Scientific and technological enterprise innovation ability evaluation method and system based on big data model

Abstract

The invention relates to the technical field of intelligent evaluation, and particularly discloses a scientific and technological enterprise innovation ability evaluation method and system based on a big data model, and the method comprises the steps: obtaining the basic information of an enterprise based on a preset statistical caliber, carrying out the data transcoding of the basic information, obtaining a data value, and generating an enterprise histogram based on the data value; obtaining product information of an enterprise, and performing two-dimensional processing on the product information to obtain a product feature map; determining an innovation capability value of the enterprise according to the enterprise histogram and the product features; constructing a sample set by taking the basic information and the product information as features and the innovation capability value as a label, and training a neural network model based on the sample set; according to the method, the enterprise is evaluated from the two aspects of indexes and products, the neural grid model is trained while the enterprise is evaluated according to the evaluation algorithm, the evaluation accuracy is high, and the evaluation efficiency is high.

Description

Technical Field

[0001] This invention relates to the field of intelligent evaluation technology, specifically a method and system for evaluating the innovation capabilities of technology companies based on a big data model. Background Technology

[0002] In today's era of rapid technological advancement, the role of technology companies in social development is becoming increasingly prominent. However, investment institutions and industry analysts consistently face significant challenges in scientifically and objectively evaluating the innovation capabilities of these companies. Traditional evaluation methods rely excessively on manual assessment and subjective judgment, lacking a scientific and unified evaluation system. This makes the evaluation results highly susceptible to subjective biases, failing to accurately reflect the true innovation capabilities of companies. Therefore, the development of a method and system for evaluating the innovation capabilities of technology companies based on big data models is urgently needed. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for evaluating the innovation capabilities of technology companies based on a big data model, so as to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A method for evaluating the innovation capability of technology companies based on a big data model, the method comprising:

[0006] Basic information of enterprises is obtained based on preset statistical criteria, the basic information is transcoded to obtain data values, and an enterprise histogram is generated based on the data values; the statistical criteria are used to characterize preset indicators.

[0007] Obtain the company's product information, perform two-dimensional processing on the product information, and obtain a product feature map;

[0008] Determine the innovation capability value of an enterprise based on its histogram and product characteristics;

[0009] A sample set is constructed by using basic information and product information as features and innovation capability values ​​as labels. A neural network model is trained based on the sample set. When the error rate of the neural network model is less than a preset threshold, the neural network model is used as the capability assessment scheme.

[0010] As a further aspect of the present invention: the steps of obtaining basic enterprise information based on a preset statistical caliber, transcoding the basic information to obtain data values, and generating an enterprise histogram based on the data values ​​include:

[0011] Receive user input of indicator types and their weights as statistical criteria;

[0012] Send information request to enterprises and receive data from enterprises regarding each indicator;

[0013] A transcoding model is selected based on the data structure of the indicator. The data is input into the transcoding model to obtain the data value of each indicator. The transcoding model contains an indicator positiveization function.

[0014] The enterprise histogram is obtained by statistically analyzing the data values ​​according to the preset indicator order; during the data value statistics process, the data values ​​are adjusted based on weights.

[0015] As a further aspect of the present invention: the step of obtaining the enterprise's product information and performing two-dimensional processing on the product information to obtain a product feature map includes:

[0016] Obtain information on a company's product types, sales revenue, and product launch dates;

[0017] The mapping shape is determined based on the product type, and the shape size is determined based on the product sales revenue. The mapping shape is a regular polygon. The number of sides of the regular polygon is inversely proportional to the regularity of the product type. The regularity of different product types is a preset value and stored in a preset product statistics table. The process of determining the product statistics table is as follows: enterprises are classified according to enterprise type. For each type of enterprise, the total number of enterprises corresponding to each product type is calculated. The regularity of the product type is determined based on the total number of enterprises, and the regularity is directly proportional to the total number of enterprises. The product types and their regularities are statistically analyzed to obtain the product statistics table.

[0018] The statistical order is determined based on the product release time, and all mapping shapes of the enterprise are statistically analyzed based on the statistical order to obtain the product feature map.

[0019] As a further aspect of the present invention: the step of determining the innovation capability value of an enterprise based on its histogram and product characteristics includes:

[0020] Read the enterprise histograms of all enterprises, calculate the enterprise similarity, and cluster the enterprises based on the enterprise similarity;

[0021] For each type of enterprise, calculate the mean histogram of all enterprises' histograms;

[0022] Input the values ​​of each indicator in the mean histogram into the preset indicator calculation model, and output the benchmark innovation value of this type of enterprise.

[0023] Read the product feature map of the enterprise, accumulate the area of ​​each mapped shape according to the preset decreasing weight, and determine the additional innovation value based on the area; the additional innovation value is proportional to the area.

[0024] The innovation capability value is obtained by summing the baseline innovation value and the additional innovation value.

[0025] As a further aspect of the present invention: the step of constructing a sample set by using basic information and product information as features and innovation capability values ​​as labels, training a neural network model based on the sample set, and using the neural network model as a capability evaluation scheme when the error rate of the neural network model is less than a preset threshold includes:

[0026] Read the basic information and product information of each enterprise, and read the enterprise's innovation capability value;

[0027] A sample set is constructed using basic information and product information as features and innovation capability values ​​as labels, and a neural network model is trained based on the sample set.

[0028] The error rate of the neural network model is read. When the error rate is less than a preset threshold, the application time of the neural network model within a preset time period is determined based on the error rate.

[0029] As a further aspect of the present invention, the method further includes:

[0030] The system displays the enterprise histogram and product feature map based on a pre-defined UI interface.

[0031] It receives encryption requests from enterprises containing encryption targets and encrypts the content of enterprise histograms and product feature maps.

[0032] The present invention also provides a technology enterprise innovation capability evaluation system based on a big data model, the system comprising:

[0033] The indicator statistics module is used to obtain basic information of enterprises based on preset statistical standards, perform data transcoding on the basic information to obtain data values, and generate enterprise histograms based on the data values; the statistical standards are used to characterize preset indicators.

[0034] The product information statistics module is used to acquire the company's product information, perform two-dimensional processing on the product information, and obtain a product feature map.

[0035] The capability value calculation module is used to determine the innovation capability value of an enterprise based on its histogram and product characteristics.

[0036] The model training application module is used to construct a sample set by using basic information and product information as features and innovation capability values ​​as labels. The neural network model is trained based on the sample set. When the error rate of the neural network model is less than a preset threshold, the neural network model is used as a capability assessment scheme.

[0037] As a further aspect of the present invention: the indicator statistics module includes:

[0038] The statistical caliber generation unit is used to receive the indicator type and its weight input by the user, and use it as the statistical caliber.

[0039] The data acquisition unit is used to send information acquisition requests to enterprises and receive data from enterprises regarding each indicator.

[0040] The data transcoding unit is used to select a transcoding model based on the data structure of the indicator, input the data into the transcoding model, and obtain the data value of each indicator; the transcoding model contains an indicator positiveization function;

[0041] The histogram generation unit is used to statistically analyze data values ​​according to a preset indicator order to obtain the enterprise histogram; during the data value statistics process, the data values ​​are adjusted based on weights.

[0042] As a further aspect of the present invention: the product information statistics module includes:

[0043] The acquisition execution unit is used to acquire information such as the enterprise's product types, product sales revenue, and product release time.

[0044] A shape mapping unit is used to determine the mapping shape based on the product type and the shape size based on the product sales revenue. The mapping shape is a regular polygon. The number of sides of the regular polygon is inversely proportional to the regularity of the product type. The regularity of different product types is a preset value and stored in a preset product statistics table. The process of determining the product statistics table is as follows: enterprises are classified according to enterprise type. For each type of enterprise, the total number of enterprises corresponding to each product type is calculated. The regularity of the product type is determined based on the total number of enterprises, and the regularity is directly proportional to the total number of enterprises. The product types and their regularities are statistically analyzed to obtain the product statistics table.

[0045] The mapping execution unit is used to determine the statistical order based on the product release time, and to count all the mapping shapes of the enterprise based on the statistical order to obtain the product feature map.

[0046] As a further aspect of the present invention: the capability value calculation module includes:

[0047] The clustering unit is used to read the enterprise histograms of all enterprises, calculate the enterprise similarity, and cluster the enterprises based on the enterprise similarity.

[0048] The mean calculation unit is used to calculate the mean histogram of all enterprises for each type of enterprise.

[0049] The indicator analysis unit is used to input the values ​​of each indicator in the mean histogram into a preset indicator calculation model and output the benchmark innovation value of this type of enterprise.

[0050] The added value calculation unit is used to read the product feature map of the enterprise, accumulate the area of ​​each mapped shape according to the preset decreasing weight, and determine the added innovation value based on the area; the added innovation value is proportional to the area.

[0051] The numerical summation unit is used to sum the baseline innovation value and the additional innovation value to obtain the innovation capability value.

[0052] Compared with the prior art, the beneficial effects of the present invention are: the present invention evaluates enterprises from the perspectives of indicators and products, and while evaluating enterprises according to the evaluation algorithm, it trains a neural grid model, resulting in high evaluation accuracy and high evaluation efficiency. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention.

[0054] Figure 1 This is a flowchart of a method for evaluating the innovation capabilities of technology companies based on a big data model.

[0055] Figure 2 This is the first sub-flowchart of a method for evaluating the innovation capabilities of technology companies based on big data models.

[0056] Figure 3 This is the second sub-process flowchart of the evaluation method for the innovation capability of technology companies based on big data models.

[0057] Figure 4 This is the third sub-process flowchart of the evaluation method for the innovation capability of technology companies based on big data models.

[0058] Figure 5 This is the fourth sub-process flowchart of the evaluation method for the innovation capability of technology companies based on big data models.

[0059] Figure 6 This is a structural diagram of a technology enterprise innovation capability evaluation system based on a big data model. Detailed Implementation

[0060] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0061] Figure 1 The flowchart illustrates a method for evaluating the innovation capabilities of technology companies based on a big data model. In this embodiment of the invention, a method for evaluating the innovation capabilities of technology companies based on a big data model includes:

[0062] Step S100: Obtain basic information of the enterprise based on a preset statistical standard, perform data transcoding on the basic information to obtain data values, and generate an enterprise histogram based on the data values; the statistical standard is used to characterize the preset indicators;

[0063] Step S200: Obtain the enterprise's product information, perform two-dimensional processing on the product information, and obtain a product feature map;

[0064] Step S300: Determine the company's innovation capability value based on the company's histogram and product characteristics;

[0065] Step S400: Construct a sample set by using basic information and product information as features and innovation capability values ​​as labels. Train a neural network model based on the sample set. When the error rate of the neural network model is less than a preset threshold, use the neural network model as a capability assessment scheme.

[0066] This invention provides a technical solution for intelligent enterprise evaluation. First, it requires acquiring enterprise data, specifically the basic information mentioned above. This basic information encompasses various types, including the responsibility to collect data from diverse data sources related to technology companies. This data covers, but is not limited to, dimensions such as R&D investment scale, patent quantity and quality, technological leadership, talent structure characteristics, and innovation culture. The collected information undergoes cleaning, integration, and standardization to ensure accuracy and consistency. The statistical definitions of the indicators involved in the basic information are numerous; in other words, each indicator corresponds to different units and even different formats—some are numerical, others are text. Transcoding this data yields dimensionless numerical data. Statistical analysis of these transcoded values ​​produces an enterprise histogram reflecting the company's basic information.

[0067] Based on this, the company's product information is obtained, including what products the company has, the sales volume of each product, and the release time of each product. The product information is then converted into data and processed in a two-dimensional form to obtain a product feature map.

[0068] After the above processing, the company's basic information is converted into a company histogram, and the company's product information is converted into a product feature map. Using the company histogram and product feature map as independent variables, an evaluation algorithm is applied to obtain an innovation capability value, which is used to evaluate the company's innovation capability.

[0069] It should be noted that the enterprise histograms and product feature maps mentioned above are data that can be directly integrated into the display process. Conventional display technologies can be used to create UI interfaces, allowing users to interact with the data and intuitively display the enterprise's information.

[0070] Based on the above, this application also introduces a neural network model. The input of the neural network model is basic information and product information. The data format of the input data is not limited. Its output is the innovation capability value. At the beginning of the neural network model, the original evaluation algorithm is used to process the basic information and product information. When the accuracy of the neural network model is high enough, the neural network model is applied. At this time, the data transcoding process and the application process of the evaluation algorithm are simplified and more efficient.

[0071] It is worth mentioning that the output of the neural grid model training process can be sent to a human reviewer to receive the corrected innovation capability value from the human reviewer, resulting in a more accurate neural network model.

[0072] Figure 2 The first sub-process flowchart of the evaluation method for the innovation capability of technology enterprises based on a big data model includes the following steps: obtaining basic information of enterprises based on a preset statistical caliber, transcoding the basic information to obtain data values, and generating enterprise histograms based on the data values.

[0073] Step S101: Receive the indicator type and its weight input by the user as the statistical caliber;

[0074] Step S102: Send an information acquisition request to the enterprise and receive data from the enterprise regarding each indicator;

[0075] Step S103: Select a transcoding model based on the data structure of the indicator, input the data into the transcoding model, and obtain the data value of each indicator; the transcoding model contains an indicator positiveization function;

[0076] Step S104: Statistical data values ​​are collected according to a preset indicator order to obtain an enterprise histogram; wherein, the data values ​​are adjusted based on weights during the data value collection process.

[0077] In one example of the technical solution of this invention, the process of generating an enterprise histogram is described. The system receives the indicator type and its weight input by the user as a statistical caliber, sends an information acquisition request to the enterprise, receives data on each indicator from the enterprise, selects a transcoding model according to the data structure of the indicator, inputs the data into the transcoding model, obtains the data value of each indicator, and statistically analyzes the data values ​​according to a preset indicator order to obtain the enterprise histogram.

[0078] In the transcoding process, all indicators need to be positively processed so that the converted values ​​are as high as possible. In addition, for the process of generating enterprise histograms, this application also introduces a weight for each data type. The more important the data, the greater its weight. The weight is multiplied by the converted value to obtain the adjusted value.

[0079] Figure 3 This is the second sub-flowchart of the method for evaluating the innovation capabilities of technology companies based on a big data model. The step of obtaining the company's product information and performing two-dimensional processing on the product information to obtain a product feature map includes:

[0080] Step S201: Obtain the company's product types, product sales volume, and product launch time;

[0081] Step S202: Determine the mapping shape based on the product type and the shape size based on the product sales revenue; the mapping shape is a regular polygon.

[0082] Step S203: Determine the statistical order based on the product release time, and statistically analyze all the mapping shapes of the enterprise based on the statistical order to obtain the product feature map.

[0083] In one example of the technical solution of the present invention, the process of analyzing product information is described. The product type, product sales and product release time of the enterprise are obtained. The mapping shape is determined according to the product type and the shape size is determined according to the product sales. The mapping shape is a regular polygon. The statistical order is determined based on the product release time. All mapping shapes of the enterprise are counted based on the statistical order to obtain the product feature map.

[0084] In the process of determining the regular polygon, the number of sides of the regular polygon is inversely proportional to the regularity of the product type. As a result, the more unconventional a product is, the more sides it has, the better its sales are, and the longer its side length is, the larger the area of ​​the corresponding mapped shape.

[0085] Furthermore, the regularity of different product types is a preset value, stored in a preset product statistics table; the process of determining the product statistics table is as follows: classify enterprises according to enterprise type, calculate the total number of enterprises corresponding to each product type for each type of enterprise, determine the regularity of product type based on the total number of enterprises, and the regularity is directly proportional to the total number of enterprises; count the product types and their regularity to obtain the product statistics table;

[0086] Figure 4 This is the third sub-process flowchart of the method for evaluating the innovation capability of technology companies based on big data models. The step of determining the innovation capability value of a company based on its histogram and product characteristics includes:

[0087] Step S301: Read the enterprise histograms of all enterprises, calculate the enterprise similarity, and cluster the enterprises based on the enterprise similarity;

[0088] Step S302: For each type of enterprise, calculate the mean histogram of all enterprises' histograms;

[0089] Step S303: Input the values ​​of each indicator in the mean histogram into the preset indicator calculation model, and output the benchmark innovation value of this type of enterprise;

[0090] Step S304: Read the product feature map of the enterprise, accumulate the area of ​​each mapped shape according to the preset decreasing weight, and determine the additional innovation value based on the area; the additional innovation value is proportional to the area.

[0091] Step S305: Sum the baseline innovation value and the additional innovation value to obtain the innovation capability value.

[0092] In one example of the technical solution of this invention, the enterprise histograms of all enterprises are read, enterprise similarity is calculated, and enterprises are clustered according to the enterprise similarity. The enterprise histogram is essentially an array, and the process of calculating enterprise similarity is an array comparison process. For each category of enterprises, the mean histogram of all enterprise histograms is calculated (the data types are the same, and the dimensions of the histograms are the same). The mean histogram is equivalent to a mean array. The values ​​of each indicator in the mean histogram are input into a preset indicator calculation model, and the baseline innovation value of the enterprise category is output. The product feature map of the enterprise is read, and the area of ​​each mapping shape is accumulated according to a preset decreasing weight. The additional innovation value is determined according to the area, and the additional innovation value is proportional to the area. Finally, the baseline innovation value and the additional innovation value are summed to obtain the innovation capability value.

[0093] Figure 5 The fourth sub-process flowchart of the technology enterprise innovation capability evaluation method based on big data model includes the following steps: constructing a sample set by using basic information and product information as features and innovation capability values ​​as labels; training a neural network model based on the sample set; and using the neural network model as the capability evaluation scheme when the error rate of the neural network model is less than a preset threshold.

[0094] Step S401: Read the basic information and product information of each enterprise, and read the innovation capability value of the enterprise;

[0095] Step S402: Construct a sample set by using basic information and product information as features and innovation capability values ​​as labels, and train a neural network model based on the sample set;

[0096] Step S403: Read the error rate of the neural network model. When the error rate is less than a preset threshold, determine the proportion of application time of the neural network model within a preset time period based on the error rate.

[0097] In one example of the technical solution of the present invention, the basic information and product information of each enterprise are read, the innovation capability value of the enterprise is read, the basic information and product information are used as features, and the innovation capability value is used as a label to construct a sample set. A neural network model is trained based on the sample set. The sample set is divided into a training set and a test set. The neural network model is trained, and the error rate of the neural network model is read. When the error rate is less than a preset threshold, the application time ratio of the neural network model within a preset time period is determined according to the error rate.

[0098] The meaning of the time percentage is based on the execution cycle. For example, if the execution cycle is one hour, the error rate is calculated and subtracted to obtain the accuracy rate. This hour is multiplied by the accuracy rate to obtain the application time of the neural network model. The remaining process uses the original evaluation algorithm, namely step S300.

[0099] As a preferred embodiment of the technical solution of the present invention, the method further includes:

[0100] The system displays the enterprise histogram and product feature map based on a pre-defined UI interface.

[0101] It receives encryption requests from enterprises containing encryption targets and encrypts the content of enterprise histograms and product feature maps.

[0102] Since this invention has converted basic information into enterprise histograms and product information into product feature maps, the display process is very easy and can be directly integrated with the UI interface. On this basis, this application introduces an encryption process, which receives an encryption request from the enterprise containing the encryption target, and encrypts the content of the enterprise histogram and product feature map. The encryption target is the part to be encrypted.

[0103] Figure 6 This is a structural block diagram of a technology enterprise innovation capability evaluation system based on a big data model. In this embodiment of the invention, a technology enterprise innovation capability evaluation system based on a big data model, system 10, includes:

[0104] The indicator statistics module 11 is used to obtain basic information of enterprises based on preset statistical standards, perform data transcoding on the basic information to obtain data values, and generate an enterprise histogram based on the data values; the statistical standards are used to characterize preset indicators.

[0105] Product information statistics module 12 is used to acquire enterprise product information, perform two-dimensional processing on the product information, and obtain product feature map;

[0106] The capability value calculation module 13 is used to determine the innovation capability value of an enterprise based on its histogram and product characteristics.

[0107] The model training application module 14 is used to construct a sample set by using basic information and product information as features and innovation capability values ​​as labels. The neural network model is trained based on the sample set. When the error rate of the neural network model is less than a preset threshold, the neural network model is used as a capability evaluation scheme.

[0108] Furthermore, the indicator statistics module 11 includes:

[0109] The statistical caliber generation unit is used to receive the indicator type and its weight input by the user, and use it as the statistical caliber.

[0110] The data acquisition unit is used to send information acquisition requests to enterprises and receive data from enterprises regarding each indicator.

[0111] The data transcoding unit is used to select a transcoding model based on the data structure of the indicator, input the data into the transcoding model, and obtain the data value of each indicator; the transcoding model contains an indicator positiveization function;

[0112] The histogram generation unit is used to statistically analyze data values ​​according to a preset indicator order to obtain the enterprise histogram; during the data value statistics process, the data values ​​are adjusted based on weights.

[0113] Specifically, the product information statistics module 12 includes:

[0114] The acquisition execution unit is used to acquire information such as the enterprise's product types, product sales revenue, and product release time.

[0115] A shape mapping unit is used to determine the mapping shape based on the product type and the shape size based on the product sales revenue. The mapping shape is a regular polygon. The number of sides of the regular polygon is inversely proportional to the regularity of the product type. The regularity of different product types is a preset value and stored in a preset product statistics table. The process of determining the product statistics table is as follows: enterprises are classified according to enterprise type. For each type of enterprise, the total number of enterprises corresponding to each product type is calculated. The regularity of the product type is determined based on the total number of enterprises, and the regularity is directly proportional to the total number of enterprises. The product types and their regularities are statistically analyzed to obtain the product statistics table.

[0116] The mapping execution unit is used to determine the statistical order based on the product release time, and to count all the mapping shapes of the enterprise based on the statistical order to obtain the product feature map.

[0117] Furthermore, the capability value calculation module 13 includes:

[0118] The clustering unit is used to read the enterprise histograms of all enterprises, calculate the enterprise similarity, and cluster the enterprises based on the enterprise similarity.

[0119] The mean calculation unit is used to calculate the mean histogram of all enterprises for each type of enterprise.

[0120] The indicator analysis unit is used to input the values ​​of each indicator in the mean histogram into a preset indicator calculation model and output the benchmark innovation value of this type of enterprise.

[0121] The added value calculation unit is used to read the product feature map of the enterprise, accumulate the area of ​​each mapped shape according to the preset decreasing weight, and determine the added innovation value based on the area; the added innovation value is proportional to the area.

[0122] The numerical summation unit is used to sum the baseline innovation value and the additional innovation value to obtain the innovation capability value.

[0123] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for evaluating the innovation capability of technology enterprises based on a big data model, characterized in that, The method includes: Basic information of enterprises is obtained based on preset statistical criteria, the basic information is transcoded to obtain data values, and an enterprise histogram is generated based on the data values; the statistical criteria are used to characterize preset indicators. Obtain the company's product information, perform two-dimensional processing on the product information, and obtain a product feature map; Determine the innovation capability value of an enterprise based on its histogram and product characteristics; A sample set is constructed by using basic information and product information as features and innovation capability values ​​as labels. A neural network model is trained based on the sample set. When the error rate of the neural network model is less than a preset threshold, the neural network model is used as the capability assessment scheme.

2. The method for evaluating the innovation capability of technology enterprises based on a big data model according to claim 1, characterized in that, The steps of obtaining basic enterprise information based on a preset statistical standard, transcoding the basic information to obtain data values, and generating an enterprise histogram based on the data values ​​include: Receive user input of indicator types and their weights as statistical criteria; Send information request to enterprises and receive data from enterprises regarding each indicator; A transcoding model is selected based on the data structure of the indicator. The data is input into the transcoding model to obtain the data value of each indicator. The transcoding model contains an indicator positiveization function. The enterprise histogram is obtained by statistically analyzing the data values ​​according to the preset indicator order; during the data value statistics process, the data values ​​are adjusted based on weights.

3. The method for evaluating the innovation capability of technology enterprises based on a big data model according to claim 1, characterized in that, The steps of obtaining the enterprise's product information and performing two-dimensional processing on the product information to obtain a product feature map include: Obtain information on a company's product types, sales revenue, and product launch dates; The mapping shape is determined based on the product type, and the shape size is determined based on the product sales revenue. The mapping shape is a regular polygon. The number of sides of the regular polygon is inversely proportional to the regularity of the product type. The regularity of different product types is a preset value and stored in a preset product statistics table. The process of determining the product statistics table is as follows: enterprises are classified according to enterprise type. For each type of enterprise, the total number of enterprises corresponding to each product type is calculated. The regularity of the product type is determined based on the total number of enterprises, and the regularity is directly proportional to the total number of enterprises. The product types and their regularities are statistically analyzed to obtain the product statistics table. The statistical order is determined based on the product release time, and all mapping shapes of the enterprise are statistically analyzed based on the statistical order to obtain the product feature map.

4. The method for evaluating the innovation capability of technology enterprises based on a big data model according to claim 1, characterized in that, The steps for determining a company's innovation capability value based on its histogram and product characteristics include: Read the enterprise histograms of all enterprises, calculate the enterprise similarity, and cluster the enterprises based on the enterprise similarity; For each type of enterprise, calculate the mean histogram of all enterprises' histograms; Input the values ​​of each indicator in the mean histogram into the preset indicator calculation model, and output the benchmark innovation value of this type of enterprise. Read the product feature map of the enterprise, accumulate the area of ​​each mapped shape according to the preset decreasing weight, and determine the additional innovation value based on the area; the additional innovation value is proportional to the area. The innovation capability value is obtained by summing the baseline innovation value and the additional innovation value.

5. The method for evaluating the innovation capability of technology enterprises based on a big data model according to claim 1, characterized in that, The steps of constructing a sample set by using basic information and product information as features and innovation capability values ​​as labels, training a neural network model based on the sample set, and using the neural network model as a capability assessment scheme when the error rate of the neural network model is less than a preset threshold include: Read the basic information and product information of each enterprise, and read the enterprise's innovation capability value; A sample set is constructed using basic information and product information as features and innovation capability values ​​as labels, and a neural network model is trained based on the sample set. The error rate of the neural network model is read. When the error rate is less than a preset threshold, the application time of the neural network model within a preset time period is determined based on the error rate.

6. The method for evaluating the innovation capability of technology enterprises based on a big data model according to claim 1, characterized in that, The method further comprises: The system displays the enterprise histogram and product feature map based on a pre-defined UI interface. It receives encryption requests from enterprises containing encryption targets and encrypts the content of enterprise histograms and product feature maps.

7. A technology enterprise innovation capability evaluation system based on a big data model, characterized in that, The system includes: The indicator statistics module is used to obtain basic information of enterprises based on preset statistical standards, perform data transcoding on the basic information to obtain data values, and generate enterprise histograms based on the data values; the statistical standards are used to characterize preset indicators. The product information statistics module is used to acquire the company's product information, perform two-dimensional processing on the product information, and obtain a product feature map. The capability value calculation module is used to determine the innovation capability value of an enterprise based on its histogram and product characteristics. The model training application module is used to construct a sample set by using basic information and product information as features and innovation capability values ​​as labels. The neural network model is trained based on the sample set. When the error rate of the neural network model is less than a preset threshold, the neural network model is used as a capability assessment scheme.

8. The technology enterprise innovation capability evaluation system based on a big data model according to claim 7, characterized in that, The indicator statistics module includes: The statistical caliber generation unit is used to receive the indicator type and its weight input by the user, and use it as the statistical caliber. The data acquisition unit is used to send information acquisition requests to enterprises and receive data from enterprises regarding each indicator. The data transcoding unit is used to select a transcoding model based on the data structure of the indicator, input the data into the transcoding model, and obtain the data value of each indicator; the transcoding model contains an indicator positiveization function; The histogram generation unit is used to statistically analyze data values ​​according to a preset indicator order to obtain the enterprise histogram; during the data value statistics process, the data values ​​are adjusted based on weights.

9. The technology enterprise innovation capability evaluation system based on a big data model according to claim 7, characterized in that, The product information statistics module includes: The acquisition execution unit is used to acquire information such as the enterprise's product types, product sales revenue, and product release time. A shape mapping unit is used to determine the mapping shape based on the product type and the shape size based on the product sales revenue. The mapping shape is a regular polygon. The number of sides of the regular polygon is inversely proportional to the regularity of the product type. The regularity of different product types is a preset value and stored in a preset product statistics table. The process of determining the product statistics table is as follows: enterprises are classified according to enterprise type. For each type of enterprise, the total number of enterprises corresponding to each product type is calculated. The regularity of the product type is determined based on the total number of enterprises, and the regularity is directly proportional to the total number of enterprises. The product types and their regularities are statistically analyzed to obtain the product statistics table. The mapping execution unit is used to determine the statistical order based on the product release time, and to count all the mapping shapes of the enterprise based on the statistical order to obtain the product feature map.

10. The technology enterprise innovation capability evaluation system based on a big data model according to claim 7, characterized in that, The capability value calculation module includes: The clustering unit is used to read the enterprise histograms of all enterprises, calculate the enterprise similarity, and cluster the enterprises based on the enterprise similarity. The mean calculation unit is used to calculate the mean histogram of all enterprises for each type of enterprise. The indicator analysis unit is used to input the values ​​of each indicator in the mean histogram into a preset indicator calculation model and output the benchmark innovation value of this type of enterprise. The added value calculation unit is used to read the product feature map of the enterprise, accumulate the area of ​​each mapped shape according to the preset decreasing weight, and determine the added innovation value based on the area; the added innovation value is proportional to the area. The numerical summation unit is used to sum the baseline innovation value and the additional innovation value to obtain the innovation capability value.

Images