Enterprise portrait calculation method, medium and program based on big data and multi-dimensional characteristics

Abstract

The present invention discloses a method, medium and program for calculating enterprise portraits based on big data and multidimensional features. The method includes the following steps: obtaining enterprise data, and performing clustering operations on the enterprise data to generate enterprise labels; using a label scoring model to score the enterprise labels corresponding to each enterprise, and generating a scoring result; based on the scoring result, obtaining the weight of each enterprise label; constructing a first enterprise portrait based on the enterprise labels and the corresponding weights; obtaining the influence factor of each enterprise label, and scoring the influence factor through an influence factor scoring model to obtain the value of the influence factor; constructing a second enterprise portrait based on the influence factor of each enterprise label and the corresponding value; combining the first enterprise portrait with the second enterprise portrait to generate a third enterprise portrait. The present invention improves the comprehensiveness and accuracy of enterprise portrait depiction, and facilitates the enterprise service industry to provide more accurate services to enterprises.

Description

Technical Field

[0001] The present invention relates to the field of big data technology and data mining, and in particular to a method, medium and program for calculating enterprise portraits based on big data and multi-dimensional features. Background Art

[0002] With the advent of the big data era, a large amount of enterprise-related data information has given rise to the concept of enterprise portraits. The so-called enterprise portrait is to label the information describing the status and behavior of the enterprise, and to "label" the enterprise based on the data related to the enterprise status and behavior information, so as to portray the overall picture of the enterprise; enterprises are the main participants in market economic activities, and the enterprise service industry is the life steward of the enterprise. More accurate portrayal of enterprise portraits has become a key issue in the enterprise service industry. Among them, the application number 202010660269.8 discloses a method for evaluating enterprise portraits based on association rules, which obtains results by analyzing the five dimensions of the company according to different weights, and analyzes the ten major related enterprises of the enterprise to be profiled by the AHP hierarchical analysis method; the application document uses the weight calculation of the five dimensions to obtain related enterprises, but does not add the numerical description of the label to the enterprise portrait. The application number 202011266790.X discloses a method for constructing enterprise portraits, which constructs enterprise portraits through data in the enterprise family relationship database and data in the enterprise information database. In the enterprise portrait constructed by this patent, the label is not accurately described using numerical values, and the accuracy of the enterprise portrait is not high. Summary of the invention

[0003] In view of this, the embodiments of the present application provide a method, medium and program for calculating enterprise portraits based on big data and multi-dimensional features, which improves the comprehensiveness and accuracy of enterprise portrait depiction and facilitates the enterprise service industry to provide more accurate services to enterprises.

[0004] The present application provides a method for calculating an enterprise portrait based on big data and multi-dimensional features, the method comprising:

[0005] Acquire enterprise data, and perform clustering operations on the enterprise data to generate enterprise tags;

[0006] Use the label scoring model to score the enterprise label corresponding to each enterprise and generate a scoring result;

[0007] Based on the scoring results, obtain the weight of each of the enterprise tags;

[0008] Based on the enterprise tags and corresponding weights, construct a first enterprise portrait;

[0009] Obtaining the impact factor of each enterprise label, and scoring the impact factor using an impact factor scoring model to obtain the value of the impact factor;

[0010] Based on the impact factor and corresponding value of each enterprise tag, a second enterprise portrait is constructed;

[0011] Combine the first corporate portrait and the second corporate portrait to generate a third corporate portrait.

[0012] In one embodiment, obtaining the weight of each of the enterprise tags based on the scoring result includes:

[0013] Compare the scoring results of each enterprise tag to obtain a comparison result;

[0014] According to the comparison results, construct a priority diagram weight calculation table;

[0015] Based on the priority diagram weight calculation table, the weight of each enterprise label is obtained.

[0016] In one embodiment, the comparing the scoring results of each enterprise tag to obtain the comparison result includes:

[0017] Compare the scoring results of each enterprise tag in pairs;

[0018] If, of the two enterprise tags, the scoring result of the first enterprise tag is greater than the scoring result of the second enterprise tag, the first comparison result obtained is the first set value;

[0019] If, of the two enterprise tags, the scoring result of the first enterprise tag is less than the scoring result of the second enterprise tag, the second comparison result obtained is the second set value;

[0020] If, of the two enterprise tags, the scoring result of the first enterprise tag is equal to the scoring result of the second enterprise tag, the third comparison result obtained is a third set value; wherein the first set value, the second set value and the third set value are all different.

[0021] In one embodiment, obtaining the weight of each enterprise tag based on the priority graph weight calculation table includes:

[0022] Obtaining the sum of the comparison results of each column in the priority diagram weight calculation table as the importance of the corresponding enterprise label;

[0023] The proportion of the importance of each of the enterprise tags to the total of the comparison results in the priority weight calculation table is calculated to obtain the weight of each of the enterprise tags.

[0024] In one embodiment, before the step of using the label scoring model to score the enterprise label corresponding to each enterprise and generating the scoring result, the step includes:

[0025] Constructing the label scoring model specifically includes:

[0026] Preprocessing the enterprise data corresponding to the enterprise tag to generate input data in a preset format;

[0027] The input data is input into a neural network for supervised training to generate the label scoring model.

[0028] In one embodiment, the preprocessing of the enterprise data corresponding to the enterprise tag to generate input data in a preset format includes:

[0029] Cleaning and counting the enterprise data corresponding to the enterprise tag;

[0030] According to the statistical results, the scoring value corresponding to the enterprise label is obtained;

[0031] Based on the enterprise label, the score value, and the corresponding relationship between the enterprise label and the score value, input data of the label scoring model is generated.

[0032] In one embodiment, the enterprise tags include at least: innovation capability, risk resistance capability, and management capability.

[0033] In one embodiment, before the step of obtaining the impact factor of each enterprise tag and scoring the impact factor using an impact factor scoring model to obtain the value of the impact factor, the step includes:

[0034] Constructing the impact factor scoring model specifically includes:

[0035] Clean and count the data corresponding to the impact factors;

[0036] Based on the statistical results, a score value corresponding to the impact factor is obtained;

[0037] Based on the impact factor, the score value, and the corresponding relationship between the impact factor and the score value, the impact factor is input into a neural network for supervised training to generate the impact factor scoring model.

[0038] In order to achieve the above-mentioned purpose, a computer storage medium is also provided, on which a program for a method for calculating a corporate portrait based on big data and multi-dimensional features is stored. When the program for calculating a method for calculating a corporate portrait based on big data and multi-dimensional features is executed by a processor, any of the steps of the method for calculating a corporate portrait based on big data and multi-dimensional features described above is implemented.

[0039] In order to achieve the above-mentioned purpose, a computer program product is also provided, including a computer program, which, when executed by a processor, implements the steps of any of the above-mentioned methods for calculating corporate portraits based on big data and multi-dimensional features.

[0040] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: obtaining enterprise data and performing clustering operations on the enterprise data to generate enterprise labels; by obtaining a large amount of enterprise data and performing clustering operations to generate enterprise labels, the enterprise labels are made more comprehensive, and the enterprise image can be portrayed from multiple dimensions, thereby improving the comprehensiveness of the enterprise portrait.

[0041] The label scoring model is used to score the enterprise label corresponding to each enterprise to generate a scoring result; the enterprise label is scored by the trained label scoring model to improve the accuracy of the scoring result.

[0042] Based on the scoring results, the weight of each enterprise label is obtained; based on the enterprise labels and the corresponding weights, a first enterprise portrait is constructed. Through the scoring results, the weight of each enterprise label is obtained using the priority diagram method, that is, the advantages and disadvantages of the enterprise label of each enterprise are obtained, and the generated first enterprise portrait is used to more comprehensively portray the enterprise portrait.

[0043] Obtain the impact factor of each enterprise label, and score the impact factor through the impact factor scoring model to obtain the value of the impact factor; construct a second enterprise portrait based on the impact factor of each enterprise label and the corresponding value; combine the first enterprise portrait with the second enterprise portrait to generate a third enterprise portrait; by combining the comprehensive first enterprise portrait and the more detailed second enterprise portrait, the generated third enterprise portrait can more comprehensively and accurately portray the enterprise image, which is convenient for the enterprise to optimize itself, and also helps the enterprise service industry to provide more accurate services to enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a flowchart of the first embodiment of the enterprise portrait calculation method based on big data and multi-dimensional features of the present application;

[0045] Figure 2 This is the specific implementation step of step S130 of the first embodiment of the enterprise portrait calculation method based on big data and multi-dimensional features of this application;

[0046] Figure 3 The specific implementation steps of step S131 of the enterprise portrait calculation method based on big data and multi-dimensional features of this application;

[0047] Figure 4This is the specific implementation step of step S133 of the enterprise portrait calculation method based on big data and multi-dimensional features of this application;

[0048] Figure 5 This is a flowchart of the second embodiment of the enterprise portrait calculation method based on big data and multi-dimensional features of the present application;

[0049] Figure 6 The specific implementation steps of step S220 in the second embodiment of the enterprise portrait calculation method based on big data and multi-dimensional features of this application;

[0050] Figure 7 The specific implementation steps of step S221 of the enterprise portrait calculation method based on big data and multi-dimensional features of this application;

[0051] Figure 8 This is a flowchart of a third embodiment of the enterprise portrait calculation method based on big data and multi-dimensional features of the present application;

[0052] Fig. 9 This is a schematic diagram of the hardware architecture of the enterprise portrait generation device based on big data and multi-dimensional features involved in the embodiments of the present application. DETAILED DESCRIPTION

[0053] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0054] The main solution of the embodiment of the present invention is: obtain enterprise data, and perform clustering operations on the enterprise data to generate enterprise labels; use the label scoring model to score the enterprise labels corresponding to each enterprise and generate a scoring result; based on the scoring result, obtain the weight of each enterprise label; based on the enterprise labels and the corresponding weights, construct a first enterprise portrait; obtain the influence factor of each enterprise label, and score the influence factor through the influence factor scoring model to obtain the value of the influence factor; based on the influence factor of each enterprise label and the corresponding value, construct a second enterprise portrait; combine the first enterprise portrait and the second enterprise portrait to generate a third enterprise portrait. The present invention improves the comprehensiveness and accuracy of enterprise portrait portrayal, making it easier for the enterprise service industry to provide more accurate services to enterprises.

[0055] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0056] Reference Figure 1 , Figure 1 This is the first embodiment of the enterprise portrait calculation method based on big data and multi-dimensional features of this application, and the method includes:

[0057] Step S110: Acquire enterprise data, and perform clustering operations on the enterprise data to generate enterprise tags.

[0058] Specifically, enterprise data may be data about an enterprise stored on the Internet, and may be obtained from the official website of the enterprise or from a third-party enterprise information management website, without limitation herein; enterprise data may be obtained through web crawler technology.

[0059] The clustering operation can be a text-related clustering algorithm, specifically an LDA topic model or a K-Means clustering algorithm; the K-Means clustering algorithm is an unsupervised clustering algorithm, which first divides the sample set into K clusters, so that the distance between clusters becomes larger and larger through iterations, and the samples in the cluster should be as close as possible. The specific algorithm is as follows:

[0060] Algorithm input: data set X, number of clusters K;

[0061] Algorithm output: cluster set C;

[0062] S1: Randomly select K sample objects from the entire data set X;

[0063] S2: Calculate the distance from each sample Xm in the data set to the cluster center Ci;

[0064] S3: Find the minimum distance from each sample object Xm to the cluster center Ci, and classify the sample object Xm into the same cluster as Ci;

[0065] S4: Calculate the mean of objects in the same cluster and update the cluster center;

[0066] S5: Repeat S2-S4, and the algorithm iterates until the cluster center object does not change.

[0067] It should be noted that enterprise labels can also be set according to user needs.

[0068] Step S120: Use the label scoring model to score the enterprise label corresponding to each enterprise and generate a scoring result.

[0069] Specifically, the label scoring model can be a mathematical prediction model or a neural network prediction model, which is not limited here.

[0070] Step S130: Based on the scoring result, obtain the weight of each enterprise tag.

[0071] Specifically, the weights of enterprise labels may be obtained by using a priority diagram method, or by using a hierarchical analysis method, which is not limited here.

[0072] Step S140: Construct a first enterprise portrait based on the enterprise tags and corresponding weights.

[0073] Specifically, the enterprise labels and corresponding weights generated by clustering are used to vividly depict the first enterprise portrait using weight values, making the first enterprise portrait more accurate.

[0074] Step S150: Obtain the impact factor of each enterprise label, and score the impact factor using an impact factor scoring model to obtain the value of the impact factor.

[0075] Specifically, the impact factor scoring model can be constructed according to the construction method of the label scoring model. Specifically, the impact factors of the enterprise label can be multiple. For example, when the enterprise label is innovation ability, the impact factors corresponding to the innovation ability can be technical research and development ability, core member examples, patent output, intellectual property accumulation, and qualification certification; in addition, an impact factor can be determined by multiple attributes, such as technical research and development ability can be determined by registered capital, number of insured persons, number of branches, high-tech enterprises, number of patents, and number of software copyrights; while the strength of core members can be determined by the number of core team members; patent output can be determined by the number of patents and the number of patents per capita; intellectual property accumulation can be determined by the number of patents, the number of software copyrights, and the number of trademarks; qualification certification can be determined by whether it is a high-tech enterprise, the number of qualification certificates, and the number of website filings. When the enterprise label is business ability, the impact factors corresponding to business ability can be team strength, financing ability, and investment strength; among which team strength can be determined by the number of core teams and the education of members; financing ability can be determined by the number of financings and the current financing round; investment strength can be the total amount of foreign investment and the number of foreign investments. When the enterprise label is risk resistance, the influencing factors corresponding to risk resistance may be media exposure, qualification certificate, bidding activity, judicial risk, business abnormalities, administrative penalties, execution status and capital mortgage status; among them, media exposure can be determined by the number of news opinions and the number of self-media; qualification certificate can be determined by the number of qualification certificates; bidding activity can be determined by the number of biddings; judicial risk can be determined by the number of legal proceedings; business abnormalities can be determined by the number of business abnormalities; administrative penalties can be determined by the number of administrative penalties; execution status can be determined by the number of executions; pledge and mortgage status can be determined by the number of pledges and mortgages.

[0076] Step S160: Construct a second enterprise portrait based on the impact factor and corresponding value of each enterprise tag.

[0077] Specifically, further using the influencing factors of each enterprise label and its corresponding value can characterize the enterprise portrait in more detail and specifically, and generate a second enterprise portrait, that is, further refine each enterprise label.

[0078] Step S170: Generate a third corporate portrait by combining the first corporate portrait and the second corporate portrait.

[0079] By combining enterprise labels and their weights, as well as the influencing factors and corresponding values ​​of enterprise labels, the enterprise portrait is portrayed to make the enterprise portrait more comprehensive and specific. By depicting the enterprise portrait through the weights of enterprise labels and the values ​​of influencing factors, the advantages and disadvantages of the enterprise are more prominent.

[0080] In the above embodiment, the beneficial effects are: obtaining enterprise data and performing clustering operations on the enterprise data to generate enterprise labels; by obtaining a large amount of enterprise data and performing clustering operations to generate enterprise labels, the enterprise labels are made more comprehensive, and the enterprise image can be portrayed from multiple dimensions, thereby improving the comprehensiveness of the enterprise portrait.

[0081] The label scoring model is used to score the enterprise label corresponding to each enterprise to generate a scoring result; the enterprise label is scored by the trained label scoring model to improve the accuracy of the scoring result.

[0082] Based on the scoring results, the weight of each enterprise label is obtained; based on the enterprise labels and the corresponding weights, a first enterprise portrait is constructed. Through the scoring results, the weight of each enterprise label is obtained using the priority diagram method, that is, the advantages and disadvantages of the enterprise label of each enterprise are obtained, and the generated first enterprise portrait is used to more comprehensively portray the enterprise portrait.

[0083] Obtain the influence factor of each enterprise label, and score the influence factor through the influence factor scoring model to obtain the value of the influence factor; construct a second enterprise portrait based on the influence factor of each enterprise label and the corresponding value; combine the first enterprise portrait with the second enterprise portrait to generate a third enterprise portrait; by combining the comprehensive first enterprise portrait and the more detailed second enterprise portrait, the generated third enterprise portrait can more comprehensively and accurately portray the enterprise image, which is convenient for the enterprise to optimize itself, and also helps the enterprise service industry to provide more accurate services to the enterprise.

[0084] Reference Figure 2 , Figure 2 The specific implementation steps of step S130 of the first embodiment of the enterprise portrait calculation method based on big data and multi-dimensional features of this application are as follows: based on the scoring results, the weight of each enterprise label is obtained by using the priority diagram method, including:

[0085] Step S131: Compare the scoring results of each enterprise tag to obtain a comparison result.

[0086] Step S132: constructing a priority diagram weight calculation table according to the comparison result.

[0087] Step S133: Based on the priority diagram weight calculation table, obtain the weight of each enterprise tag.

[0088] Specifically, the priority graph method compares multiple indicators or goals in pairs and finally gives an order of importance or priority. In this embodiment, the priority graph method is used to obtain a priority graph weight calculation table, and the weight of each enterprise label is obtained according to the priority graph weight calculation table to reflect the importance of each enterprise label among all labels, more accurately reflect the strengths and weaknesses of the enterprise, provide a more accurate optimization direction in the process of enterprise development, and promote the rapid development of the enterprise.

[0089] Reference Figure 3 , Figure 3 The specific implementation steps of step S131 of the enterprise portrait calculation method based on big data and multi-dimensional features of this application are to compare the scoring results of each enterprise label to obtain the comparison results, including:

[0090] Step S131 - 1: performing pairwise comparisons on the scoring results of each enterprise tag;

[0091] Step S131 - 2 : If, of the two enterprise tags, the scoring result of the first enterprise tag is greater than the scoring result of the second enterprise tag, the first comparison result obtained is a first set value; wherein the first set value may be 1.

[0092] If, of the two enterprise tags, the scoring result of the first enterprise tag is smaller than the scoring result of the second enterprise tag, the second comparison result obtained is the second set value; wherein the second set value may be 0.5.

[0093] If the scoring result of the first enterprise tag is equal to the scoring result of the second enterprise tag, the third comparison result is a third set value, where the third set value may be 0.

[0094] The first setting value, the second setting value and the third setting value are all different.

[0095] Specifically, when the score result of enterprise label 1 is 81, the score result of enterprise label 2 is 67, and the score result of enterprise label 3 is 59, the generated priority graph weight calculation table is shown in Table 1 below:

[0096]

[0097]

[0098] Table 1

[0099] Reference Figure 4 , Figure 4 The specific implementation steps of step S133 of the enterprise portrait calculation method based on big data and multi-dimensional features of this application, based on the priority diagram weight calculation table, obtain the weight of each enterprise label, including:

[0100] Step S133 - 1 : Obtain the sum of the comparison results of each column in the priority diagram weight calculation table as the importance of the corresponding enterprise label.

[0101] Step S133 - 2 : Calculate the proportion of the importance of each of the enterprise tags to the total of the comparison results in the priority weight calculation table to obtain the weight of each of the enterprise tags.

[0102] Specifically, as shown in Table 1, the importance of enterprise label 1 is 2.5, the importance of enterprise label 2 is 1.5, and the importance of enterprise label 3 is 0.5; the weight of each enterprise label is obtained by calculating the comprehensive proportion of the comparison results of the importance of each enterprise label in the priority weight calculation table, and the weight of enterprise label 1 is 55.6%, the weight of enterprise label 2 is 33.3%, and the weight of enterprise label 3 is 11.1%.

[0103] In the above embodiment, the beneficial effects are as follows: the weights of enterprise labels obtained by the priority diagram method are more reasonable, and the calculation process is simple, which further improves the calculation efficiency of enterprise portrait generation.

[0104] Reference Figure 5 , Figure 5 This is the second embodiment of the enterprise portrait calculation method based on big data and multi-dimensional features of this application. Before the step of using the label scoring model to score the enterprise label corresponding to each enterprise and generating the scoring result, it includes:

[0105] Step S210: Acquire enterprise data, and perform clustering operations on the enterprise data to generate enterprise tags.

[0106] Step S220: constructing the label scoring model;

[0107] Specifically, the label scoring model can be a sequence-to-sequence model based on a long short-term memory network, wherein the long short-term memory network can obtain feature information of the context in the enterprise data, further improving the accuracy of the scoring results.

[0108] Step S230: Use the label scoring model to score the enterprise label corresponding to each enterprise and generate a scoring result.

[0109] Step S240: Based on the scoring result, obtain the weight of each enterprise tag.

[0110] Step S250: Construct a first enterprise portrait based on the enterprise tags and corresponding weights.

[0111] Step S260: Obtain the impact factor of each enterprise label, and score the impact factor using an impact factor scoring model to obtain the value of the impact factor.

[0112] Step S270: Construct a second enterprise portrait based on the impact factor and corresponding value of each enterprise tag.

[0113] Step S280: Generate a third corporate portrait by combining the first corporate portrait and the second corporate portrait.

[0114] Compared with the first embodiment, the second embodiment includes step S220, and the other steps have been described in the first embodiment and will not be described again.

[0115] In the above embodiment, the beneficial effect is as follows: by constructing a label scoring model based on a sequence-to-sequence model of a long short-term memory network, the scoring results of enterprise labels are made more reliable, further improving the accuracy of enterprise portrait portrayal.

[0116] Reference Figure 6 , Figure 6 The specific implementation steps of step S220 in the second embodiment of the enterprise portrait calculation method based on big data and multi-dimensional features of this application are as follows:

[0117] Step S221: pre-processing the enterprise data corresponding to the enterprise tag to generate input data in a preset format.

[0118] Specifically, the preset format may be a correspondence between an enterprise label and a scoring value, wherein the scoring value may be obtained based on manual labeling or through statistics of the amount of data corresponding to the enterprise label.

[0119] Step S222: input the input data into a neural network for supervised training to generate the label scoring model.

[0120] Specifically, the neural network model can be a neural network prediction model, wherein the neural network prediction model uses a neural network model to predict unknown data. The neural network prediction model includes but is not limited to a convolutional neural network (CNN), a recurrent neural network (RNN), a long short-term memory network (LSTM), etc.

[0121] Reference Figure 7 , Figure 7The specific implementation steps of step S221 of the enterprise portrait calculation method based on big data and multi-dimensional features of this application are as follows: the enterprise data corresponding to the enterprise tag is preprocessed to generate input data in a preset format, including:

[0122] Step S221 - 1 : Clean and count the enterprise data corresponding to the enterprise tag.

[0123] Specifically, performing data cleansing on enterprise data may be performing a deduplication operation on the enterprise data.

[0124] Step S221 - 2: Obtain the score value corresponding to the enterprise tag according to the statistical results.

[0125] Specifically, for example, if the data volume of enterprise label 1 of enterprise A is 200K, a score value can be set according to the size of the data volume. The same is true for other enterprise labels, that is, the larger the data volume corresponding to the enterprise label, the larger the corresponding score value.

[0126] Step S221 - 3 : Based on the enterprise label, the score value, and the corresponding relationship between the enterprise label and the score value, generate input data for the label scoring model.

[0127] Specifically, the input data conforms to a preset format. In this embodiment, the enterprise label, the scoring value, and the corresponding relationship between the enterprise label and the scoring value may be used.

[0128] In one embodiment, the enterprise labels include at least: innovation capability, risk resistance capability, and management capability.

[0129] Specifically, corporate labels are not limited to the above-mentioned innovation capabilities, risk resistance capabilities, and operational capabilities.

[0130] Reference Figure 8 , Figure 8 This is the third embodiment of the enterprise portrait calculation method based on big data and multi-dimensional features of the present application, and the method further includes:

[0131] Step S310: Acquire enterprise data, and perform clustering operations on the enterprise data to generate enterprise tags.

[0132] Step S320: Use the label scoring model to score the enterprise label corresponding to each enterprise and generate a scoring result.

[0133] Step S330: Based on the scoring result, obtain the weight of each enterprise tag.

[0134] Step S340: Construct a first enterprise portrait based on the enterprise tags and corresponding weights.

[0135] Step S350: constructing the impact factor scoring model.

[0136] Specifically, the impact factor scoring model can be a sequence-to-sequence model based on a long short-term memory network, wherein the long short-term memory network can obtain feature information of the context in the enterprise data, further improving the accuracy of the scoring results.

[0137] Constructing the impact factor scoring model specifically includes:

[0138] S1: Clean and count the data corresponding to the impact factors;

[0139] S2: Based on the statistical results, obtain the score value corresponding to the impact factor;

[0140] S3: Based on the impact factor, the score value, and the corresponding relationship between the impact factor and the score value, the impact factor is input into a neural network for supervised training to generate the impact factor scoring model.

[0141] Specifically, the construction process of the impact factor scoring model is similar to that of the label scoring model, and the label scoring model can be referred to.

[0142] Step S360: Obtain the impact factor of each enterprise tag, and score the impact factor using an impact factor scoring model to obtain the value of the impact factor.

[0143] Step S370: Construct a second enterprise portrait based on the impact factor and corresponding value of each enterprise tag.

[0144] Step S380: Combine the first corporate portrait and the second corporate portrait to generate a third corporate portrait.

[0145] Compared with the first embodiment, the third embodiment includes step S350, and the other steps have been described in the first embodiment and will not be described again.

[0146] In the above embodiment, there is a beneficial effect: by constructing a neural network impact factor scoring model, the scoring accuracy of the impact factor is improved, and the accurate portrayal of the second enterprise portrait is guaranteed, thereby ensuring the accuracy of the third enterprise portrait.

[0147] The present application also protects a computer storage medium, on which is stored a program for a method for calculating a corporate portrait based on big data and multi-dimensional features. When the program for calculating a corporate portrait based on big data and multi-dimensional features is executed by a processor, the program implements any of the steps of the method for calculating a corporate portrait based on big data and multi-dimensional features described above.

[0148] The present application also protects a computer program product, including a computer program, which, when executed by a processor, implements the steps of any of the above-mentioned methods for calculating corporate portraits based on big data and multi-dimensional features.

[0149] This application relates to a corporate portrait generation device based on big data and multi-dimensional features 010 including: Fig. 9 Shown: at least one processor 012 and a memory 011.

[0150] Processor 012 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in processor 012 or instructions in software form. The above processor 012 may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The disclosed methods, steps and logic block diagrams in the embodiments of the present invention can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The software module may be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 011, and the processor 012 reads the information in the memory 011 and completes the steps of the above method in combination with its hardware.

[0151] It can be understood that the memory 011 in the embodiment of the present invention can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct RAM bus random access memory (DRRAM). The memory 011 of the system and method described in the embodiments of the present invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0152] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0153] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0154] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0155] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0156] It should be noted that in the claims, any reference signs placed between brackets shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claim. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention may be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third etc. does not indicate any order. These words may be interpreted as names.

[0157] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0158] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for calculating enterprise portraits based on big data and multi-dimensional features, characterized in that: The method comprises: Acquire enterprise data, and perform clustering operations on the enterprise data to generate enterprise tags, and / or set the enterprise tags according to user requirements; Scoring the enterprise label corresponding to each enterprise using a label scoring model to generate a scoring result, wherein the label scoring model is a sequence-to-sequence model based on a long short-term memory network, wherein the long short-term memory network can obtain feature information of the context in the enterprise data; Based on the scoring results, obtain the weight of each of the enterprise tags; Based on the enterprise tags and corresponding weights, construct a first enterprise portrait; Obtaining the impact factor of each enterprise label, and scoring the impact factor using an impact factor scoring model to obtain the value of the impact factor; Based on the impact factor and corresponding value of each enterprise tag, a second enterprise portrait is constructed; Combining the first enterprise portrait with the second enterprise portrait to generate a third enterprise portrait; Wherein, obtaining the weight of each of the enterprise tags based on the scoring result includes: Compare the scoring results of each of the enterprise tags to obtain a comparison result; According to the comparison results, construct a priority diagram weight calculation table; Based on the priority map weight calculation table, the weight of each enterprise tag is obtained, specifically including: obtaining the sum of the comparison results of each column in the priority map weight calculation table as the importance of the corresponding enterprise tag; calculating the proportion of the importance of each enterprise tag to the sum of the comparison results in the priority map weight calculation table to obtain the weight of each enterprise tag; The step of using the label scoring model to score the enterprise label corresponding to each enterprise and generating the scoring result includes: Constructing the label scoring model specifically includes: Preprocessing the enterprise data corresponding to the enterprise tag to generate input data in a preset format; Inputting the input data into a neural network for supervised training to generate the label scoring model; The preprocessing of the enterprise data corresponding to the enterprise tag to generate input data in a preset format includes: Cleaning and counting the enterprise data corresponding to the enterprise tag; According to the statistical results, the scoring value corresponding to the enterprise label is obtained; Based on the enterprise label, the score value, and the corresponding relationship between the enterprise label and the score value, generating input data of the label scoring model; Before the step of obtaining the impact factor of each enterprise label and scoring the impact factor using an impact factor scoring model to obtain the value of the impact factor, the step includes: Constructing the impact factor scoring model specifically includes: Clean and count the data corresponding to the impact factors; Based on the statistical results, a score value corresponding to the impact factor is obtained; Based on the impact factor, the score value, and the corresponding relationship between the impact factor and the score value, the impact factor is input into a neural network for supervised training to generate the impact factor scoring model.

2. The enterprise portrait calculation method based on big data and multi-dimensional features as claimed in claim 1, characterized in that: The comparing the scoring results of each of the enterprise tags to obtain a comparison result includes: Compare the scoring results of each enterprise tag in pairs; If, of the two enterprise tags, the scoring result of the first enterprise tag is greater than the scoring result of the second enterprise tag, the first comparison result obtained is the first set value; If, of the two enterprise tags, the scoring result of the first enterprise tag is less than the scoring result of the second enterprise tag, the second comparison result obtained is the second set value; If the scoring result of the first enterprise tag is equal to the scoring result of the second enterprise tag in the two enterprise tags, the third comparison result is a third set value; wherein the first set value, the second set value and the third set value are all different.

3. The enterprise portrait calculation method based on big data and multi-dimensional features as claimed in claim 1, characterized in that: The enterprise labels include at least: innovation capability, risk resistance capability, and management capability.

4. A computer storage medium, characterized in that: The computer storage medium stores a program for a method for calculating a corporate portrait based on big data and multi-dimensional features. When the program for calculating a corporate portrait based on big data and multi-dimensional features is executed by a processor, the steps of the method for calculating a corporate portrait based on big data and multi-dimensional features described in any one of claims 1 to 3 are implemented.

5. A computer program product, characterized in that It includes a computer program, which, when executed by a processor, implements the steps of the enterprise portrait calculation method based on big data and multi-dimensional features as described in any one of claims 1-3.

Images