Technical life cycle identification method, system and device based on science and technology big data, and medium

Abstract

The invention relates to a technology life cycle identification method, system and device based on technology big data and a medium, and the method comprises the steps: collecting initial multi-source data, classifying the initial multi-source data according to technology categories, obtaining multi-source data, and determining the field attribute of each category of technology and the annual quality index of each piece of data based on the multi-source data; acquiring identification characteristic parameters for each domain attribute, wherein the identification characteristic parameters comprise quality sensitivity and a periodic conversion coefficient; calculating an annual quality weight based on the quality sensitivity and the annual quality index; for each technology, performing S curve fitting based on the annual quality weight to obtain an initial S curve model and obtain core parameters; based on the core parameter and the cycle conversion coefficient, performing domain adaptive cycle life cycle threshold calculation, based on the cycle life cycle threshold, obtaining a stage node, and drawing a life cycle curve; the system, the equipment and the medium are used for implementing the method. Compared with the prior art, the life cycle identification method can realize accurate life cycle identification in any different technical fields.

Description

Technical Field

[0001] This invention relates to the field of science and technology development analysis and technology assessment technology, and in particular to a method, system, device and medium for identifying the technology life cycle based on big data of science and technology. Background Technology

[0002] Technology Life Cycle Theory (TCLT) is an early theoretical framework applied in academia for technology identification. This theory posits that technological development, over time and in response to changes in the industry environment, exhibits phased transformations resembling a life cycle. Therefore, TCLT can describe the complete process of technological development from its inception to its demise. By dividing technological development into stages within its life cycle according to TCLT, researchers can better grasp technological trends and proactively identify promising technological areas. After long-term refinement and development, TCLT is increasingly being applied to technology identification to help governments, research institutions, and enterprises make appropriate methodological choices and allocate resources more rationally at different stages of technological development. Because patents, documents, and papers record the progress of emerging technologies and the main achievements of technological innovation, patent documents and papers are important data sources and references for determining the stage of technological development. By analyzing key information such as keywords, patent classifications, applicants, and the number of applications in patent documents and papers, the stage of technological development can be divided based on relevant characteristics, and the development trend of the technology can be analyzed. For example, Chinese patent application CN120197748A provides a method for identifying and recognizing the technology life cycle based on the analysis of scientific papers and patent documents. This includes: obtaining time-series data on the number of published scientific papers and patent applications, performing S-curve fitting to identify the time nodes of the nascent, growth, maturity, and decline stages of basic research and technological application in the technology field; and then using the basic research stage and technological application stage as the horizontal and vertical axes, constructing a life cycle matrix based on the time overlap areas of the basic research stage and technological application stage. While this scheme achieves the identification and recognition of the development stages of science and technology, it has the following drawbacks: 1) When performing S-curve fitting, it only uses the number of publications as the fitting input, ignoring the impact of the quality of the publications on the cycle. This drawback is particularly evident when identifying the cycle of emerging technologies. Emerging fields may be misjudged as being in the growth stage due to an inflated number of low-quality conference papers, when in reality, there is a lack of high-quality core research, and the field is in its nascent stage, leading to a disconnect between stage identification and reality; 2) When dividing the life cycle, it relies on a fixed threshold, ignoring the fact that the duration of each development stage varies in different technological fields, which may lead to problems such as the stage being identified being too late or too early.

[0003] Therefore, providing a method that can accurately identify the lifecycle of any different technological field is a technical problem that needs to be solved. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a solution.

[0005] The objective of this invention can be achieved through the following technical solutions: According to a first aspect of the present invention, a method for identifying the technology lifecycle based on big data in science and technology is provided, the method comprising: Initial multi-source data, including patents, papers, and project data, is collected and classified according to technology categories to obtain multi-source data. Based on the multi-source data, the domain attributes of each technology category and the annual quality indicators of each data are determined. For each of the aforementioned domain attributes, identification feature parameters are obtained, including quality sensitivity and periodic conversion coefficient; based on the quality sensitivity and annual quality indicators, an annual quality weight is calculated. For each technology, an initial S-curve model is obtained by fitting an S-curve based on the multi-source data and annual quality weights, and core parameters are obtained, including saturation value, inversion point and growth time. Based on the core parameters and cycle conversion coefficients, the domain-adaptive cycle lifecycle threshold is calculated, and the stage nodes are obtained based on the cycle lifecycle threshold, and the lifecycle curve is plotted.

[0006] As a preferred technical solution, the method for determining the aforementioned domain attributes is as follows: For each type of technology, the publication time of the first core paper and the release time of the first core patent are obtained based on patent and paper data. The time difference between the publication time of the first core paper and the release time of the first core patent is calculated as the technology transformation cycle. The total number of patent applications and the total number of papers published in the past N years are statistically analyzed from the multi-source data mentioned above, and the patent-to-paper ratio is calculated. The number of industry promotion projects in the past N years is statistically analyzed, and the ratio of the number of industry promotion projects to the total number of projects is calculated as the proportion of industrialization projects. If the technology transformation cycle is less than the first transformation cycle threshold, and at the same time the patent-to-paper ratio is greater than the first number threshold, and the proportion of industrialization projects is greater than the first proportion threshold, then the corresponding technology is determined to be a technology-oriented technology. If the technology transformation cycle is greater than the second transformation cycle threshold, and at the same time the patent-to-paper ratio is less than the second number threshold, and the proportion of industrialization projects is less than the second proportion threshold, then the corresponding technology is determined to be a research-oriented technology. If the technology transformation cycle is between the first transformation cycle threshold and the second transformation cycle threshold, and the number of patents-papers is between the first number threshold and the second number threshold, and the proportion of industrialization projects is between the first proportion threshold and the second proportion threshold, then the corresponding technology is determined to be a hybrid-oriented technology. The first conversion cycle threshold is less than the second conversion cycle threshold, the first quantity threshold is greater than the second quantity threshold, and the first proportion threshold is greater than the second proportion threshold.

[0007] As a preferred technical solution, the quality sensitivity includes adjustable patent quality sensitivity, paper quality sensitivity, and project quality sensitivity; and each quality sensitivity is a preset value when performing the initial S-curve model fitting, and the quality sensitivity always satisfies the following condition: The project quality sensitivity value is fixed, and the sum of the patent quality sensitivity and the paper quality sensitivity is 1. For the aforementioned technology-oriented technologies, the sensitivity to patent quality is greater than the sensitivity to paper quality. For the research-oriented technologies mentioned above, the sensitivity to paper quality is greater than that to patent quality. For the aforementioned hybrid-guided technology, the sensitivity values ​​for patent quality and paper quality are equal.

[0008] As a preferred technical solution, the formula for calculating the cycle conversion coefficient is as follows: , , , in, This represents the cycle conversion coefficient of technology-oriented technologies; Indicates the calculation period; This represents the growth rate of the number of patents or papers related to technology-oriented technologies in year i. This represents the growth rate of the number of patents or papers related to technology-oriented technologies over the past N years. This represents the cycle conversion coefficient of research-oriented technologies; This represents the growth rate of the number of patents or papers related to research-oriented technologies in year i. This represents the growth rate of the number of patents or papers related to research-oriented technologies over the past N years. This represents the cycle conversion coefficient of hybrid-oriented technologies; Indicates quality sensitivity.

[0009] As a preferred technical solution, the method for obtaining the initial S-curve model includes: For each of the aforementioned technologies, the number of papers, patents, and projects is obtained by year. Based on the domain attributes corresponding to the technology, and combined with the annual quality weight and quality sensitivity, the annual score for each year is calculated. The annual scores are then accumulated by year to obtain the true annual comprehensive score. Based on the aforementioned actual annual comprehensive score, the general patterns of the life cycle are extracted using the S-curve model, and an initial S-curve model is constructed as follows: , in, This represents the annual composite quality fit value for year t; Indicates the saturation value; Indicates the time of growth; Indicates the inflection point; This represents the mean of quality indicators for multi-source data; Indicates the growth acceleration factor; This represents the inflection point adjustment factor, and is relevant to technology-oriented technologies. For research-oriented technologies For hybrid guidance technology , Indicates sensitivity to patent quality. Indicates sensitivity to paper quality; Calculate the goodness of fit and the normal distribution of residuals of the initial S-curve model. If the goodness of fit and the normal distribution of residuals meet the threshold requirements, the fitting is successful; otherwise, adjust the quality sensitivity and refit the S-curve.

[0010] As a preferred technical solution, the method for calculating the domain-adaptive lifecycle threshold includes: obtaining the cycle conversion coefficient of each technology according to the domain attributes, calculating the lifecycle threshold ratio of each stage based on the cycle conversion coefficient, and calculating the lifecycle threshold based on the lifecycle threshold ratio and the saturation value, as follows: ,in, Indicates the saturation value. This represents the lifecycle threshold proportion of stage x.

[0011] As a preferred technical solution, the method for calculating the lifecycle threshold ratio is as follows: , in, This represents the initial value of the lifecycle threshold ratio for stage x, which is a preset value. Indicates the adjustment factor; Represents the periodic conversion coefficient, and , This represents the cycle conversion coefficient of technology-oriented technologies. This represents the cycle conversion coefficient of research-oriented technologies. This represents the cycle conversion coefficient of hybrid-oriented technologies; This represents the average value of the periodic conversion coefficient.

[0012] According to a second aspect of the present invention, a technology lifecycle identification system based on big data of science and technology is provided, the system comprising: The data acquisition and processing module is used to collect initial multi-source data, including patent, paper and project data, and classify them according to technology categories to obtain multi-source data. Based on the multi-source data, the domain attributes of each technology and the annual quality indicators of each data are determined. The weight calculation module is used to obtain identification feature parameters for each of the aforementioned domain attributes, including quality sensitivity and periodic conversion coefficient; and to calculate the annual quality weight based on the quality sensitivity and annual quality indicators. The S-curve fitting module is used to perform S-curve fitting for each technology based on the multi-source data and annual quality weight to obtain an initial S-curve model and acquire core parameters, including saturation value, inversion point and growth time. The lifecycle curve plotting module is used to calculate the domain-adaptive lifecycle threshold based on the core parameters and the cycle conversion coefficient, obtain the stage nodes based on the lifecycle threshold, and plot the lifecycle curve.

[0013] According to a third aspect of the present invention, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the program to implement the method described thereon.

[0014] According to a fourth aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method described thereon.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1) To address the shortcomings of existing technologies, this invention first classifies technologies by domain attributes, categorizing them into technology-oriented, research-oriented, and hybrid-oriented types. Different quality sensitivities are assigned to technologies with different domain attributes, and different calculation methods are used to calculate the cycle conversion coefficient. Finally, when fitting the S-curve, the corresponding quality sensitivity and cycle conversion coefficient are incorporated into the fitting process according to the domain attributes, ensuring that life cycle curves that are more consistent with reality can be generated for different types of technologies, effectively avoiding misjudgments of life cycle caused by a large amount of low-quality data.

[0016] 2) In this invention, the iteration speed is quantified according to the characteristics of different fields. Specifically, for technology-oriented fields, the variance of the patent number growth rate is used as the benchmark for iteration speed; for research-oriented fields, the variance of the paper number growth rate is used as the benchmark; and for hybrid-oriented fields, both are combined and weighted by the field quality sensitivity coefficient to accurately characterize the differences in the development speed of different fields. The life cycle threshold of different stages is corrected by the calculated cycle conversion coefficient to make the stage duration characteristics of different fields more consistent with reality. Attached Figure Description

[0017] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0019] Example 1 Current S-curve fitting methods often use the total number of published papers as the core input indicator, without classifying and weighting the quality of the papers. This deficiency can easily lead to biases in identifying the cyclical nature of emerging technologies. In the early stages of exploration, emerging technologies are often accompanied by a rapid influx of low-quality literature, such as short conference papers that haven't undergone rigorous peer review, opinion pieces lacking empirical data, or highly repetitive review articles. This concentrated publication of literature can make the curve appear to rise rapidly, characteristic of a growth phase. However, in reality, the field may lack high-quality research published in core journals, possessing original theoretical breakthroughs or key technological verifications. The true technological core remains in the nascent stage of conceptual exploration and feasibility verification. Such judgments based solely on quantity can lead to a severe disconnect between technological stage identification and actual development status, potentially misleading policymakers or investors into investing resources in areas that lack a mature development foundation, resulting in resource misallocation.

[0020] Furthermore, existing lifecycle segmentation methods generally rely on preset fixed thresholds, such as using the inflection point of the S-curve or a specific slope as the critical value for stage division. However, this approach ignores the significant differences in development pace across different technological fields. The duration of each stage of the technology lifecycle is influenced by multiple factors, including disciplinary foundation, R&D difficulty, and market demand, exhibiting strong domain specificity. For example, the nascent stage of a new technology primarily driven by research may last for decades, while a technology primarily driven by technological development may complete the transition from the growth stage to the maturity stage in just a few years. If a uniform fixed threshold is used to determine the stage of all technologies, technologies with long nascent stages may be misjudged as being in the growth stage before they mature; for technologies rapidly advancing in the growth stage, the optimal intervention or strategic positioning may be missed due to the lag in the threshold, ultimately leading to significant problems of premature or delayed stage determination and affecting the scientific nature of technology development strategies.

[0021] To address the two technical problems mentioned above, this invention provides a technology lifecycle identification method based on big data in science and technology. This method involves fitting S-curves to collected multi-source data according to domain attributes, and incorporating data quality guidance during the fitting process to generate an S-curve that better reflects the actual situation. The process is as follows: Figure 1 ,include: S1. Collect initial multi-source data including patents, papers and project data and classify them according to technology categories to obtain multi-source data. Based on the multi-source data, determine the domain attributes of each technology and the annual quality indicators of each data.

[0022] S11. Initial multi-source data acquisition and preprocessing.

[0023] Initial multi-source data was collected, including patent, paper, and project data. Patent data included patent classification numbers, patent texts, patent time data, and the number of patent families; paper data included paper abstracts, paper time data, and the number of citations; and project data included project start and end dates, project keywords, project funding levels, and project implementation status.

[0024] S111. Perform targeted retrieval and preliminary screening on the initial multi-element data, and retain the initial multi-source data with a technical relevance greater than or equal to 0.7, where the technical relevance is the ratio of the frequency of core keywords extracted based on the initial multi-source data to the total frequency of keywords.

[0025] S112. Clean the data after initial screening and the reverse data, including deduplication, outlier removal and missing value imputation.

[0026] S113. Data standardization: For data of different magnitudes, such as the number of patent families and the frequency of paper citations, min-max standardization is used to map them to the [0,1] interval.

[0027] S114. For standardized initial multi-source data, classify them according to technical categories based on patent classification numbers, paper abstracts, and project keywords.

[0028] S12. Determine the technical field attributes.

[0029] S121. For each type of technology, based on patent and paper data, obtain the publication time of the first core paper and the release time of the first core patent, and calculate the time difference between the publication time of the first core paper and the release time of the first core patent as the technology transformation cycle.

[0030] S122. Calculate the patent-to-paper ratio by analyzing the total number of patent applications and the total number of papers published in the past 5 years from multiple sources of data.

[0031] S123. Calculate the ratio of the number of industrial promotion projects to the total number of projects in the past 5 years from the statistical project data, and use this ratio as the proportion of industrialization projects.

[0032] S124. Based on the technology transformation cycle, the patent-to-paper ratio, and the proportion of industrialization projects, the domain attributes are determined. In this step, a first transformation cycle threshold, a second transformation cycle threshold, a first quantity threshold, a second quantity threshold, a first proportion threshold, and a second proportion threshold are set. The first transformation cycle threshold is less than the second transformation cycle threshold, the first quantity threshold is greater than the second quantity threshold, and the first proportion threshold is greater than the second proportion threshold.

[0033] Specifically, if the technology transfer cycle is less than the first transfer cycle threshold, and simultaneously the patent-to-paper ratio is greater than the first threshold, and the proportion of industrialization projects is greater than the first proportion threshold, then the corresponding technology is determined to be a technology-oriented technology. If the technology transfer cycle is greater than the second transfer cycle threshold, and simultaneously the patent-to-paper ratio is less than the second threshold, and the proportion of industrialization projects is less than the second proportion threshold, then the corresponding technology is determined to be a research-oriented technology. If the technology transfer cycle is between the first and second transfer cycle thresholds, and simultaneously the patent-to-paper ratio is between the first and second thresholds, and the proportion of industrialization projects is between the first and second proportion thresholds, then the corresponding technology is determined to be a hybrid-oriented technology.

[0034] S13. Calculation of annual quality indicators.

[0035] For the data in the papers, the annual quality indicators are: ,in, Represents the balance coefficient. This represents the average normalized citation count of papers in year i. This represents the average journal quartile score of papers published in year i, such as... .

[0036] For patent data, the annual quality indicators are: , Indicates the balance coefficient; This represents the average standardized value of the same family in year i; denoted as the average legal status score of the patent in year i, where the score is set to 1 for granted patents, 0.5 for patents under examination, and 0.1 for rejected patents.

[0037] For project data, the annual quality indicators are: ,in, Indicates the balance coefficient; This represents the average funding level score for year i, where national level is set to 1, provincial level to 0.8, and municipal level to 0.6. This represents the average acceptance score of the project in year i, where 1 is set to excellent, 0.6 to qualified, and 0 to unqualified.

[0038] S2. Obtain identification feature parameters for each domain attribute. The identification feature parameters include quality sensitivity and periodic conversion coefficient. Calculate the annual quality weight based on quality sensitivity and annual quality indicators.

[0039] S21. Obtaining the quality sensitivity coefficient.

[0040] Quality sensitivity includes adjustable patent quality sensitivity, paper quality sensitivity, and project quality sensitivity; and each quality sensitivity is preset to a value when performing the initial S-curve model fitting, with the corresponding value set based on the current situation. The quality sensitivity always meets the following conditions: The project quality sensitivity value is fixed, and the sum of the patent quality sensitivity and the paper quality sensitivity is 1. For technology-oriented technologies, the patent quality sensitivity is greater than the paper quality sensitivity. For research-oriented technologies, the paper quality sensitivity is greater than the patent quality sensitivity. For hybrid-oriented technologies, the patent quality sensitivity and the paper quality sensitivity are equal.

[0041] S22, Obtaining the periodic conversion coefficient.

[0042] In this invention, considering that the duration of each stage of the lifecycle of different domain attributes is not the same, and their corresponding cycle conversion coefficients are also different, different methods are used to calculate the cycle conversion coefficient for technologies of different domain attributes. The calculation formula is as follows: , , , in, This represents the cycle conversion coefficient of technology-oriented technologies; Indicates the calculation period; This represents the growth rate of the number of patents or papers related to technology-oriented technologies in year i. This represents the growth rate of the number of patents or papers related to technology-oriented technologies over the past N years. This represents the cycle conversion coefficient of research-oriented technologies; This represents the growth rate of the number of patents or papers related to research-oriented technologies in year i. This represents the growth rate of the number of patents or papers related to research-oriented technologies over the past N years. This represents the cycle conversion coefficient of hybrid-oriented technologies; Indicates quality sensitivity.

[0043] S23, Calculation of annual quality weights.

[0044] In this invention, after normalizing the annual quality indicators, the corresponding annual quality weight is obtained by multiplying the quality sensitivity by the corresponding annual quality indicator according to the domain attribute.

[0045] S3. For each technology, an initial S-curve model is obtained by fitting an S-curve based on the multi-source data and annual quality weight, and core parameters, including saturation value, inversion point and growth time, are obtained.

[0046] S31. For each type of technology, obtain the number of papers, patents, and projects by year. Based on the domain attribute corresponding to the technology, and combined with the annual quality weight and quality sensitivity, calculate the annual score for each year. Accumulate the annual scores by year to obtain the true annual comprehensive score, expressed as follows: , in, This represents the annual comprehensive score calculated based on multi-source data up to year t. Indicates the starting year of the first paper or patent for the technology to be analyzed; Let $i$ represent the annual score for year $i$, and , , as well as These represent sensitivity to patent quality, sensitivity to academic paper quality, and sensitivity to project quality, respectively. , as well as These represent the number of patents, papers, and projects in year i, respectively. , as well as The weights represent the annual quality weights of patents, papers, and projects in year i.

[0047] S32. Based on the actual annual comprehensive score, extract the general patterns of the life cycle through the S-curve model, and construct the initial S-curve model as follows: , in, This represents the annual composite quality fit value for year t; Indicates the saturation value; Indicates the time of growth; Indicates the inflection point; This represents the mean of quality indicators for multi-source data; Indicates the growth acceleration factor; This represents the inflection point adjustment factor, and is relevant to technology-oriented technologies. For research-oriented technologies For hybrid guidance technology , Indicates sensitivity to patent quality. This indicates sensitivity to the quality of the paper.

[0048] S33. Calculate the goodness of fit and the normal distribution of residuals of the initial S-curve model. If the goodness of fit and the normal distribution of residuals meet the threshold requirements, the fitting is successful; otherwise, adjust the quality sensitivity and refit the S-curve.

[0049] S34. After successful fitting, directly obtain the saturation value, inversion point, and growth time. The saturation value represents the ceiling of the technology's life cycle growth, the inversion point represents the dividing year between the growth and maturity stages, and the growth time represents the cumulative value from... Growth to The effort required is used to reflect the speed of growth in the field.

[0050] S4. Calculate the domain-adaptive cycle lifecycle threshold based on core parameters and cycle conversion coefficients, obtain stage nodes based on the cycle lifecycle threshold, and draw the lifecycle curve.

[0051] S41. Obtain the cycle conversion coefficient for each technology according to the domain attribute.

[0052] S42. Calculate the life cycle threshold ratio for each stage based on the cycle conversion coefficient. The expression is as follows: , in, This represents the initial value of the lifecycle threshold ratio for stage x, which is a preset value. Indicates the adjustment factor; Represents the periodic conversion coefficient, and , This represents the cycle conversion coefficient of technology-oriented technologies. This represents the cycle conversion coefficient of research-oriented technologies. This represents the cycle conversion coefficient of hybrid-oriented technologies; This represents the average value of the periodic conversion coefficient.

[0053] S43. Calculate the lifecycle threshold based on the lifecycle threshold ratio and saturation value, as follows: ,in, Indicates the saturation value. This represents the lifecycle threshold proportion of stage x.

[0054] S44, Found The corresponding year is used as a stage node, and the rationality of the node is verified. If it is not reasonable, it is refitted; otherwise, a life cycle curve is drawn.

[0055] Example 2 This invention also provides a technology lifecycle identification system based on big data in science and technology, characterized in that the system includes: The data acquisition and processing module is used to collect initial multi-source data, including patents, papers and project data, and classify them according to technology categories to obtain multi-source data. Based on the multi-source data, the domain attributes of each technology and the annual quality indicators of each data are determined.

[0056] The weight calculation module is used to obtain identification feature parameters for each domain attribute, including quality sensitivity and periodic conversion coefficient; and to calculate the annual quality weight based on quality sensitivity and annual quality indicators.

[0057] The S-curve fitting module is used to perform S-curve fitting for each technology based on the multi-source data and annual quality weights to obtain an initial S-curve model and acquire core parameters, including saturation value, inversion point, and growth time.

[0058] The lifecycle curve plotting module is used to calculate the domain-adaptive lifecycle threshold based on core parameters and cycle conversion coefficients, obtain stage nodes based on the lifecycle threshold, and plot the lifecycle curve.

[0059] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the described module can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0060] The present invention also provides an electronic device including a central processing unit (CPU) that can perform various appropriate actions and processes according to computer program instructions stored in read-only memory (ROM) or loaded from a storage unit into random access memory (RAM). The RAM may also store various programs and data required for device operation. The CPU, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0061] Multiple components in the device are connected to the I / O interface, including: input units such as keyboards and mice; output units such as various types of displays and speakers; storage units such as disks and optical discs; and communication units such as network interface cards (NICs), modems, and wireless transceivers. The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0062] The processing unit executes the various methods and processes described above, such as methods S1 to S4. For example, in some embodiments, methods S1 to S4 may be implemented as computer software programs tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and / or installed on the device via ROM and / or a communication unit. When the computer program is loaded into RAM and executed by the CPU, one or more steps of methods S1 to S4 described above may be performed. Alternatively, in other embodiments, the CPU may be configured to execute methods S1 to S4 by any other suitable means (e.g., by means of firmware).

[0063] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0064] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0065] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0066] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and such modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for identifying the technology lifecycle based on big data in science and technology, characterized in that, The methods include: Initial multi-source data, including patents, papers, and project data, is collected and classified according to technology categories to obtain multi-source data. Based on the multi-source data, the domain attributes of each technology category and the annual quality indicators of each data are determined. For each of the aforementioned domain attributes, identification feature parameters are obtained, including quality sensitivity and periodic conversion coefficient; based on the quality sensitivity and annual quality indicators, an annual quality weight is calculated. For each technology, an initial S-curve model is obtained by fitting an S-curve based on the multi-source data and annual quality weights, and core parameters are obtained, including saturation value, inversion point and growth time. Based on the core parameters and cycle conversion coefficients, the domain-adaptive cycle lifecycle threshold is calculated, and the stage nodes are obtained based on the cycle lifecycle threshold, and the lifecycle curve is plotted.

2. The technology lifecycle identification method based on big data in science and technology according to claim 1, characterized in that, The method for determining the aforementioned domain attributes is as follows: For each type of technology, the publication time of the first core paper and the release time of the first core patent are obtained based on patent and paper data. The time difference between the publication time of the first core paper and the release time of the first core patent is calculated as the technology transformation cycle. The total number of patent applications and the total number of papers published in the past N years are statistically analyzed from the multi-source data mentioned above, and the patent-to-paper ratio is calculated. The number of industry promotion projects in the past N years is statistically analyzed, and the ratio of the number of industry promotion projects to the total number of projects is calculated as the proportion of industrialization projects. If the technology transformation cycle is less than the first transformation cycle threshold, and at the same time the patent-to-paper ratio is greater than the first number threshold, and the proportion of industrialization projects is greater than the first proportion threshold, then the corresponding technology is determined to be a technology-oriented technology. If the technology transformation cycle is greater than the second transformation cycle threshold, and at the same time the patent-to-paper ratio is less than the second number threshold, and the proportion of industrialization projects is less than the second proportion threshold, then the corresponding technology is determined to be a research-oriented technology. If the technology transformation cycle is between the first transformation cycle threshold and the second transformation cycle threshold, and the number of patents-papers is between the first number threshold and the second number threshold, and the proportion of industrialization projects is between the first proportion threshold and the second proportion threshold, then the corresponding technology is determined to be a hybrid-oriented technology. The first conversion cycle threshold is less than the second conversion cycle threshold, the first quantity threshold is greater than the second quantity threshold, and the first proportion threshold is greater than the second proportion threshold.

3. The method for identifying the technology lifecycle based on big data in science and technology according to claim 2, characterized in that, The aforementioned quality sensitivity includes numerically adjustable patent quality sensitivity, paper quality sensitivity, and project quality sensitivity; and each quality sensitivity is a preset value when performing the initial S-curve model fitting, and the quality sensitivity always satisfies the following condition: The project quality sensitivity value is fixed, and the sum of the patent quality sensitivity and the paper quality sensitivity is 1. For the aforementioned technology-oriented technologies, the sensitivity to patent quality is greater than the sensitivity to paper quality. For the research-oriented technologies mentioned above, the sensitivity to paper quality is greater than that to patent quality. For the aforementioned hybrid-guided technology, the sensitivity values ​​for patent quality and paper quality are equal.

4. The technology lifecycle identification method based on big data in science and technology according to claim 1, characterized in that, The formula for calculating the periodic conversion coefficient is as follows: ， ， ， in, This represents the cycle conversion coefficient of technology-oriented technologies; Indicates the calculation period; This represents the growth rate of the number of patents or papers related to technology-oriented technologies in year i. This represents the growth rate of the number of patents or papers related to technology-oriented technologies over the past N years. This represents the cycle conversion coefficient of research-oriented technologies; This represents the growth rate of the number of patents or papers related to research-oriented technologies in year i. This represents the growth rate of the number of patents or papers related to research-oriented technologies over the past N years. This represents the cycle conversion coefficient of hybrid-oriented technologies; Indicates quality sensitivity.

5. The technology lifecycle identification method based on big data in science and technology according to claim 1, characterized in that, The method for obtaining the initial S-curve model includes: For each of the aforementioned technologies, the number of papers, patents, and projects is obtained by year. Based on the domain attributes corresponding to the technology, and combined with the annual quality weight and quality sensitivity, the annual score for each year is calculated. The annual scores are then accumulated by year to obtain the true annual comprehensive score. Based on the aforementioned actual annual comprehensive score, the general patterns of the life cycle are extracted using the S-curve model, and an initial S-curve model is constructed as follows: ， in, This represents the annual composite quality fit value for year t; Indicates the saturation value; Indicates the time of growth; Indicates the inflection point; This represents the mean of quality indicators for multi-source data; Indicates the growth acceleration factor; This represents the inflection point adjustment factor, and is relevant to technology-oriented technologies. For research-oriented technologies For hybrid guidance technology , Indicates sensitivity to patent quality. Indicates sensitivity to paper quality; Calculate the goodness of fit and the normal distribution of residuals of the initial S-curve model. If the goodness of fit and the normal distribution of residuals meet the threshold requirements, the fitting is successful; otherwise, adjust the quality sensitivity and refit the S-curve.

6. The method for identifying the technology lifecycle based on big data in science and technology according to claim 1, characterized in that, The method for calculating the adaptive lifecycle threshold for a specific domain includes: obtaining the cycle conversion coefficient for each technology according to domain attributes, calculating the lifecycle threshold ratio for each stage based on the cycle conversion coefficient, and calculating the lifecycle threshold based on the lifecycle threshold ratio and the saturation value. ,in, Indicates the saturation value. This represents the lifecycle threshold proportion of stage x.

7. The technology lifecycle identification method based on big data in science and technology according to claim 5, characterized in that, The calculation method for the aforementioned lifecycle threshold ratio is as follows: ， in, This represents the initial value of the lifecycle threshold ratio for stage x, which is a preset value. Indicates the adjustment factor; Represents the periodic conversion coefficient, and , This represents the cycle conversion coefficient of technology-oriented technologies. This represents the cycle conversion coefficient of research-oriented technologies. This represents the cycle conversion coefficient of hybrid-oriented technologies; This represents the average value of the periodic conversion coefficient.

8. A technology lifecycle identification system based on big data in science and technology, characterized in that, The system includes: The data acquisition and processing module is used to collect initial multi-source data, including patent, paper and project data, and classify them according to technology categories to obtain multi-source data. Based on the multi-source data, the domain attributes of each technology and the annual quality indicators of each data are determined. The weight calculation module is used to obtain identification feature parameters for each of the aforementioned domain attributes, including quality sensitivity and periodic conversion coefficient; and to calculate the annual quality weight based on the quality sensitivity and annual quality indicators. The S-curve fitting module is used to perform S-curve fitting for each technology based on the multi-source data and annual quality weight to obtain an initial S-curve model and acquire core parameters, including saturation value, inversion point and growth time. The lifecycle curve plotting module is used to calculate the domain-adaptive lifecycle threshold based on the core parameters and the cycle conversion coefficient, obtain the stage nodes based on the lifecycle threshold, and plot the lifecycle curve.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 7.

Images