&D investment of the current settled enterprises is decreasing year by year, and the R&D capabilities are decreasing year by year;

[0026] At the same time, according to the increase, decrease and replacement of legal representatives for n consecutive years, the strategic planning and operation mode changes of the settled enterprises are analyzed, and the enterprise transfer analysis table is obtained:

[0027] If the R&D investment increases after the legal representative is updated, it means that the strategic focus of the settled enterprise is biased towards R&D; if the R&D investment remains unchanged after the legal representative is updated, it means that the settled enterprise has not changed its strategic focus; if the R&D investment decreases after the legal representative is updated, it means that the strategic focus of the settled enterprise has shifted towards R&D;

[0028] Finally, based on the R&D proportion line chart, R&D analysis table and enterprise mobilization analysis table, and combined with the changes in the enterprise's intellectual property data, a comprehensive analysis and evaluation was conducted to obtain the scientific research evaluation table.

[0029] Preferably, the policy parsing layer uses a word segmentation tool to remove stop words in the time-series science and technology innovation policy list, and divides the time-series science and technology innovation policy list without stop words into several independent phrases; then, according to the custom dictionary, uncommon characters or incorrect characters in all phrases are cleaned, and the similarity of all cleaned phrases is calculated, and the similarities are sorted in descending order, and then the key information of the science and technology innovation policy is extracted to obtain a key table of policy information.

[0030] Preferably, the fusion evaluation layer first extracts the word vector features of the scientific research evaluation table according to the preset data dimensions to obtain multi-dimensional scientific research vector features;

[0031] F=σ(WX s +b)+Q E (σ(WX s ),Y)+Q F (σ(WX s ),Y) (2)

[0032] Among them, T represents the multidimensional scientific research vector feature, σ() represents the activation function, W represents the weight value, X represents the data sample in the scientific research evaluation table, s represents the real data dimension, b represents the deviation vector, Switch() represents the selection function, Q E represents the data dimension reduction function, Y represents the true label, Q F Represents the data dimension-raising function;

[0033] Then, the information parsing algorithm is used to parse the key policy information tables according to the preset information importance, and the primary necessary information data set, the secondary necessary information data set and the non-essential information data set are obtained:

[0034]

[0035] Among them, ε represents the parsing error rate, i represents the first-level information sequence number, n represents the number of first-level information, γ represents the first-level vector matrix, and α represents each first-level necessary information. represents the primary necessary information data set, j represents the secondary information sequence number, m represents the number of secondary information, δ represents the secondary vector matrix, β represents each secondary information, Represents a secondary necessary information data set, represents the Krone product, ∑σ represents the set of non-essential information data;

[0036] Then, three network branches are used to extract deep vector features from the primary necessary information data set, the secondary necessary information data set, and the non-essential information data set, respectively, to obtain the primary necessary vector features, the secondary necessary vector features, and the non-essential vector features; the first network branch includes 2 3×3 convolutional layers, 2 5×5 convolutional layers, and 1 7×7 average pooling layer; the second network branch includes 3 convolutional blocks consisting of 3×1 convolutional layers and 1×3 convolutional layers, and 2 5×5 maximum pooling layers; the third network branch includes 3 3×3 convolutional layers, 1 7×7 convolutional layer, and 2 3×3 adaptive pooling layers;

[0037] Finally, the hierarchical evaluation algorithm is used to compare and evaluate the multidimensional scientific research vector characteristics with the first-level necessary vector characteristics, the secondary necessary vector characteristics and the non-essential vector characteristics, respectively, to obtain the first-level compliance degree, the secondary compliance degree and the non-essential compliance degree; then the policy matching degree is calculated based on the first-level compliance degree, the secondary compliance degree and the non-essential compliance degree, and the science and technology innovation support policies of the settled enterprises are screened and matched according to the policy matching degree in the time-series science and technology innovation policy list, and the corresponding policy application template is downloaded.

[0038] Preferably, the sci-tech innovation recommendation layer first extracts the enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitive product data and enterprise risk information of the past X years from the enterprise profile table, and uses the prediction fitting algorithm to predict the enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitive product data and enterprise risk information of the settled enterprises in the next Y years, and makes corrections in combination with the enterprise financing data to obtain the enterprise forecast development table; then, the policy matching degree of all sci-tech innovation support policies is calculated according to the enterprise forecast development table, and the non-compliant data in the sci-tech innovation support policies are marked; then, a number of sci-tech innovation projects are recommended for the settled enterprises according to the marked non-compliant data, enterprise industry categories and sci-tech innovation support policies; sci-tech innovation projects include basic research projects, applied research projects, scientific and technological innovation projects, social service projects, industrial innovation projects, strategic-oriented projects, decision-making support projects and special research projects; finally, corresponding sci-tech innovation services are provided according to the sci-tech innovation projects; sci-tech innovation services include technical research and development support, intellectual property services, project application guidance, industry-university-research cooperation docking, innovation resource sharing, market expansion assistance, financing channel docking and policy and regulatory consultation.

[0039] Preferably, the service query module includes a keyword extraction unit, a project matching unit and a progress display unit; the keyword extraction unit uses the TextRank algorithm to extract keywords for the retrieval query content, and searches for similar keywords from the scientific and technological innovation service contracts to extract all relevant scientific and technological innovation service project contracts; the content matching unit uses the collaborative filtering algorithm to eliminate the scientific and technological innovation service contracts that are not within the retrieval query time from all relevant scientific and technological innovation service project contracts according to the selected retrieval query time, and obtains all nearby scientific and technological innovation service contracts; the graph display unit interactively displays the service status of all nearby scientific and technological innovation service contracts.

[0040] Preferably, the data security module includes an information encryption unit and an information management unit; the information encryption unit uses a dynamic encryption algorithm to encrypt the information of the settled enterprises and the scientific and technological innovation service contract; the information management unit uses the metadata database to store the encrypted information of the settled enterprises and the scientific and technological innovation service contract, and backs them up on the blockchain and the cloud respectively, so as to provide subsequent service tracking and data backtracking.

[0041] The present invention uses a scientific and technological innovation reasoning model to conduct a detailed analysis of enterprise information and a detailed interpretation of scientific and technological innovation policies, thereby matching each enterprise with the corresponding scientific and technological innovation policies at each stage of enterprise development, and recommending a number of scientific and technological innovation projects and providing a number of scientific and technological innovation services for each enterprise based on the scientific and technological innovation policies; reducing the difficulty of existing manual interpretation of scientific and technological innovation policies; and improving the accuracy of manual recommendations of scientific and technological innovation policies for enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0043] Figure 1 Flowchart of the enterprise science and technology innovation management platform;

[0044] Figure 2 This is a schematic diagram of the principle of the scientific and technological innovation reasoning model. DETAILED DESCRIPTION

[0045] The present invention is described in detail below with reference to the accompanying drawings and embodiments:

[0046] like Figure 1-Figure 2 As shown, the enterprise science and technology innovation management platform based on artificial intelligence described in the present invention includes an enterprise management module, a policy collection module, a matching diagnosis module, a service query module, a science and technology innovation mall module and a data security module; wherein,

[0047] The enterprise management module is used to add, delete and modify the information of settled enterprises and obtain the final version of enterprise information; and to classify settled enterprises according to enterprise types and obtain several enterprise profile tables;

[0048] The policy collection module is used to obtain all science and technology innovation policies from the State Council policy document library, the government official website and the national laws and regulations database, and classify and grade them to obtain a time-series science and technology innovation policy list;

[0049] In this embodiment, the science and technology innovation policy, i.e., the science and technology innovation policy, refers to a series of policy measures formulated by the government to promote scientific and technological innovation. It is a branch of public policy, which aims to promote scientific and technological innovation activities, optimize the innovation environment, cultivate innovative talents, and promote the transformation of scientific and technological achievements;

[0050] Science and technology innovation policy is a public policy formulated by the government to promote the production, diffusion and application of scientific and technological knowledge. It involves many different fields and is related to a wide range of industrial policies such as trade, finance, fiscal policy, and foreign investment. Science and technology innovation policy can be divided into science policy, technology policy and innovation policy. There is an inherent organic connection between the three, and they work together in the whole process of scientific and technological innovation activities;

[0051] A matching diagnosis module is used to match the corresponding science and technology innovation policies in the time series science and technology innovation policy list according to the enterprise profile table by using the covariate-propensity value algorithm or the science and technology innovation reasoning model, and recommend a number of science and technology innovation projects and provide a number of science and technology innovation services for the settled enterprises according to the science and technology innovation policies;

[0052] The service query module is used by real customers to manage the technology innovation service contracts and query the service status of the technology innovation projects of the settled enterprises;

[0053] The Science and Technology Innovation Mall module is used by potential customers to browse the Science and Technology Innovation policy section, Science and Technology Innovation service section and successful cases of enterprise services;

[0054] The data security module is used to store and encrypt the information of settled enterprises and scientific and technological innovation service contracts, so as to provide subsequent service tracking and data backtracking.

[0055] In the present invention, enterprise information includes enterprise basic information, enterprise financing data, enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitive product data and enterprise risk information; enterprise basic information includes enterprise annual report and industrial and commercial information; industrial and commercial information includes unified social credit code, organization code, registered address and legal representative;

[0056] In this embodiment, basic enterprise information refers to the basic information of the enterprise, including but not limited to the name, registered address, establishment time, legal representative, registered capital, business scope, company type, unified social credit code, etc. This information usually comes from the official records when the enterprise is registered;

[0057] Corporate financing data: related information about the company's fundraising in the financial market, including financing time, financing method (such as equity financing, debt financing), financing amount, investor information, financing purpose, etc. These data can help understand the company's financial status and capital structure;

[0058] Business operating data: refers to the data generated by the company in its daily operations, such as operating income, net profit, cost, gross profit margin, debt-to-asset ratio, cash flow and other financial data, as well as production, sales, inventory, number of employees and other business data;

[0059] Enterprise intellectual property data: including information on the enterprise's intellectual property, such as patents, trademarks, copyrights, trade secrets, etc. These data reflect the enterprise's innovation ability, brand value and market competitiveness;

[0060] Enterprise label data: refers to labels given based on the characteristics, attributes or behaviors of enterprises, such as industry classification, enterprise size, market position, technology field, operating status, etc. These labels help to quickly identify and classify enterprises;

[0061] Enterprise competitive product data: related information about other enterprises or products that compete with the enterprise's products or services, including the market share, product features, pricing strategies, marketing activities, customer reviews, etc. of the competitors. These data help enterprises analyze competitors and the market environment;

[0062] Enterprise risk information: refers to information about events or situations that may have a negative impact on the operation of an enterprise, such as legal proceedings, administrative penalties, financial difficulties, credit defaults, negative market news, etc. Enterprise risk information is an important basis for enterprise risk assessment and management.

[0063] The enterprise management module is used for managers of settled enterprises to register enterprise accounts; after completing the enterprise account registration, the enterprise management module has a built-in electronic template for basic enterprise information. After the manager completes the filling of the basic enterprise information and confirms the automatically filled information, the enterprise management module uses the unified social credit code and organization code in the industrial and commercial information and the optical character extraction algorithm to automatically fill in the electronic templates of enterprise financing data, enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitor data and enterprise risk information in turn according to the enterprise information queried in the enterprise information query tool to obtain the initial enterprise information template; the enterprise management module transmits each initial enterprise information template to the manager for inspection and confirmation in turn: if the information data in the initial enterprise information template is wrong and / or missing, the manager directly modifies the erroneous information data and / or missing information data until the initial enterprise information template is completely correct to obtain the final version of the enterprise information; if there is no error and / or omission in the information data in the initial enterprise information template, the initial enterprise information template is used as the final version of the enterprise information;

[0064] The enterprise management module divides the enterprise information according to the final version and the enterprise category to obtain several enterprise industry lists; the enterprise categories include agriculture, industry and service industry; then divide all enterprise industry lists into several enterprise ownership lists according to the ownership form of the enterprise; the ownership forms include state-owned, private, joint-stock and collective; finally, divide all enterprise ownership lists into several enterprise profile tables according to the registered address, province, city, district, county and township;

[0065] In this embodiment, the enterprise categories include post and telecommunications, communications, community services, wholesale and retail, transportation, construction and installation, medical care, urban construction, tourism, hotels and restaurants, etc.

[0066] In the present invention, the policy collection module automatically obtains all science and technology innovation policies and policy application templates through the external interface opened by the official websites of government departments at all levels; governments at all levels include national, provincial, municipal and county levels; the policy application templates and science and technology innovation policies are then associated by a data association algorithm to obtain a number of aggregated policy lists; and all aggregated policy lists are divided into a number of science and technology innovation-specific policy lists according to the government level; and all science and technology innovation-specific policy lists are divided into a number of science and technology innovation single-category exclusive policy lists according to the policy type; policy types include tax reduction and exemption policies, venture capital preferential policies, preferential policies for attracting and cultivating talents, preferential policies for research and experimental development, preferential policies for achievement transformation and preferential policies for the entire industrial chain; and all science and technology innovation single-category exclusive policy lists are sorted by year according to the policy release time to obtain a time-series science and technology innovation policy list.

[0067] In the present invention, the matching diagnosis module first uses the enterprise evaluation layer of the scientific and technological innovation reasoning model to evaluate the scientific research capabilities of the settled enterprises according to the enterprise profile table to obtain a scientific research evaluation table; then uses the policy analysis layer of the scientific and technological innovation reasoning model to extract the key information of the scientific and technological innovation policy according to the time-series scientific and technological innovation policy list to obtain a policy information key table; then uses the fusion evaluation layer of the scientific and technological innovation reasoning model to match the scientific and technological innovation support policy for the settled enterprises according to the scientific research evaluation table and the policy information key table; finally, uses the scientific and technological innovation recommendation layer of the scientific and technological innovation reasoning model to recommend a number of scientific and technological innovation projects to the settled enterprises according to the scientific and technological innovation support policy, and provides a number of scientific and technological innovation services.

[0068] In the present invention, the enterprise evaluation layer first uses the information analysis algorithm to analyze the enterprise profile table of the settled enterprise and obtain the enterprise annual report and legal representative;

[0069]

[0070] Among them, T represents the enterprise profile table, ∧ represents the logical AND operation, t represents the year, i represents the year collection sequence number, n represents the total number of year collections, α represents the enterprise annual report, represents the Kroger product operation, j represents the order number of the legal person drawn, m represents the total number of legal persons drawn, and β represents the legal representative;

[0071] Then, the balance sheet, income statement, cash flow statement, total operating income, main business income, total tax and R&D investment are extracted from the annual reports of the company for n consecutive years, and the proportion of R&D investment in each year is calculated respectively. The line chart of R&D proportion is drawn and analyzed to obtain the R&D analysis table:

[0072] If the broken line shows an upward trend, it means that the R&D investment of the current settled enterprises is increasing year by year, and the R&D capabilities are strengthening year by year; if the broken line shows a stable trend, it means that the R&D investment of the current settled enterprises remains stable, and the R&D capabilities are improving steadily; if the broken line shows a downward trend, it means that the R&D investment of the current settled enterprises is decreasing year by year, and the R&D capabilities are decreasing year by year;

[0073] At the same time, according to the increase, decrease and replacement of legal representatives for n consecutive years, the strategic planning and operation mode changes of the settled enterprises are analyzed, and the enterprise transfer analysis table is obtained:

[0074] If the R&D investment increases after the legal representative is updated, it means that the strategic focus of the settled enterprise is biased towards R&D; if the R&D investment remains unchanged after the legal representative is updated, it means that the settled enterprise has not changed its strategic focus; if the R&D investment decreases after the legal representative is updated, it means that the strategic focus of the settled enterprise has shifted towards R&D;

[0075] Finally, based on the R&D proportion line chart, R&D analysis table and enterprise mobilization analysis table, and combined with the changes in the enterprise's intellectual property data, a comprehensive analysis and evaluation was conducted to obtain the scientific research evaluation table.

[0076] In the present invention, the policy parsing layer uses a word segmentation tool to remove stop words in the time-series science and technology innovation policy list, and divides the time-series science and technology innovation policy list without stop words into a number of independent phrases; then, according to a custom dictionary, uncommon characters or incorrect characters in all phrases are cleaned, and the similarity of all cleaned phrases is calculated, and the similarities are sorted in order from large to small, and then the key information of the science and technology innovation policy is extracted to obtain a key table of policy information.

[0077] In the present invention, the fusion evaluation layer first extracts the word vector features of the scientific research evaluation table according to the preset data dimensions to obtain multi-dimensional scientific research vector features;

[0078] F=σ(WX s +b)+Q E (σ(WX s ),Y)+Q F (σ(WX s ),Y) (2)

[0079] Among them, T represents the multidimensional scientific research vector feature, σ() represents the activation function, W represents the weight value, X represents the data sample in the scientific research evaluation table, s represents the real data dimension, b represents the deviation vector, Switch() represents the selection function, Q E represents the data dimension reduction function, Y represents the true label, Q F Represents the data dimension-raising function;

[0080] Then, the information parsing algorithm is used to parse the key policy information tables according to the preset information importance, and the primary necessary information data set, the secondary necessary information data set and the non-essential information data set are obtained:

[0081]

[0082] Among them, ε represents the parsing error rate, i represents the first-level information sequence number, n represents the number of first-level information, γ represents the first-level vector matrix, and α represents each first-level necessary information. represents the primary necessary information data set, j represents the secondary information sequence number, m represents the number of secondary information, δ represents the secondary vector matrix, β represents each secondary information, Represents a secondary necessary information data set, represents the Krone product, ∑σ represents the set of non-essential information data;

[0083] Then, three network branches are used to extract deep vector features from the primary necessary information data set, the secondary necessary information data set, and the non-essential information data set, respectively, to obtain the primary necessary vector features, the secondary necessary vector features, and the non-essential vector features; the first network branch includes 2 3×3 convolutional layers, 2 5×5 convolutional layers, and 1 7×7 average pooling layer; the second network branch includes 3 convolutional blocks consisting of 3×1 convolutional layers and 1×3 convolutional layers, and 2 5×5 maximum pooling layers; the third network branch includes 3 3×3 convolutional layers, 1 7×7 convolutional layer, and 2 3×3 adaptive pooling layers;

[0084] Finally, the hierarchical evaluation algorithm is used to compare and evaluate the multidimensional scientific research vector features with the first-level necessary vector features, the second-level necessary vector features, and the non-essential vector features in turn, and the first-level compliance, the second-level compliance, and the non-essential compliance are obtained respectively; then the policy matching degree is calculated based on the first-level compliance, the second-level compliance, and the non-essential compliance; in the time-series scientific and technological innovation policy list, the scientific and technological innovation support policies of the enterprises are screened and matched according to the policy matching degree, and the corresponding policy application template is downloaded;

[0085] In the present invention, the sci-tech innovation recommendation layer first extracts the enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitive product data and enterprise risk information of the past X years from the enterprise profile table, and uses the prediction fitting algorithm to predict the enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitive product data and enterprise risk information of the settled enterprises in the next Y years, and makes corrections in combination with the enterprise financing data to obtain the enterprise forecast development table; then, the policy matching degree of all sci-tech innovation support policies is calculated according to the enterprise forecast development table, and the unqualified data in the sci-tech innovation support policies are marked; then, a number of sci-tech innovation projects are recommended for the settled enterprises according to the marked unqualified data, enterprise industry categories and sci-tech innovation support policies; finally, corresponding sci-tech innovation services are provided according to the sci-tech innovation projects;

[0086] In this embodiment, scientific and technological innovation projects include basic research projects, applied research projects, scientific and technological innovation projects, social service projects, industrial innovation projects, strategic-oriented projects, decision-making support projects and special research projects; scientific and technological innovation services include technical research and development support, intellectual property services, project application guidance, industry-university-research cooperation docking, innovation resource sharing, market development assistance, financing channel docking and policy and regulatory consultation;

[0087] There is a close relationship between an enterprise's scientific and technological innovation projects and scientific and technological innovation services. Scientific and technological innovation services are designed to provide support for enterprise innovation and help enterprises better implement scientific and technological innovation projects.

[0088] First, enterprise science and technology innovation projects are important carriers of scientific and technological innovation. These projects usually involve the research and development of new technologies, the development of new products or the innovation of new service models, and aim to promote the technological progress and industrial upgrading of enterprises through innovative technologies, innovative models and innovative capabilities. Science and technology innovation services exist precisely to support the smooth progress of these science and technology innovation projects.

[0089] Scientific and technological innovation services cover many aspects, including but not limited to intellectual property services (such as trademark registration, patent application), technical services, expert consultation, industry-university-research cooperation, and application for government policy support projects. These services can provide enterprises with all-round support and help them solve various problems that may be encountered in the implementation of scientific and technological innovation projects.

[0090] In the present invention, the service query module includes a keyword extraction unit, a project matching unit and a progress display unit; the keyword extraction unit uses the TextRank algorithm to extract keywords for the search query content, and searches for similar keywords from the scientific and technological innovation service contracts to extract all relevant scientific and technological innovation service project contracts; the content matching unit uses the collaborative filtering algorithm to eliminate the scientific and technological innovation service contracts that are not within the search query time from all relevant scientific and technological innovation service project contracts according to the selected search query time, and obtains all nearby scientific and technological innovation service contracts; the graph display unit interactively displays the service status of all nearby scientific and technological innovation service contracts.

[0091] In the present invention, the data security module includes an information encryption unit and an information management unit; the information encryption unit uses a dynamic encryption algorithm to encrypt the information of the settled enterprises and the scientific and technological innovation service contract; the information management unit uses a metadata database to store the encrypted information of the settled enterprises and the scientific and technological innovation service contract, and backs them up on the blockchain and the cloud respectively, so as to provide subsequent service tracking and data backtracking.

[0092] Example:

[0093] The enterprise management module is used for managers of settled enterprises to register enterprise accounts; after completing the enterprise account registration, the enterprise management module has a built-in electronic template for basic enterprise information. After the manager completes the filling of the basic enterprise information and confirms the automatically filled information, the enterprise management module uses the unified social credit code and organization code in the industrial and commercial information and the optical character extraction algorithm to automatically fill in the electronic templates of enterprise financing data, enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitor data and enterprise risk information in turn according to the enterprise information queried in the enterprise information query tool to obtain the initial enterprise information template; the enterprise management module transmits each initial enterprise information template to the manager for inspection and confirmation in turn: if the information data in the initial enterprise information template is wrong and / or missing, the manager directly modifies the erroneous information data and / or missing information data until the initial enterprise information template is completely correct to obtain the final version of the enterprise information; if there is no error and / or omission in the information data in the initial enterprise information template, the initial enterprise information template is used as the final version of the enterprise information;

[0094] Enterprise information includes basic enterprise information, enterprise financing data, enterprise operation data, enterprise intellectual property data, enterprise label data, enterprise competitive product data and enterprise risk information; basic enterprise information includes annual report and industrial and commercial information; industrial and commercial information includes unified social credit code, organization code, registered address and legal representative;

[0095] The enterprise management module divides the enterprise information according to the final version and the enterprise category to obtain several enterprise industry lists; the enterprise categories include agriculture, industry and service industry; then divide all enterprise industry lists into several enterprise ownership lists according to the ownership form of the enterprise; the ownership forms include state-owned, private, joint-stock and collective; finally, divide all enterprise ownership lists into several enterprise profile tables according to the registered address, province, city, district, county and township;

[0096] At the same time, the policy collection module automatically obtains all science and technology innovation policies and policy application templates through the external interfaces opened on the official websites of government departments at all levels; governments at all levels include national, provincial, municipal and county levels; then use the data association algorithm to associate the policy application templates with science and technology innovation policies to obtain several aggregated policy lists; and divide all aggregated policy lists into several science and technology innovation-specific policy lists according to the government level; then divide all science and technology innovation-specific policy lists into several science and technology innovation-specific policy lists according to the policy type; policy types include tax reduction and exemption policies, venture capital preferential policies, preferential policies for attracting and cultivating talents, preferential policies for research and experimental development, preferential policies for achievement transformation and preferential policies for the entire industrial chain; then sort all science and technology innovation-specific policy lists by year according to the policy release time to obtain a time-series science and technology innovation policy list;

[0097] Then, the matching diagnosis module first uses the enterprise evaluation layer of the scientific and technological innovation reasoning model to evaluate the scientific research capabilities of the settled enterprises according to the enterprise profile table, and obtains the scientific research evaluation table; then uses the policy analysis layer of the scientific and technological innovation reasoning model to extract the key information of the scientific and technological innovation policy according to the time-series scientific and technological innovation policy list, and obtains the policy information key table; then uses the fusion evaluation layer of the scientific and technological innovation reasoning model to match the scientific and technological innovation support policy for the settled enterprises according to the scientific research evaluation table and the policy information key table; finally, uses the scientific and technological innovation recommendation layer of the scientific and technological innovation reasoning model to recommend a number of scientific and technological innovation projects for the settled enterprises according to the scientific and technological innovation support policy, and provides a number of scientific and technological innovation services;

[0098] Among them, the enterprise assessment layer first uses the information analysis algorithm to analyze the enterprise profile table of the settled enterprises and obtain the annual report and legal representative of the enterprise;

[0099] Then, the balance sheet, income statement, cash flow statement, total operating income, main business income, total tax and R&D investment are extracted from the annual reports of the company for n consecutive years, and the proportion of R&D investment in each year is calculated respectively. The line chart of R&D proportion is drawn and analyzed to obtain the R&D analysis table:

[0100] If the broken line shows an upward trend, it means that the R&D investment of the current settled enterprises is increasing year by year, and the R&D capabilities are strengthening year by year; if the broken line shows a stable trend, it means that the R&D investment of the current settled enterprises remains stable, and the R&D capabilities are improving steadily; if the broken line shows a downward trend, it means that the R&D investment of the current settled enterprises is decreasing year by year, and the R&D capabilities are decreasing year by year;

[0101] At the same time, according to the increase, decrease and replacement of legal representatives for n consecutive years, the strategic planning and operation mode changes of the settled enterprises are analyzed, and the enterprise transfer analysis table is obtained:

[0102] If the R&D investment increases after the legal representative is updated, it means that the strategic focus of the settled enterprise is biased towards R&D; if the R&D investment remains unchanged after the legal representative is updated, it means that the settled enterprise has not changed its strategic focus; if the R&D investment decreases after the legal representative is updated, it means that the strategic focus of the settled enterprise has shifted towards R&D;

[0103] Finally, based on the R&D ratio line chart, R&D analysis table and enterprise transfer analysis table, combined with the change in the enterprise's intellectual property data, a comprehensive analysis and evaluation was conducted to obtain the scientific research evaluation table;

[0104] The policy analysis layer uses word segmentation tools to remove stop words in the time-series science and technology innovation policy list, and divides the time-series science and technology innovation policy list without stop words into several independent phrases; then, according to the custom dictionary, the uncommon characters or wrong characters in all phrases are cleaned, and the similarity of all cleaned phrases is calculated, and the similarity is sorted in descending order, and then the key information of the science and technology innovation policy is extracted to obtain the key table of policy information;

[0105] The fusion evaluation layer first extracts the word vector features of the scientific research evaluation table according to the preset data dimensions to obtain multi-dimensional scientific research vector features;

[0106] Then, the information parsing algorithm is used to parse the key policy information tables according to the preset information importance, and the primary necessary information data set, the secondary necessary information data set and the non-essential information data set are obtained:

[0107] Then, three network branches are used to extract deep vector features from the primary necessary information data set, the secondary necessary information data set, and the non-essential information data set, respectively, to obtain the primary necessary vector features, the secondary necessary vector features, and the non-essential vector features;

[0108] Finally, the hierarchical evaluation algorithm is used to compare and evaluate the multidimensional scientific research vector features with the first-level necessary vector features, the second-level necessary vector features, and the non-essential vector features in turn, and the first-level compliance, the second-level compliance, and the non-essential compliance are obtained respectively; then the policy matching degree is calculated based on the first-level compliance, the second-level compliance, and the non-essential compliance; in the time-series scientific and technological innovation policy list, the scientific and technological innovation support policies of the enterprises are screened and matched according to the policy matching degree, and the corresponding policy application template is downloaded;

[0109] The sci-tech innovation recommendation layer first extracts the enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitive product data and enterprise risk information of the past X years from the enterprise profile table, and uses the prediction fitting algorithm to predict the enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitive product data and enterprise risk information of the settled enterprises in the next Y years, and makes corrections based on the enterprise financing data to obtain the enterprise forecast development table; then calculates the policy matching degree of all sci-tech innovation support policies based on the enterprise forecast development table, and marks the non-compliant data in the sci-tech innovation support policies; then recommends a number of sci-tech innovation projects to the settled enterprises based on the marked non-compliant data, enterprise industry categories and sci-tech innovation support policies; sci-tech innovation projects include basic research projects, applied research projects, scientific and technological innovation projects, social service projects, industrial innovation projects, strategic-oriented projects, decision-making support projects and special research projects; finally, provides corresponding sci-tech innovation services based on sci-tech innovation projects; sci-tech innovation services include technical research and development support, intellectual property services, project application guidance, industry-university-research cooperation docking, innovation resource sharing, market expansion assistance, financing channel docking and policy and regulatory consultation;

[0110] In reality, customers use the TextRank algorithm to extract keywords for search query content through the keyword extraction unit of the service query module, and search for similar keywords from the science and technology service contracts to extract all relevant science and technology service project contracts; the content matching unit uses the collaborative filtering algorithm to remove science and technology service contracts that are not within the search query time from all relevant science and technology service project contracts according to the selected search query time, and obtains all nearby science and technology service contracts; the graph display unit interactively displays the service status of all nearby science and technology service contracts;

[0111] Potential customers can browse the science and technology innovation policy section, science and technology innovation service section and enterprise service success cases through the science and technology innovation mall module;

[0112] Finally, the information encryption unit of the data security module uses a dynamic encryption algorithm to encrypt the information of the settled enterprises and the scientific and technological innovation service contracts; the information management unit uses the metadata database to store the encrypted information of the settled enterprises and the scientific and technological innovation service contracts, and backs them up on the blockchain and the cloud respectively, so as to provide subsequent service tracking and data backtracking.

Claims

1. An enterprise science and technology innovation management platform based on artificial intelligence, characterized by: It includes enterprise management module, policy collection module, matching diagnosis module, service query module, science and technology innovation mall module and data security module; among them, The enterprise management module is used to add, delete and modify the information of settled enterprises and obtain the final version of enterprise information; and to classify settled enterprises according to enterprise types and obtain several enterprise profile tables; The policy collection module is used to obtain all science and technology innovation policies from the State Council policy document library, the official government website and the national laws and regulations database, and classify and grade them to obtain a time-series science and technology innovation policy list; A matching diagnosis module is used to match the corresponding science and technology innovation policies in the time series science and technology innovation policy list according to the enterprise profile table by using the covariate-propensity value algorithm or the science and technology innovation reasoning model, and recommend a number of science and technology innovation projects and provide a number of science and technology innovation services for the settled enterprises according to the science and technology innovation policies; The service query module is used by real customers to manage the technology innovation service contracts and query the service status of the technology innovation projects of the settled enterprises; The Science and Technology Innovation Mall module is used by potential customers to browse the Science and Technology Innovation policy section, Science and Technology Innovation service section and successful cases of enterprise services; The data security module is used to store and encrypt the information of settled enterprises and scientific and technological innovation service contracts, so as to provide subsequent service tracking and data backtracking.

2. The enterprise science and technology innovation management platform based on artificial intelligence according to claim 1 is characterized by: The enterprise information includes enterprise basic information, enterprise financing data, enterprise operation data, enterprise intellectual property data, enterprise label data, enterprise competitor data and enterprise risk information; Basic enterprise information includes annual reports and industrial and commercial information; industrial and commercial information includes unified social credit code, organization code, registered address and legal representative; The enterprise management module is used for managers of settled enterprises to register enterprise accounts; after completing the enterprise account registration, the enterprise management module has a built-in electronic template for basic enterprise information. After the manager completes the filling of the basic enterprise information and confirms the automatically filled information, the enterprise management module uses the unified social credit code and organization code in the industrial and commercial information, and uses the optical character extraction algorithm to automatically fill in the electronic templates of enterprise financing data, enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitor data and enterprise risk information according to the enterprise information queried in the enterprise information query tool, and obtains the initial enterprise information template; the enterprise management module transmits each initial enterprise information template to the manager for inspection and confirmation in turn: if there are errors and / or omissions in the information data in the initial enterprise information template, the manager directly modifies the erroneous information data and / or the missing information data until the initial enterprise information template is completely correct, and the final version of the enterprise information is obtained; If there are no errors and / or omissions in the information data in the initial enterprise information template, the initial enterprise information template will be used as the final version of the enterprise information; The enterprise management module divides the enterprise information according to the final version and the enterprise category to obtain several enterprise industry lists; enterprise categories include agriculture, industry and service industries; then according to the ownership form of the enterprise, all enterprise industry lists are divided into several enterprise ownership lists; ownership forms include state-owned, private, joint-stock and collective; finally, according to the registered address, according to province, city, district, county and township, all enterprise ownership lists are divided into several enterprise profile tables.

3. The enterprise science and technology innovation management platform based on artificial intelligence according to claim 1 is characterized by: The policy collection module automatically obtains all science and technology innovation policies and policy application templates through the external interface opened by the official websites of government departments at all levels; governments at all levels include national, provincial, municipal and county levels; then the policy application templates and science and technology innovation policies are associated by the data association algorithm to obtain a number of aggregated policy lists; and all aggregated policy lists are divided into a number of science and technology innovation-specific policy lists according to the government level; then all science and technology innovation-specific policy lists are divided into a number of science and technology innovation single-category exclusive policy lists according to the policy type; policy types include tax reduction and exemption policies, venture capital preferential policies, preferential policies for attracting and cultivating talents, preferential policies for research and experimental development, preferential policies for achievement transformation and preferential policies for the entire industrial chain; then all science and technology innovation single-category exclusive policy lists are sorted by year according to the policy release time to obtain a time-series science and technology innovation policy list.

4. The enterprise science and technology innovation management platform based on artificial intelligence according to claim 1 is characterized by: The matching diagnosis module first uses the enterprise evaluation layer of the scientific and technological innovation reasoning model to evaluate the scientific research capabilities of the settled enterprises according to the enterprise profile table to obtain a scientific research evaluation table; then uses the policy analysis layer of the scientific and technological innovation reasoning model to extract key information of the scientific and technological innovation policies according to the time-series scientific and technological innovation policy list to obtain a policy information key table; Subsequently, the fusion evaluation layer of the scientific and technological innovation reasoning model is used to match scientific and technological innovation support policies for the settled enterprises according to the scientific research evaluation form and the policy information key table; finally, the scientific and technological innovation recommendation layer of the scientific and technological innovation reasoning model is used to recommend a number of scientific and technological innovation projects to the settled enterprises according to the scientific and technological innovation support policies, and provide a number of scientific and technological innovation services.

5. The enterprise science and technology innovation management platform based on artificial intelligence according to claim 4 is characterized by: The enterprise evaluation layer first uses information analysis algorithms to analyze the enterprise profile of the settled enterprises and obtain the annual reports and legal representatives of the enterprises; Then from the continuous The balance sheet, income statement, cash flow statement, total operating income, main business income, total tax payment and R&D investment are extracted from the annual report of the company in 2017, and the proportion of R&D investment in each year is calculated respectively. The line graph of R&D proportion is drawn and analyzed to obtain the R&D analysis table: If the broken line shows an upward trend, it means that the R&D investment of the current settled enterprises is increasing year by year, and the R&D capabilities are strengthening year by year; If the broken line shows a stable trend, it means that the R&D investment of the current settled enterprises remains stable and the R&D capabilities are steadily improving; If the broken line shows a downward trend, it means that the R&D investment of the current companies is decreasing year by year, and their R&D capabilities are decreasing year by year; At the same time, according to the continuous Based on the increase, decrease and replacement of legal representatives in 2017, the strategic planning and operation mode changes of the settled enterprises were analyzed, and the enterprise transfer analysis table was obtained: If the R&D investment increases after the legal representative is updated, it means that the strategic focus of the settled enterprise is biased towards R&D; if the R&D investment remains unchanged after the legal representative is updated, it means that the settled enterprise has not changed its strategic focus; If R&D investment decreases after the legal representative is updated, it means that the strategic focus of the settled enterprise has shifted to R&D; Finally, based on the R&D proportion line chart, R&D analysis table and enterprise mobilization analysis table, and combined with the changes in the enterprise's intellectual property data, a comprehensive analysis and evaluation was conducted to obtain the scientific research evaluation table.

6. The enterprise science and technology innovation management platform based on artificial intelligence according to claim 4 is characterized by: The policy analysis layer uses a word segmentation tool to remove stop words in the time-series science and technology innovation policy list, and divides the time-series science and technology innovation policy list without stop words into several independent phrases; then, according to a custom dictionary, uncommon characters or incorrect characters in all phrases are cleaned, and the similarity of all cleaned phrases is calculated, and the similarities are sorted in descending order, and then the key information of the science and technology innovation policy is extracted to obtain a key table of policy information.

7. The enterprise science and technology innovation management platform based on artificial intelligence according to claim 4 is characterized by: The fusion evaluation layer first extracts the word vector features of the scientific research evaluation table according to the preset data dimensions to obtain multi-dimensional scientific research vector features; Then, the information parsing algorithm is used to parse the key policy information tables according to the preset information importance, and the primary necessary information data set, the secondary necessary information data set and the non-essential information data set are obtained: Then, three network branches are used to extract deep vector features from the primary necessary information data set, the secondary necessary information data set, and the non-essential information data set, respectively, to obtain the primary necessary vector features, the secondary necessary vector features, and the non-essential vector features; Finally, the hierarchical evaluation algorithm is used to compare and evaluate the multidimensional scientific research vector characteristics with the first-level necessary vector characteristics, the secondary necessary vector characteristics and the non-essential vector characteristics, respectively, to obtain the first-level compliance degree, the secondary compliance degree and the non-essential compliance degree; then the policy matching degree is calculated based on the first-level compliance degree, the secondary compliance degree and the non-essential compliance degree, and the science and technology innovation support policies of the settled enterprises are screened and matched according to the policy matching degree in the time-series science and technology innovation policy list, and the corresponding policy application template is downloaded.

8. The enterprise science and technology innovation management platform based on artificial intelligence according to claim 4 is characterized by: The science and technology innovation recommendation layer first extracts the enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitive product data and enterprise risk information of the past X years from the enterprise profile table, and uses the prediction fitting algorithm to predict the enterprise operating data, enterprise intellectual property data, enterprise label data, enterprise competitive product data and enterprise risk information of the settled enterprises in the next Y years, and makes corrections in combination with the enterprise financing data to obtain the enterprise forecast development table; Then, the policy matching degree of all science and technology innovation support policies is calculated based on the enterprise forecast development table, and the data that does not meet the standards in the science and technology innovation support policies are marked; Subsequently, a number of scientific and technological innovation projects are recommended to the settled enterprises based on the marked non-compliant data, enterprise industry categories and scientific and technological innovation support policies; finally, corresponding scientific and technological innovation services are provided based on the scientific and technological innovation projects.

9. The enterprise science and technology innovation management platform based on artificial intelligence according to claim 1 is characterized by: The service query module includes a keyword extraction unit, a project matching unit and a progress display unit; the keyword extraction unit uses the TextRank algorithm to extract keywords for the search query content, and searches for similar keywords from the scientific and technological innovation service contracts to extract all relevant scientific and technological innovation service project contracts; the content matching unit uses the collaborative filtering algorithm to eliminate the scientific and technological innovation service contracts that are not within the search query time from all relevant scientific and technological innovation service project contracts according to the selected search query time, and obtains all nearby scientific and technological innovation service contracts; the graph display unit interactively displays the service status of all nearby scientific and technological innovation service contracts.

10. The enterprise science and technology innovation management platform based on artificial intelligence according to claim 1 is characterized by: The data security module includes an information encryption unit and an information management unit; the information encryption unit uses a dynamic encryption algorithm to encrypt the information of the settled enterprises and the scientific and technological innovation service contract; The information management unit uses the metadata database to store the encrypted information of the settled enterprises and the scientific and technological innovation service contracts, and backs them up on the blockchain and the cloud respectively, so as to provide subsequent service tracking and data backtracking.

Images