Multi-field information data fusion method and system and medium

Abstract

The invention relates to the technical field of data processing, in particular to a multi-field information data fusion method and system and a medium, and the method comprises the steps: collecting first multi-dimensional project data of a target project in each project field; respectively determining a standardized project attenuation coefficient and a target development stage of the target project in each project field based on the first multi-dimensional project data; correcting the standardized project attenuation coefficient based on project innovation uniqueness of the target project in each target development stage to obtain a data importance attenuation coefficient of each project field; and fusing the first multi-dimensional project data of the plurality of project fields according to the data importance attenuation coefficient. The technical problem of insufficient accuracy of project value evaluation can be solved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to a multi-domain information data fusion method, system and medium. Background Art

[0002] In the field of asset valuation, project value assessment and risk control have long relied on single-dimensional analysis such as financial data, market trends, and industry experience, facing pain points such as data isolation, decision-making lag, and information asymmetry. With the development of information technology, Internet multi-modal data analysis technologies (such as macroeconomic indicators, market changes, technological progress, etc.) have shown exponential growth. However, traditional project valuation models are difficult to effectively perform multi-domain data fusion analysis for the current complex business modalities, resulting in insufficient prediction of the potential value and risks of projects.

[0003] Existing data fusion methods mostly perform technical fusion analysis based on the data structure of the data itself. For example, the Bayesian estimation method fuses uncertain data by establishing a probability model of the data, thereby performing data fusion. However, in the field of project valuation, there are many types of data, and there are many pieces of information in different domains that interact with each other. The importance of data in different domains changes significantly at different stages of the project. When existing technologies perform data fusion in the investment field, they often face problems such as real-time performance and insufficient interpretability, resulting in insufficient accuracy of project value assessment. Summary of the Invention

[0004] In order to solve the technical problem of insufficient accuracy of project value assessment, the purpose of the present invention is to provide a multi-domain information data fusion method, system and medium, and the specific technical solutions adopted are as follows: In the first aspect, the present invention provides a multi-domain information data fusion method, and the method includes: Collect first multi-dimensional project data of a target project in each project domain; Based on the first multi-dimensional project data, respectively determine the standardized project attenuation coefficient and the target development stage of the target project in each project domain; Based on the project innovation uniqueness of the target project in each target development stage, correct the standardized project attenuation coefficient to obtain the data importance attenuation coefficient of each project domain; According to the data importance attenuation coefficient, fuse the first multi-dimensional project data of multiple project domains.

[0005] Optionally, the step of respectively determining the standardized project attenuation coefficient and the target development stage of the target project in each project domain based on the first multi-dimensional project data includes: Determine the set of projects of the same type for the target project under each of the project fields based on the first multi-dimensional project data; Calculate the standardized project decay coefficient of the target project under each of the project fields according to the first multi-dimensional project data and the second multi-dimensional project data of at least one project of the same type in the set of projects of the same type; Calculate the industry stage matching degree of the target project under each field dimension of the same project field based on the first multi-dimensional project data and the industry division criteria under multiple field dimensions corresponding to each project field; Determine the target development stage of the target project under each of the project fields by statistically analyzing the industry stage matching degrees of the target project under each field dimension of the same project field.

[0006] Optionally, the determining the set of projects of the same type for the target project under each of the project fields based on the first multi-dimensional project data includes: Based on the first multi-dimensional project data, extract the first project keyword vector of the target project under each of the project fields, and extract the second project keyword vector in the second multi-dimensional project data corresponding to other projects; Based on the first project keyword vector and the second project keyword vector, calculate the project type similarity between the target project under each of the project fields and any other project; Determine the projects of the same type corresponding to the target project as the other projects for which the project type similarity is greater than the first preset threshold, and use the projects of the same type to construct the set of projects of the same type for the target project under each of the project fields.

[0007] Optionally, the calculating the standardized project decay coefficient of the target project under each of the project fields according to the first multi-dimensional project data and the second multi-dimensional project data of at least one project of the same type in the set of projects of the same type includes: Use the second multi-dimensional project data of the at least one project of the same type to calculate the mean value of the second dimension performance data of the at least one project of the same type under each field dimension of each of the project fields; Based on the first dimension performance data of the target project under each field dimension in the first multi-dimensional project data and the mean value of the second dimension performance data, calculate the standardized project decay coefficient of the target project under each of the project fields.

[0008] Optionally, the determining the target development stage of the target project under each of the project fields by statistically analyzing the industry stage matching degrees of the target project under each field dimension of the same project field includes: Determine the preset development stage in which the matching degree of the target project corresponding to each field dimension in the same project field to the industry stage is greater than the second preset threshold; Statistically calculate the proportion of successful matches of each of the preset development stages in multiple field dimensions of each project field, and determine the preset development stage corresponding to the largest proportion of successful matches as the target development stage of the target project in the corresponding project field.

[0009] Optionally, correct the standardized project attenuation coefficient based on the project innovation uniqueness of the target project in each target development stage to obtain the data importance attenuation coefficient of each project field, including: Calculate the project innovation uniqueness of the target project in each target development stage according to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the same target development stage; Determine the product of the reciprocal of the project innovation uniqueness and the standardized project attenuation coefficient in each project field as the data importance attenuation coefficient corresponding to the project field.

[0010] Optionally, calculate the project innovation uniqueness of the target project in each target development stage according to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the same target development stage, including: Determine at least one representative data dimension corresponding to each project field in the target development stage; Construct the first time series data of the target project in the representative data dimension and the second time series data of the same-type project in the representative data dimension according to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the same target development stage; Calculate the difference value between the first time series data and the second time series data; Determine the slope value of the target project in the representative data dimension by fitting the first time series data; Calculate the project innovation uniqueness of the target project in each target development stage based on the difference value and the slope value corresponding to the at least one representative data dimension.

[0011] Optionally, fuse the first multi-dimensional project data of multiple project fields according to the data importance attenuation coefficient, including: Input the data importance attenuation coefficient into the pre-trained neural network data fusion model to obtain the data weight of each project field; Weightedly fuse the first multi-dimensional project data of the multiple project fields based on the data weights.

[0012] In a second aspect, an embodiment of the present invention further provides a multi-field information data fusion system, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, the steps of the multi-field information data fusion method in the first aspect are implemented.

[0013] In a third aspect, an embodiment of the present invention further provides a computer-readable storage medium for storing a computer program, and the computer program causes a computer to execute the multi-field information data fusion method in the first aspect.

[0014] The present invention has the following beneficial effects: Through the technical solution provided by the present invention, after collecting the first multi-dimensional project data of the target project in each project field, the standardized project attenuation coefficient and the target development stage of the target project in each project field can be determined respectively based on the first multi-dimensional project data; further, the standardized project attenuation coefficient is corrected based on the project innovation uniqueness of the target project in each target development stage to obtain the data importance attenuation coefficient of each project field; finally, the first multi-dimensional project data of multiple project fields are fused according to the data importance attenuation coefficient. By determining the standardized project attenuation coefficient and the target development stage, the present invention can measure the importance change of data in each field according to the characteristics of the project in different stages. This overcomes the problems of the prior art that only fuses based on the data structure, does not fully consider the diverse data types in the project valuation field, the mutual influence of information in different fields, and the change of data importance with the project stage, and can better capture the core value driving factors of the project. Further, by accurately measuring the data importance and reasonably fusing multi-field data, the accuracy of project value evaluation can be significantly improved, making the evaluation process more interpretable. Investors can clearly understand the contribution of each data factor in project value evaluation, and thus make more reliable decisions.

[0015] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present invention. Other features and advantages of the present invention will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 Schematic flowchart of a multi - domain information data fusion method provided by an embodiment of the present invention; Figure 2 Schematic flowchart of a multi - domain information data fusion method provided by another embodiment of the present invention. Detailed implementation manners

[0018] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details a multi - domain information data fusion method, system and medium proposed according to the present invention, including its specific implementation manners, structures, features and effects. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0020] The following specifically describes the specific solutions of a multi - domain information data fusion method, system and medium provided by the present invention with reference to the accompanying drawings.

[0021] Please refer to Figure 1 , which shows the method flowchart of a multi - domain information data fusion method provided by an embodiment of the present invention. The method includes the following steps: Step 110: Collect the first multi - dimensional project data of the target project in each project domain.

[0022] Among them, the target project is the core object around which the entire data collection and analysis work revolves, that is, a specific project that needs to conduct value assessment, risk analysis, or other research. For example, in an investment scenario, an emerging technology startup project seeking investment, or a traditional enterprise project planning for strategic transformation, etc., can all become the target project; the target project usually involves multiple different professional categories, and these categories are the project fields. Common project fields include the financial field, which involves the project's financial status, profitability level, cost structure, etc.; the market field, which includes market size, market share, competitor analysis, etc.; the technology field, which covers the core technology relied on by the project, technological innovation ability, technology patent situation, etc.; the team field, which focuses on the professional skills, experience, team collaboration ability of the project team members, etc. Each project field reflects some characteristics of the target project from a specific perspective; each project field corresponds to a first multi-dimensional project data, and the first multi-dimensional project data is multi-dimensional data directly obtained from project-related sources, indicating the diversity of data, and reflecting the project characteristics of the target project in this project field from different angles and levels. Taking the market field as an example, there is not only data on the dimension of market size, but also multiple dimensions such as market growth rate, market distribution in different regions, and preferences of different customer groups.

[0023] Step 120: Based on the first multi-dimensional project data, respectively determine the standardized project decay coefficient and the target development stage of the target project in each project field.

[0024] Since the target project involves multiple different project fields, and each field has its own unique characteristics and development laws, it is necessary to conduct separate analysis and calculation for each field to obtain specific indicators applicable to that field, namely the standardized project decay coefficient and the target development stage.

[0025] Among them, the standardized project attenuation coefficient is an indicator used to measure the relative performance of a target project among projects of the same type. When calculating, it will refer to the mean value of the performance data of projects of the same type in each field dimension and combine the performance of the target project itself in these dimensions to determine. Generally speaking, if the performance of the target project in a certain field is better than the average level of projects of the same type, its standardized project attenuation coefficient is relatively small, which means that the attenuation effect of data on project value evaluation in this field is small, and the project value is less affected; on the contrary, if the performance is poor, the coefficient is relatively large, the attenuation effect of data is more obvious, and the negative impact on project value evaluation is greater. For example, in the smartphone market, if the key indicators such as the market share and user satisfaction rate of a certain brand of mobile phone are higher than the average level of products of the same type, its standardized project attenuation coefficient in the market field will be relatively low; the target project will go through different development stages in different project fields. For example, in the product life cycle, it may go through the R & D period, the introduction period, the growth period, the maturity period and the decline period. The target development stage is to determine the specific stage where the target project is currently in each field. The determination of this stage is usually based on the matching situation of the project with the industry division standard in multiple field dimensions. For example, for a technology startup project, in the technology field, it may be in the stage of technological R & D breakthrough, and the product has not been widely launched into the market; in the market field, it may be in the market introduction period, with a small market share but great growth potential; in the financial field, it may be in a loss state, but with a large R & D investment and in the accumulation stage. Clearly defining the target development stage helps to understand the development trend of the project in different fields and provides an important reference for subsequent decision-making.

[0026] Step 130: Based on the project innovation uniqueness of the target project in each target development stage, correct the standardized project attenuation coefficient to obtain the data importance attenuation coefficient for each project field.

[0027] For the embodiments of the present disclosure, the previously determined standardized project attenuation coefficient can be adjusted and optimized according to the unique innovation characteristics demonstrated by the target project in each target development stage. Through such a correction process, the data importance attenuation coefficient exclusive to each project field is obtained, and this coefficient can more accurately reflect the actual importance degree of data in each field during project evaluation.

[0028] Among them, the innovation uniqueness of a project refers to the unique features and advantages of the target project in terms of innovation compared to other projects of the same type. This may be reflected in technological innovation, such as having unique patented technologies and innovative R & D processes; it may also be manifested in business model innovation, such as creating a brand-new profit model and unique marketing strategies. Taking an Internet e-commerce enterprise as an example, its innovation uniqueness may lie in the use of big data algorithms to achieve precise personalized product recommendations, thereby increasing the user purchase conversion rate. This innovation point enables it to stand out in the highly competitive e-commerce market. When calculating the innovation uniqueness of a project, relevant data of projects of the same type at the same development stage are usually compared, and the differences and advantages are analyzed to determine it; the standardized project attenuation coefficient is obtained based on the comparative analysis of the target project and projects of the same type in the early stage, and is used to measure the relative performance of the project among projects of the same type. However, this coefficient does not fully consider the innovation uniqueness of the project itself. The correction process is to make adjustments by combining the innovation uniqueness of the project on the basis of the standardized project attenuation coefficient. If the innovation uniqueness of the project is relatively high, it indicates that the project has greater advantages and potential in market competition. Then, when evaluating the project value, the importance of data in related fields should be enhanced, that is, the standardized project attenuation coefficient needs to be correspondingly reduced; on the contrary, if the innovation uniqueness is relatively low, the standardized project attenuation coefficient may need to be increased. After correcting the standardized project attenuation coefficient by combining the innovation uniqueness of the project, the data importance attenuation coefficient exclusive to each project field is finally obtained. This coefficient comprehensively considers the development stage of the project in this field, the innovation uniqueness, and the comparison with projects of the same type, and can more accurately reflect the importance degree of data in each project field in the project value evaluation.

[0029] Step 140: Integrate the first multi-dimensional project data of multiple project fields according to the data importance attenuation coefficient.

[0030] For the embodiments of the present disclosure, the data importance attenuation coefficients of each project field obtained in the previous steps can be used to integrate and process the first multi-dimensional project data of the target project in multiple different project fields collected initially. During the integration process, different weights are assigned according to the importance degree of data in each field, so that the integrated data can more accurately reflect the true situation and value of the project in different aspects.

[0031] In summary, according to a multi - domain information data fusion method provided by the present invention, after collecting the first multi - dimensional project data of the target project in each project domain; based on the first multi - dimensional project data, respectively determine the standardized project attenuation coefficient and the target development stage of the target project in each project domain; further correct the standardized project attenuation coefficient based on the project innovation uniqueness of the target project in each target development stage to obtain the data importance attenuation coefficient of each project domain; finally, fuse the first multi - dimensional project data of multiple project domains according to the data importance attenuation coefficient. By determining the standardized project attenuation coefficient and the target development stage, the present invention can measure the change of the importance of data in each domain according to the characteristics of the project in different stages. This overcomes the problem of the prior art that only fuses based on the data structure, without fully considering the diverse data types in the project valuation domain, the mutual influence of information in different domains, and the change of data importance with the project stage, and can better capture the core value - driving factors of the project. Further, by accurately measuring the data importance and reasonably fusing multi - domain data, the accuracy of project value evaluation can be significantly improved, making the evaluation process more interpretable. Investors can clearly understand the contribution of each data factor in project value evaluation, and thus make more reliable decisions.

[0032] Based on Figure 1 the embodiments shown, as a refinement and extension of the above - mentioned embodiments, in order to fully illustrate the specific implementation process of the method of this embodiment, this embodiment provides a specific method as shown in Figure 2 shown. Figure 2 Based on Figure 1 the embodiments shown. As shown in 2, the method includes the following steps: Step 210, collect the first multi - dimensional project data of the target project in each project domain.

[0033] In related scenarios such as project evaluation or investment decision - making, in order to conduct a comprehensive and in - depth analysis of the target project, for each different professional domain involved in the target project, collect the most original and basic data that reflects the project characteristics from multiple angles and multiple levels, that is, the first multi - dimensional project data.

[0034] For the embodiments of the present disclosure, the first multi - dimensional project data of the target project in each project domain can be collected through techniques such as questionnaire surveys, database queries, web crawlers, sensor monitoring, on - site research, etc., to provide strong support for subsequent accurate analysis and decision - making.

[0035] Step 220, based on the first multi - dimensional project data, determine the set of same - type projects of the target project in each project domain.

[0036] For the embodiments of the present disclosure, step 220 of determining the set of same - type projects of the target project in each project domain based on the first multi - dimensional project data may include the following steps: Step 220-1: Based on the first multi-dimensional project data, extract the first project keyword vectors of the target project in each project field, and extract the second project keyword vectors in the corresponding second multi-dimensional project data of other projects.

[0037] Among them, the first project keyword vector is the result obtained by extracting key information and converting it into vector form for the target project in each project field; other projects refer to other relevant projects participating in the comparative analysis except the target project. These projects are in the same industry as the target project or have similarities. By comparing with them, the advantages and disadvantages, market positioning, etc. of the target project can be better understood; the second multi-dimensional project data is similar to the first multi-dimensional project data and is the original data that reflects the characteristics of other projects from multiple perspectives and levels; the second project keyword vector is extracted from the second multi-dimensional project data of other projects and is a vector used to represent the key characteristics of these projects in each project field. It has a similar function to the first project keyword vector, but is for other projects.

[0038] For the embodiments of the present disclosure, when extracting the first project keyword vector and the second project keyword vector, natural language processing (NLP) techniques can be used to extract the words or phrases that best represent the document theme and key information from the text content and convert them into vector form. Natural language processing (NLP) techniques may include, for example, the term frequency-inverse document frequency (TF-IDF) algorithm, latent semantic analysis (LSA), etc.

[0039] Step 220-2: Based on the first project keyword vector and the second project keyword vector, calculate the project type similarity between the target project in each project field and any other project.

[0040] Among them, the project type similarity is a quantitative index calculated by comparing the keyword vectors of the target project and other projects in each project field, and is used to measure the similarity degree between two projects in terms of business model, technology application, market positioning, etc. A high project type similarity means that the two projects may have more common points in terms of business model, technology application, market positioning, etc.; otherwise, the differences are relatively large. The first project keyword vector is extracted from the first multi-dimensional project data of the target project and is a vector representing the key characteristics of the target project in each project field; the second project keyword vector is extracted from the second multi-dimensional project data of other projects and is used to reflect the key characteristics of other projects in each field.

[0041] For the embodiments of the present disclosure, a specific algorithm can be used to process the first project keyword vector and the second project keyword vector to obtain a quantitative value representing their similarity. Common methods for calculating vector similarity include the cosine similarity algorithm, the Euclidean distance algorithm, etc. Taking the cosine similarity algorithm as an example, this algorithm measures similarity by calculating the cosine value of the included angle between two vectors, and the value range is between -1 and 1. The closer the cosine value is to 1, the more similar the directions of the two vectors are, that is, the higher the project type similarity; the closer it is to 0, the greater the difference between the two vectors, and the lower the project type similarity.

[0042] Step 220-3: Determine other projects with project type similarity greater than the first preset threshold as the same type projects corresponding to the target project, and use the same type projects to construct a set of the same type projects of the target project in each project field.

[0043] Among them, the first preset threshold is a numerical standard set manually and is used to determine whether other projects are of the same type as the target project. The setting of this threshold needs to be combined with specific business scenarios and analysis purposes, and is generally determined based on experience or the analysis of a large amount of historical data; the same type projects refer to other projects that have high similarity with the target project in a certain project field and have many common characteristics in terms of business, technology, market, etc. These projects can be used as a reference for the target project to compare and analyze the advantages and disadvantages of the target project and its position in the industry, etc.; the set of the same type projects refers to a set formed by concentrating the same type projects that meet the similarity standard (i.e., similarity greater than the first preset threshold) for each project field of the target project.

[0044] In a specific application scenario, when determining the same-type projects of a target project, the similarity scores of the target project and all other projects in each project field can be traversed. For each project field, select other projects with similarity scores greater than the first preset threshold. Suppose the target project is an enterprise adopting the social e-commerce model. In the market field, through calculation, it is found that the similarity scores of Project A, Project B, and Project C with the target project are 0.75, 0.8, and 0.72 respectively, all of which are greater than the preset threshold of 0.7. Then Project A, Project B, and Project C are determined as the same-type projects of the target project in the market field. For each project field of the target project, summarize the selected same-type projects to form a corresponding set of same-type projects. The above-mentioned Project A, Project B, and Project C are determined as the same-type projects of the target project in the market field, then these three projects can be grouped into a set, and this set is the set of same-type projects of the target project in the market field; similarly, taking the technology field as an example, suppose after screening, Project D, Project E, and Project F are determined as the same-type projects of the target project in the technology field, then these three projects are grouped into a set, and this set is the set of same-type projects of the target project in the technology field.

[0045] Step 230: Calculate the standardized project decay coefficient of the target project in each project field according to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the set of same-type projects.

[0046] For the embodiments of the present disclosure, calculating the standardized project decay coefficient of the target project in each project field according to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the set of same-type projects in Step 230 may include the following steps: Step 230-1: Use the second multi-dimensional project data of at least one same-type project to calculate the mean value of the second-dimension performance data of at least one same-type project in each field dimension of each project field.

[0047] Among them, the second-dimension performance data is the specific performance data of the same-type project in each field dimension. Taking the "market share" dimension in the market field as an example, the second-dimension performance data is the actual market share value of each same-type project; in the "R & D investment" dimension in the technology field, the second-dimension performance data is the amount of funds invested in R & D by each same-type project.

[0048] For the embodiments of the present disclosure, after determining the same-type projects of the target project, all the same-type projects in the set of same-type projects can be selected. For the second multi-dimensional project data owned by these same-type projects, according to each project field involved in the target project and each field dimension included therein, the average value of the second dimension performance data of all the same-type projects in the set of same-type projects under each field dimension is calculated respectively. This average value will be used as a reference standard for comparative analysis of the performance of the target project in the corresponding dimension.

[0049] Step 230-2: Calculate the standardized project decay coefficient of the target project under each project field based on the first dimension performance data of the target project under each field dimension in the first multi-dimensional project data and the average value of the second dimension performance data.

[0050] Among them, the first dimension performance data refers to the original data of the actual performance of the target project under each field dimension. Taking the "market share" dimension in the market field as an example, if the market share of the target project is 15%, this 15% is the first dimension performance data under this dimension; in the "R & D investment" dimension in the technology field, if the R & D investment of the target project is 5 million yuan, 5 million yuan is the first dimension performance data under this dimension; the standardized project decay coefficient is used to measure the relative performance of the target project among the same-type projects, and is determined by comparing the data of the target project itself with the average value of the same-type projects, providing a basis for subsequent evaluation of the project value and importance.

[0051] For the embodiments of the present disclosure, when calculating the standardized project decay coefficient of the target project under each project field, the first dimension performance data of each field dimension of the target project under each project field and the average value of the second dimension performance data of all the same-type projects in the set of same-type projects under each field dimension can be substituted into the first calculation formula, and then the standardized project decay coefficient of the target project under the corresponding project field can be calculated. Among them, the formula feature description of the first calculation formula is: In the formula, represents the standardized project decay coefficient of the target project under the mth project field; The function is a mathematical function used to uniformly convert project indicators of different types and magnitudes into a relatively comparable numerical range; represents the average value of the second dimension performance data of all the same-type projects in the set of same-type projects corresponding to the mth project field under the xth field dimension; represents the first dimension performance data of the target project under the xth field dimension of the mth project field; X represents that there are X field dimensions in the mth project field, such as the proportion of the market occupancy of the project product, sales data, market growth situation, etc.

[0052] Step 240: Calculate the industry stage matching degree of the target project in each domain dimension of the same project domain based on the first multi-dimensional project data and the industry division criteria under multiple domain dimensions corresponding to each project domain.

[0053] Among them, the industry division criteria are formulated based on industry experience, market research, and relevant industry norms, and are used to divide the different development stages of the industry. These criteria set different-stage quantitative indicators or feature descriptions for each domain dimension. For example, in the market domain, the introduction period, growth period, maturity period, and decline period are divided according to market share and market growth rate; in the technology domain, the technology development stage is divided based on R & D investment intensity, technological leadership, etc.; the industry stage matching degree is a value obtained by comparing the data of the target project in each domain dimension with the industry division criteria, which measures the degree of fit between the target project and different industry stages in this domain dimension. The higher the matching degree, the more the performance of the target project in this dimension conforms to the characteristics of the corresponding industry stage.

[0054] For the embodiments of the present disclosure, when calculating the industry stage matching degree, the data of the target project in each domain dimension can be compared with the industry division criteria to determine its matching degree with different industry stages. Taking market share and user growth rate as an example, the market share of this enterprise is 12% and the user growth rate is 20%. Comparing with the criteria, it is more in line with the characteristics of the growth period, and the matching degree can be set to 0.8 (with a full score of 1 point, scored according to the degree of compliance). The course update frequency is 3 times a month, which is in the medium frequency range and corresponds to the growth stage, and the matching degree is set to 0.7. Then, the industry stage matching degrees of the target project in each domain dimension of the same project domain are summarized.

[0055] Step 250: Determine the target development stage of the target project in each project domain by statistically analyzing the industry stage matching degrees of the target project in each domain dimension of the same project domain.

[0056] For the embodiments of the present disclosure, step 250 of determining the target development stage of the target project in each project domain by statistically analyzing the industry stage matching degrees of the target project in each domain dimension of the same project domain may include the following steps: Step 250-1: Determine the preset development stages corresponding to the industry stage matching degrees of the target project in each domain dimension of the same project domain that are greater than the second preset threshold.

[0057] Among them, the second preset threshold is a numerical standard set artificially and is used to screen the preset development stages with relatively high industry stage matching degrees. The setting of this threshold needs to combine industry characteristics and the actual situation of the project, and is generally determined based on experience or the analysis of a large amount of historical data; the preset development stages are different industry development stages preset according to industry classification standards, such as the introduction period, growth period, maturity period, and decline period in the market field; the start-up period, early growth period, late growth period, maturity period, etc. in the technology field.

[0058] For the embodiments of the present disclosure, on the basis of having calculated the industry stage matching degrees of the target project in each domain dimension of the same project domain, a second preset threshold is set to screen out the preset development stages with matching degrees greater than this threshold. This step is to initially screen out the industry development stages that are more compatible with the performance of the target project in each dimension, exclude the obviously incompatible stages, and narrow the scope of subsequent analysis.

[0059] Step 250-2: Statistically calculate the proportion of successful matches of each preset development stage in multiple domain dimensions of each project domain, and determine the preset development stage with the largest corresponding proportion of successful matches as the target development stage of the target project in the corresponding project domain.

[0060] Among them, the proportion of successful matches is the ratio of the number of dimensions with matching degrees greater than the second preset threshold to the total number of dimensions of a certain preset development stage in multiple domain dimensions of the project domain, and is used to quantify the overall degree of fit between this stage and the performance of the project in each dimension; the target development stage is the development stage of the target project in the corresponding project domain finally determined by comprehensively considering various factors, and can intuitively reflect the development status and trend of the project in this domain.

[0061] For the embodiments of the present disclosure, for the screened preset development stages, the proportion of successful matches of them in multiple domain dimensions of each project domain can be statistically calculated. The proportion of successful matches represents the proportion of the number of dimensions with matching degrees greater than the second preset threshold to the total number of dimensions in all participating domain dimensions of a certain preset development stage. By comparing these proportions, the preset development stage with the largest proportion is determined as the target development stage of the target project in the corresponding project domain, so as to comprehensively judge the actual development stage of the project in this domain.

[0062] For example, the industry stage matching degrees of the above online education enterprise in the three domain dimensions of the market domain: market share, user growth rate, and course update frequency are 0.8, 0.8, and 0.7 respectively, all of which are greater than the second preset threshold of 0.7. And the preset development stages corresponding to the market share, user growth rate, and course update frequency are the growth stage, the maturity stage, and the growth stage respectively. Then, the proportion of successful matches in the 3 domain dimensions of the market domain for each preset development stage can be counted. For example, in the three domain dimensions, the growth stage appears 2 times, and the number of appearances in other stages is less than 2 times. Finally, according to the industry stage with the largest proportion of successful matches or the highest frequency of appearance, the target development stage of the target project in this project domain can be determined. The comprehensive matching degree of the above online education enterprise in the market domain shows that it is closer to the growth stage, and the proportion of successful matches corresponding to the growth stage is the largest and the frequency of appearance is the highest. Therefore, the target development stage of this online education enterprise in the market domain is determined to be the growth stage.

[0063] Step 260: Based on the project innovation uniqueness of the target project in each target development stage, correct the standardized project decay coefficient to obtain the data importance decay coefficient for each project domain.

[0064] For the embodiments of the present disclosure, the step of correcting the standardized project decay coefficient based on the project innovation uniqueness of the target project in each target development stage in step 260 to obtain the data importance decay coefficient for each project domain may include the following steps: Step 260-1: Calculate the project innovation uniqueness of the target project in each target development stage according to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the same target development stage.

[0065] For the embodiments of the present disclosure, the embodiment steps may include: determining at least one representative data dimension corresponding to the target development stage for each project domain; constructing the first time series data of the target project in the representative data dimension and the second time series data of the same-type project in the representative data dimension according to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the same target development stage; calculating the difference value between the first time series data and the second time series data; determining the slope value of the target project in the representative data dimension by fitting the first time series data; calculating the project innovation uniqueness of the target project in each target development stage based on the difference values and slope values corresponding to at least one representative data dimension.

[0066] Among them, the representative data dimension is the data dimension selected under the corresponding target development stage in each project field, which can prominently reflect the innovation characteristics of the project or have a key impact on the project development. For example, the number of patents in the technology field, the growth rate of R & D investment, the diversity of new customer acquisition channels in the market field, etc.; the first time series data is a data series formed by arranging the first multi-dimensional project data of the target project in chronological order under the representative data dimension, which is used to show the change of the target project over time in this dimension; the second time series data is a data series formed by arranging the same type of projects in chronological order under the representative data dimension, corresponding to the first time series data, which is used for comparative analysis with the target project; the difference value is a numerical value obtained by calculating the degree of difference between the first time series data and the second time series data through a specific algorithm, which reflects the difference between the target project and the same type of projects in this dimension; the slope value is the slope of the straight line obtained by fitting the first time series data, which is used to measure the change trend of the data over time. Specifically, a positive slope indicates that the data is on the rise, and a negative slope indicates that the data is on the decline. Further, the positive and negative of the slope k can be adjusted according to the correlation between the data, which can comprehensively consider the relationship between multi-dimensional data and market performance, and more accurately measure the impact of the development trend of the project in each dimension on the innovation uniqueness and data importance of the project, providing more reliable data support for calculating the attenuation coefficients of the project innovation uniqueness and data importance; the project innovation uniqueness is an index used to measure the unique degree of the target project in innovation compared with the same type of projects. The higher the value, the more unique the target project is in innovation and may have greater advantages in market competition.

[0067] For the embodiments of the present disclosure, when adjusting the positive and negative of the slope k according to the correlation between the data, a three-dimensional space model can be first constructed with the data of different field dimensions of the same type of projects as the X-axis, the market performance as the Y-axis, and multiple data within the field as the Z-axis; then the Pearson correlation coefficient is used to calculate the correlation between each field dimension data in the X-axis and the market performance in the Y-axis, and the field dimension data with the largest positive correlation and an upward trend is selected as the key data dk. For the c-th field dimension data of the target project, calculate its Pearson correlation coefficient with the key data dk . If >0, it means that the c-th field dimension data is positively correlated with the key data (the larger the data, the more beneficial to the project); if <0, it means that the c-th field dimension data is negatively correlated with the key data (the smaller the data, the more beneficial to the project). Finally, the correlation coefficient can be used to adjust the slope k of the c-th field dimension: ; in the formula, is the final slope value of the target project under the representative data dimension c; k is the slope of the straight line obtained by fitting the first time series data of the target project under the representative data dimension c; is the Pearson correlation coefficient of the target project under the representative data dimension c.

[0068] Correspondingly, when calculating the project innovation uniqueness of the target project at each target development stage, the innovation uniqueness contribution value of each representative data dimension can be calculated. For example, the innovation uniqueness contribution value of the R & D investment growth rate dimension c in the m-th project field is , where is the innovation uniqueness contribution value of the R & D investment growth rate dimension c in the m-th project field; is the final slope value of the target project under the representative data dimension c; is the difference value between the first time series data and the second time series data under the R & D investment growth rate dimension c in the m-th project field. If there are multiple representative data dimensions, traverse and calculate the contribution values of each dimension, accumulate and normalize. Suppose the innovation uniqueness contribution value in the dimension of the number of new marketing channels expanded is 2.7×0.3 = 0.81. If there are multiple representative data dimensions, traverse and calculate the contribution values of each dimension, accumulate and normalize to obtain the unique project innovation uniqueness value of the target project at this target development stage. For example, if the contribution value calculated in the dimension of the patent technology iteration speed in the technology field is 0.6, the sum of the two dimensions is 0.81 + 0.6 = 1.41. After normalization, the project innovation uniqueness of this intelligent hardware enterprise in the market growth stage is approximately 0.81÷1.41≈0.57. In this way, it is ensured that each target development stage of the target project corresponds to a project innovation uniqueness value that accurately reflects its innovation uniqueness degree, so as to perform subsequent operations such as correcting the standardized project decay coefficient based on this value.

[0069] Step 260-2: Determine the data importance decay coefficient corresponding to each project field by multiplying the reciprocal of the project innovation uniqueness by the standardized project decay coefficient under each project field.

[0070] Among them, the data importance decay coefficient is obtained by comprehensively considering the project innovation uniqueness and the standardized project decay coefficient, and is an index reflecting the importance degree of the data in each project field in the project value evaluation. The larger the coefficient, the greater the attenuation degree of the importance of the data in this field in the project value evaluation; the smaller the coefficient, the smaller the attenuation degree of the data importance.

[0071] Step 270: Fuse the first multi-dimensional project data of multiple project fields according to the data importance decay coefficient.

[0072] For the embodiments of the present disclosure, fusing the first multi-dimensional project data of multiple project fields in step 270 according to the data importance decay coefficient may include the following steps: Step 270-1: Input the data importance attenuation coefficient into the pre-trained neural network data fusion model to obtain the data weights for each project area.

[0073] Among them, a neural network is a computational model that mimics the structure and function of the biological nervous system and learns patterns and regularities in data through training with a large amount of data. The pre-trained neural network data fusion model refers to a neural network model that has been trained on relevant data and is capable of processing and analyzing the input data to achieve the data fusion function. For example, in project data processing, this model can calculate appropriate data weights based on the data characteristics and importance information of different project areas in the input, so as to achieve effective data fusion; the data weight is a quantitative representation of the importance of the data in each project area in the overall data during the data fusion process. Through the processing of the data importance attenuation coefficient by the neural network data fusion model, the specific data weights for each project area are obtained. For example, the data weight of the market area is 0.4, and the data weight of the technology area is 0.3, etc. The larger the weight, the greater the proportion of the data in this area in the result after data fusion.

[0074] Step 270-2: Perform weighted fusion on the first multi-dimensional project data in multiple project areas based on the data weights.

[0075] In summary, the technical solution in this application can measure the change in the importance of data in each area according to the characteristics of the project in different stages by determining the standardized project attenuation coefficient and the target development stage. This overcomes the problems of the prior art that only focuses on data structure fusion and does not fully consider the diverse data types in the project valuation field, the mutual influence of information in different fields, and the change in data importance with the project stage, and can better capture the core value driving factors of the project. Further, by accurately measuring the data importance and reasonably fusing multi-field data, the accuracy of project value assessment can be significantly improved, making the assessment process more interpretable. Investors can clearly understand the contributions of various data factors in project value assessment, and thus make more reliable decisions.

[0076] Based on the same inventive concept as the above method, an embodiment of the present invention further provides a multi-field information data fusion system, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of any one of the multi-field information data fusion methods in the first aspect above.

[0077] Based on the same inventive concept as the above method, an embodiment of the present invention further provides a computer-readable storage medium for storing a computer program, and the computer program enables a computer to execute the steps of any one of the multi-field information data fusion methods in the first aspect above.

[0078] In summary, the embodiments of the present invention provide a multi-domain information data fusion method, system, and medium. By determining the standardized project attenuation coefficient and the target development stage, it is possible to measure the change in the importance of data in each domain based on the characteristics of the project at different stages. This overcomes the problem in the prior art that only data structure fusion is considered, without fully taking into account the diverse data types in the project valuation domain, the mutual influence of information in different domains, and the change in data importance with the project stage, and is more capable of capturing the core value driving factors of the project. Further, by accurately measuring the data importance and reasonably fusing multi-domain data, the accuracy of project value assessment can be significantly improved, making the assessment process more interpretable. Investors can clearly understand the contribution of each data factor in project value assessment, and thus make more reliable decisions.

[0079] It should be noted that the above sequence of the embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. And the above description of specific embodiments of this specification is provided. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired result. In some embodiments, multi-task processing and parallel processing are also possible or may be advantageous.

[0080] Each embodiment in this specification is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and the key point of each embodiment is to illustrate the differences from other embodiments.

[0081] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A multi-domain information data fusion method, characterized in that, The method includes: Collecting first multi-dimensional project data of a target project in each project field; Based on the first multi-dimensional project data, respectively determining a standardized project attenuation coefficient and a target development stage of the target project under each of the project fields; Based on the project innovation uniqueness of the target project at each of the target development stages, correcting the standardized project attenuation coefficient to obtain a data importance attenuation coefficient for each of the project fields; Fusing the first multi-dimensional project data of multiple project fields according to the data importance attenuation coefficient.

2. The multi-domain information data fusion method according to claim 1, wherein The respectively determining a standardized project attenuation coefficient and a target development stage of the target project under each of the project fields based on the first multi-dimensional project data includes: Based on the first multi-dimensional project data, determining a set of same-type projects of the target project under each of the project fields; According to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the set of same-type projects, calculating the standardized project attenuation coefficient of the target project under each of the project fields; Based on the first multi-dimensional project data and the industry division criteria under multiple field dimensions corresponding to each of the project fields, calculating the industry stage matching degree of the target project under each field dimension of the same project field; By statistically counting the industry stage matching degrees of the target project under each field dimension of the same project field, determining the target development stage of the target project under each of the project fields.

3. The multi-domain information data fusion method according to claim 2, wherein The determining a set of same-type projects of the target project under each of the project fields based on the first multi-dimensional project data includes: Based on the first multi-dimensional project data, extracting a first project keyword vector of the target project under each of the project fields, and extracting a second project keyword vector in the second multi-dimensional project data corresponding to other projects; Based on the first project keyword vector and the second project keyword vector, calculating the project type similarity between the target project under each of the project fields and any other project; Determining other projects with the project type similarity greater than a first preset threshold as the same-type projects corresponding to the target project, and using the same-type projects to construct a set of same-type projects of the target project under each of the project fields.

4. The multi-domain information data fusion method according to claim 2, wherein The calculating the standardized project attenuation coefficient of the target project under each of the project fields according to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the set of same-type projects includes: Using the second multi-dimensional project data of the at least one same-type project to calculate the mean value of the second dimension performance data of the at least one same-type project under each field dimension of each of the project fields; Based on the first dimension performance data of the target project under each field dimension in the first multi-dimensional project data and the mean value of the second dimension performance data, calculating the standardized project attenuation coefficient of the target project under each of the project fields.

5. The multi-domain information data fusion method according to claim 2, wherein Determining the target development stage of the target project in each of the project fields by statistically calculating the industry stage matching degree of the target project under each field dimension in the same project field, including: Determining the preset development stages corresponding to the industry stage matching degree of the target project greater than the second preset threshold under each field dimension in the same project field; Statistically calculating the matching success ratio of each of the preset development stages in multiple field dimensions of each of the project fields, and determining the preset development stage corresponding to the largest matching success ratio as the target development stage of the target project in the corresponding project field.

6. The multi-domain information data fusion method according to claim 2, wherein Based on the project innovation uniqueness of the target project in each of the target development stages, correcting the standardized project attenuation coefficient to obtain the data importance attenuation coefficient of each of the project fields, including: Calculating the project innovation uniqueness of the target project in each of the target development stages according to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the same target development stage; Determining the product of the reciprocal of the project innovation uniqueness and the standardized project attenuation coefficient of each of the project fields as the data importance attenuation coefficient corresponding to the project field.

7. The multi-domain information data fusion method according to claim 6, wherein Calculating the project innovation uniqueness of the target project in each of the target development stages according to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the same target development stage, including: Determining at least one representative data dimension corresponding to the target development stage of each of the project fields; Constructing the first time series data of the target project under the representative data dimension and the second time series data of the same-type project under the representative data dimension according to the first multi-dimensional project data and the second multi-dimensional project data of at least one same-type project in the same target development stage; Calculating the difference value between the first time series data and the second time series data; Determining the slope value of the target project under the representative data dimension by fitting the first time series data; Calculating the project innovation uniqueness of the target project in each of the target development stages based on the difference value and the slope value corresponding to the at least one representative data dimension.

8. The multi-domain information data fusion method according to claim 1, characterized in that Fusing the first multi-dimensional project data of multiple project fields according to the data importance attenuation coefficient, including: Inputting the data importance attenuation coefficient into the pre-trained neural network data fusion model to obtain the data weight of each of the project fields; Performing weighted fusion on the first multi-dimensional project data of multiple project fields based on the data weight.

9. A multi-domain information data fusion system, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the computer program is executed by a processor, it implements the steps of a multi-field information data fusion method according to any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, For storing a computer program, the computer program causes a computer to execute the multi-field information data fusion method according to any one of claims 1-8.

Images