Search

CN-122020588-A - Multi-dimensional analysis-based scientific creation big data model generation method and system

CN122020588ACN 122020588 ACN122020588 ACN 122020588ACN-122020588-A

Abstract

The invention discloses a method and a system for generating a scientific and invasive big data model based on multidimensional analysis, and particularly relates to the field of generating the scientific and invasive big data model. The data acquisition step comprises a model input data acquisition unit and a model feedback data acquisition unit, and is used for acquiring target data in real time and transmitting the acquired data to the data analysis step. The model input data acquisition unit is used for acquiring technical basic data, resource association data and association characteristic data, the feedback data acquisition unit is used for acquiring innovation theory result data, model stability data and model adaptability data, and scientificity and reliability of model generation data sources are effectively improved by dividing large model-based scientific creation large data model generation data into various batches.

Inventors

  • CHEN WEI

Assignees

  • 湖北科惠通科技有限公司

Dates

Publication Date
20260512
Application Date
20260128

Claims (10)

  1. 1. A method for generating a scientific big data model based on multidimensional analysis is characterized by comprising the following steps: S1, data batch division, namely determining data to be acquired as target data, dividing the target data into different batches according to an equal time division mode, and marking the different batches as 1, 2. And S2, a data acquisition step, which comprises a model input data acquisition unit and a model feedback data acquisition unit, and is used for acquiring target data in real time and transmitting the acquired data to a data analysis step. The model input data acquisition unit is used for acquiring technical basic data, resource association data and association characteristic data; the feedback data acquisition unit is used for acquiring innovation theory result data, model stability data and model adaptability data; And S3, a data analysis step, which comprises a model input data analysis unit and a model feedback data analysis unit, and is used for collecting target data in real time and transmitting the collected data to a data comprehensive evaluation step. The feedback data analysis unit comprises an innovation theory result data analysis node, a model stability data analysis node and a model adaptability data analysis node; S4, a data comprehensive evaluation step, which comprises a data analysis unit for generating a scientific large data model, wherein the data analysis unit is used for comprehensively analyzing the analysis result transmitted in the data analysis step and transmitting the analysis result to an early warning feedback step; And S5, early warning feedback, namely establishing a science-created big data model to generate a comprehensive evaluation index preset value, judging the comprehensive evaluation index value generated by the science-created big data model according to the comprehensive evaluation index preset value generated by the science-created big data model, and sending a corresponding signal according to a judging result.
  2. 2. The method for generating the scientific and large data model based on the multidimensional analysis according to claim 1, wherein the technical basic data comprise algorithm complexity Ac, data dimension Ds and data sparsity Sd, the resource association data comprise energy consumption rate Ec and data transmission bandwidth demand Dt, the association feature data comprise feature correlation coefficient Fc and feature entropy value Er, the innovation theory result data comprise theory logic association strength St, theory novelty Nt and theory expansion dimension Te, the model stability data comprise model convergence speed Ms and variance expansion factor Vf, and the model adaptability data comprise new data adaptation rate Ar and data scale expansibility Dc.
  3. 3. The method for generating a large scientific data model based on multidimensional analysis as claimed in claim 1, wherein the technical basic data analysis node is used for establishing a technical basic data calculation model, and importing the technical basic data transmitted in the data acquisition step into the technical basic data calculation model to obtain a model input complexity coefficient value, and the technical basic data calculation model is specifically expressed as: , Wherein, the Representing the model input complexity coefficient value of the ith calculation, ac i represents the algorithm complexity of the ith acquisition, ds i represents the data dimension of the ith acquisition, and Sd i represents the data sparsity of the ith acquisition.
  4. 4. The method for generating a large scientific data model based on multidimensional analysis as claimed in claim 1, wherein the resource-associated data analysis node is configured to build a resource-associated data calculation model, import the resource-associated data transmitted in the data acquisition step into the resource-associated data calculation model, and obtain a resource fitness coefficient value, and the resource-associated data calculation model is specifically expressed as: , Wherein, the The resource adaptation degree coefficient value representing the ith calculation, ec i represents the energy consumption rate of the ith acquisition, dt i represents the data transmission bandwidth requirement of the ith acquisition, σ represents the standard deviation of the data transmission bandwidth requirement, and μ represents the average value of the data transmission bandwidth requirement.
  5. 5. The method for generating a large scientific data model based on multidimensional analysis as claimed in claim 1, wherein the associated feature data analysis node is used for establishing an associated feature data calculation model, and importing the associated feature data transmitted in the data acquisition step into the associated feature data calculation model to obtain a feature validity coefficient value, and the associated feature data calculation model is specifically expressed as: , Wherein, the Representing the feature validity coefficient value of the ith calculation, fc i representing the feature correlation coefficient of the ith acquisition, er i representing the feature entropy value of the ith acquisition.
  6. 6. The method for generating a large data model of science creation based on multidimensional analysis of claim 1, wherein the innovation theory result data analysis node is used for establishing an innovation theory result data calculation model, and importing innovation theory result data transmitted in the data acquisition step into the innovation theory result data calculation model to obtain model innovation coefficient values, and the innovation theory result data calculation model is specifically expressed as: , Wherein, the Representing the model innovation coefficient value of the ith calculation, st i represents the theoretical logical association strength of the ith acquisition, nt i represents the theoretical novelty of the ith acquisition, and Te i represents the theoretical expansion dimension of the ith acquisition.
  7. 7. The method for generating a large scientific data model based on multidimensional analysis as claimed in claim 1, wherein the model stability data analysis node is used for establishing a model stability data calculation model, and importing the model stability data transmitted in the data acquisition step into the model stability data calculation model to obtain a model stability coefficient value, and the model stability data calculation model is specifically expressed as: , Wherein, the Representing the model stability coefficient value of the ith calculation, ms i represents the model convergence rate of the ith acquisition, vf i represents the variance expansion factor of the ith acquisition.
  8. 8. The method for generating a large scientific data model based on multidimensional analysis as claimed in claim 1, wherein the model adaptability data analysis node is used for establishing a model adaptability data calculation model, and importing the model adaptability data transmitted in the data acquisition step into the model adaptability data calculation model to obtain a model adaptability coefficient value, and the model adaptability data calculation model is specifically expressed as: , Wherein, the Representing the model adaptation capability coefficient value of the ith calculation, ar i represents the new data adaptation rate of the ith acquisition, and Dc i represents the data scale expansibility of the ith acquisition.
  9. 9. The method for generating a large data model of science popularization based on multidimensional analysis of claim 1, wherein the large data model of science popularization generation data analysis unit is used for establishing a large data model of science popularization generation data calculation model, and importing the model input complexity coefficient value, the resource adaptation coefficient value, the feature validity coefficient value, the model innovation coefficient value, the model stability coefficient value and the model adaptation capability coefficient value transmitted in the data analysis step into the large data model of science popularization generation data calculation model to obtain a comprehensive evaluation index value of the large data model of science popularization generation, wherein the large data model of science popularization generation data calculation model is specifically expressed as: , wherein A represents the calculated scientific creation big data model to generate a comprehensive evaluation index value, The model representing the ith calculation inputs a complexity coefficient value, A resource fitness coefficient value representing the ith calculation, A feature validity coefficient value representing the ith calculation, The model innovation coefficient value representing the ith calculation, Representing the calculated model innovation coefficient minimum, Representing the calculated model innovation coefficient maximum, The model stability coefficient value representing the ith calculation, Representing the calculated minimum value of the stability coefficient of the model, Representing the calculated maximum value of the stability coefficient of the model, The model adaptability coefficient value representing the ith calculation, Representing the calculated model adaptation capability coefficient minimum, The calculated maximum value of the model adaptability coefficient is represented, i represents the number from the ith number, and n represents the number to the nth number.
  10. 10. A multi-dimensional analysis-based large scientific data model generating system for implementing the multi-dimensional analysis-based large scientific data model generating method according to any one of claims 1 and 9, comprising: The data batch dividing module is used for determining the data to be acquired as target data, dividing the target data into different batches according to an equal time dividing mode, and marking the different batches as1, 2 in sequence; the data acquisition module comprises a model input data acquisition unit and a model feedback data acquisition unit and is used for acquiring target data in real time and transmitting the acquired data to the data analysis module. The model input data acquisition unit is used for acquiring technical basic data, resource association data and association characteristic data; the feedback data acquisition unit is used for acquiring innovation theory result data, model stability data and model adaptability data; The data analysis module comprises a model input data analysis unit and a model feedback data analysis unit, and is used for collecting target data in real time and transmitting the collected data to the data comprehensive evaluation module. The feedback data analysis unit comprises an innovation theory result data analysis node, a model stability data analysis node and a model adaptability data analysis node; the data comprehensive evaluation module comprises a scientific large data model generation data analysis unit and a warning feedback module, wherein the scientific large data model generation data analysis unit is used for comprehensively analyzing the analysis result transmitted by the data analysis module and transmitting the analysis result to the warning feedback module; The early warning feedback module is used for establishing a scientific creation big data model to generate a comprehensive evaluation index preset value, judging the comprehensive evaluation index value generated by the scientific creation big data model according to the comprehensive evaluation index preset value generated by the scientific creation big data model, and sending out a corresponding signal according to a judging result.

Description

Multi-dimensional analysis-based scientific creation big data model generation method and system Technical Field The invention relates to the technical field of scientific and invasive big data model generation, in particular to a method and a system for generating a scientific and invasive big data model based on multidimensional analysis. Background With the vigorous development of technological innovation and the continuous promotion of digital transformation, the position of the field of science and technology creation in the global economy is increasingly highlighted, and the requirements on the excavation, analysis and application of science and technology creation data are also reached to unprecedented heights. The multi-dimensional analysis-based scientific creation big data model generation technology brings new challenges and opportunities to the scientific creation field due to the key effects of the technology on scientific creation decision and development. Accurate big data model generation can ensure the degree of depth insight and the effective utilization of scientific data, satisfies the demand of scientific high-quality development. The existing method for generating the scientific large data model mainly comprises a data acquisition module, a data preprocessing module, a model construction module and a model evaluation module. The method comprises the steps of acquiring data including scientific research project information, technical innovation results, market trend data, talent flowing conditions and the like by arranging intelligent acquisition tools at key positions of scientific research institution databases, innovation enterprise information systems, patent platforms and the like through a data acquisition module, collecting the scientific research data, denoising, standardizing and the like through a data preprocessing module by utilizing a data cleaning algorithm to acquire the acquired data, guaranteeing the quality and usability of the data in a model construction stage, constructing a scientific research big data model through a model construction module according to a specific algorithm and a specific framework by combining the preprocessed data, excavating internal association and potential modes among the data, providing model support for strategic planning of scientific research development, intuitively knowing the performance condition of the model through a model evaluation module, realizing adjustment and optimization of model parameters, timely correcting a model structure and improving the accuracy and stability of the scientific research big data model. However, this method still has some drawbacks in practical applications. For example, in terms of data acquisition, due to the diversity and complexity of data sources and the compatibility of acquisition tools, there may be some omission of the acquisition of the science-created data, which results in insufficient comprehensive acquired data, and this has adverse effects on the subsequent model construction, and in terms of model evaluation, facing the complex science-created environment and diversified evaluation indexes, there is a problem that the evaluation results are easily inaccurate and comprehensive, which may result in the failure to discover potential defects in the model in time, and reduce the capability of the science-created large data model to guide the science-created practice. Therefore, there is an urgent need to provide a method and a system for generating a large scientific data model based on multidimensional analysis, so as to solve the problems of insufficient comprehensive data acquisition and inaccurate evaluation result of the existing method and system for generating the large scientific data model, and further improve the quality and reliability of the scientific activity. Disclosure of Invention In order to overcome the defects in the prior art, the embodiment of the invention provides a method and a system for generating a scientific and invasive big data model based on multidimensional analysis, which solve the problems in the background art through the following scheme. In order to achieve the purpose, the invention provides the following technical scheme that the method for generating the scientific creation big data model based on multidimensional analysis comprises the following steps: S1, data batch division, namely determining data to be acquired as target data, dividing the target data into different batches according to an equal time division mode, and marking the different batches as 1, 2. And S2, a data acquisition step, which comprises a model input data acquisition unit and a model feedback data acquisition unit, and is used for acquiring target data in real time and transmitting the acquired data to a data analysis step. The model input data acquisition unit is used for acquiring technical basic data, resource association data and association characteristic data; the feed