Patent data comprehensive evaluation system and evaluation method for industrial gathering and divergence trend

Abstract

The invention discloses a patent data comprehensive evaluation system for industry gathering and divergence trend, and the system comprises a data collection module which is used for determining the patent data analysis dimension of a target industry chain, constructing a retrieval strategy for patent data, and downloading the retrieval strategy; the data processing module is used for cleaning data information, performing data indexing on the data and determining a patent data set; the spatial weight matrix construction module is used for analyzing the geographic spatial critical state of the target industrial chain in the target area based on the patent data set information and constructing a spatial weight matrix; the Moran index calculation module is used for calculating application data of a target professional chain according to patent data by adopting a global Moran index and local Moran index calculation method; and the result analysis module is used for analyzing the technical concentration degree of the target industrial chain in combination with the Moran index result. The system uses an analysis framework of patent data and economic geography Moran index to accurately measure and analyze the industrial technology concentration degree.

Description

Technical Field

[0001] This invention relates to the technical field of industrial technology concentration analysis, and in particular to a comprehensive patent data evaluation system and method for assessing industrial agglomeration and divergence trends. Background Technology

[0002] Currently, scholars both domestically and internationally have conducted extensive research on the measurement of industrial agglomeration, which mainly focuses on the following aspects: ① Research on industrial agglomeration measurement based on location entropy (Ru et al., 2019; Wu et al., 2021; Han et al., 2021; Zhang Caiyun, 2020; Tang Jianrong, 2021) measures the degree of industrial agglomeration by the spatial distribution of regional factors, and is used to determine whether industrial agglomeration exists. Commonly used measurement indicators include total industrial output, industrial added value, number of employees, and sales revenue. ② Research using industrial agglomeration index (Xie et al., 2021; Sheng Bin et al., 2022; Zhang Zhixin et al., 2022) dynamically measures the industrial agglomeration and dispersion trends of a certain industry in a certain region over a certain period of time, and predicts the industrial agglomeration situation in the region in the future. It measures the industrial agglomeration situation and comparative advantage industries of a certain region in a specific period of time, and infers the trend of agglomeration and dispersion of this industry in the region in the future. ③ Research on industry concentration (Elisabetta et al., 2020; Ferschli et al., 2021) measures the level of industry agglomeration from a market space perspective. It can also be used to analyze the sensitivity of changes in the market share of the largest companies in an industry. ④ Research based on the Herfindahl-Hirschman Index (Liu et al., 2018) can be used to examine the market concentration of an industry or firm. When measuring the regional concentration of an industry, it considers two influencing factors: the total number of firms and the size of firms, thus overcoming the shortcomings of the industry concentration index to some extent. It is also used to reflect the degree of market monopoly and competition changes in an industry and to study the sensitivity of changes in firms within an industry (Pan Yufeng, 2021; Deng Zhongqi et al., 2022). ⑤ Research using the spatial Gini coefficient (Fu Shuke et al., 2018; Chen et al., 2021) uses the ratio of the size of all industries in a region to the total size of all industries in the country as a variable, considering the impact of differences in the size of geographical units in different regions on industry concentration. ⑥ The E-G index is a method proposed by Ellision and Glaeser (1997) to overcome the shortcomings of the spatial Gini coefficient mentioned above, and is used to measure the level of regional industrial agglomeration. Due to its numerous advantages in measuring the level of industrial agglomeration, the EG index has been used by more and more scholars (Wei et al., 2018).In addition to the six methods mentioned above, researchers at home and abroad have developed new measurement tools, such as the DO index and the DEA-Malmquist model. For example, Ding Pengfei (2019) used the DO index based on point-to-point distance to calculate the degree of agglomeration of major three-digit manufacturing industries in Shanghai at different spatial scales and found that high-tech industries such as electronic information, automobiles, and biomedicine have the greatest agglomeration. Kong et al. (2021) used the DEA-Malmquist model to analyze the development efficiency, spatiotemporal evolution characteristics, and spatial improvement of China's innovative industrial clusters.

[0003] While existing research has yielded excellent results, these methods also have certain limitations. For example, location entropy and spatial Gini coefficient do not take into account differences in firm size, and the Herschman-Hirschman-Frandall index requires precision down to the firm level. Therefore, when using these methods to study the degree of industrial agglomeration, it is often difficult to obtain accurate data, or the research results are often inaccurate. Thus, there is an urgent need to find new methods suitable for measuring industrial agglomeration. Summary of the Invention

[0004] In view of the shortcomings and deficiencies of existing technologies, the technical problem that this invention aims to solve is mainly to analyze the degree of industrial technology concentration by using patent data and Moran's index in economic geography.

[0005] The specific technical solution of this invention is as follows:

[0006] This invention provides a comprehensive patent data evaluation system for industrial agglomeration and divergence trends, the system comprising:

[0007] The data collection module is configured to determine the patent data analysis dimensions of the target industry chain, construct search strategies for patent data in each dimension, and download them.

[0008] The data processing module is configured to clean the downloaded data, index the cleaned data, and determine the patent dataset of the target industry chain in different analytical dimensions.

[0009] The spatial weight matrix construction module is configured to analyze the geographic spatial critical state of the target industry chain within the target region based on patent dataset information, and construct a spatial weight matrix based on geographic proximity.

[0010] The Moran index calculation module is configured to use the calculation methods of global Moran index and local Moran index, and calculates the spatial autocorrelation of the degree of industrial technology concentration based on the application data of the target professional chain using patent data.

[0011] The results analysis module is configured to combine Moran's index results to analyze the degree of technology concentration in the target industry chain and identify areas of clustering and dispersion.

[0012] Preferably, the data collection module includes:

[0013] The patent analysis dimension determination unit is configured to determine the patent data analysis dimensions of the target industry based on the upstream, midstream, and downstream relationships of the industry chain. These dimensions include patent application data, authorized data, valid invention patent data, patent asset index data, patent citation and cited data, and patent litigation and transfer information.

[0014] The search strategy construction unit is configured to construct search strategies by combining keywords and classification numbers, limiting time ranges, applicant and inventor information, and filtering information based on patent legal status.

[0015] The data download unit is configured to execute a search strategy in the patent database and download the required patent data.

[0016] Preferably, the data processing module includes a data cleaning unit and a data indexing unit. The data cleaning specifically includes format conversion, field extraction and format processing, normalization processing, data deduplication, outlier processing, and missing value processing. The data indexing unit includes sub-units for reading and indexing data one by one, batch indexing, collaborative indexing, specialized classification indexing, subject indexing, and patent indexing for special professional information.

[0017] Preferably, the spatial weight matrix construction module includes a geospatial critical state analysis unit and a spatial weight matrix construction unit;

[0018] The geospatial critical state analysis unit includes:

[0019] The spatial data analysis subunit is configured to use geographic information systems and spatial analysis methods to identify and analyze key features and patterns in geospatial space.

[0020] The critical point identification subunit is configured to identify critical points by analyzing geospatial data, the critical points including places where significant changes occur in ecosystems, economic activities or social structures;

[0021] Spatial heterogeneity sub-units are configured to add geographic spatial heterogeneity data during the analysis process, i.e., the differences between different regions;

[0022] The Earth Boundary Concept Subunit is configured to analyze the impact of human activities on the Earth system and how these impacts approach or exceed the Earth's carrying capacity, with reference to the Earth Boundary Concept Framework.

[0023] The spatial weight matrix construction unit specifically includes:

[0024] The adjacency definition sub-unit is configured to define the adjacency relationship between spatial units, which can be based on boundary contact or corner contact;

[0025] The weighting subunit is configured to assign weights to the adjacency relationships between spatial units; the weights are binary, indicating whether an adjacency relationship exists; or the weights are continuous, indicating the strength or distance of the adjacency relationship.

[0026] GIS software uses sub-cells to create spatial weight matrices;

[0027] The spatial statistics software processing subunit is configured to use spatial statistics software to process and analyze the spatial weight matrix;

[0028] The data format conversion subunit is configured to convert the spatial weight matrix from a GAL file to a CSV file.

[0029] The weight matrix verification subunit is configured to verify the accuracy and rationality of the spatial weight matrix.

[0030] Preferably, the Moran index calculation module includes a global Moran index calculation unit and a local Moran index calculation unit;

[0031] The global Moran's index calculation unit is configured to use the global Moran's index formula to calculate the average correlation between the entire region and surrounding areas:

[0032] The formula for calculating the global Moran index is as follows:

[0033]

[0034] in, n is the total number of spatial units, y i and y j Let represent the attribute values ​​of the i-th and j-th spatial units respectively, y be the mean of the attribute values ​​of all spatial units, and w be the mean of the attribute values ​​of all spatial units. ij This represents the spatial weight value;

[0035] The local Moran index calculation unit is configured to use the local Moran index formula to calculate the correlation between the attribute value of a specific spatial unit and the surrounding area;

[0036] The formula for calculating the local Moran index is as follows:

[0037]

[0038] Where n is the total number of spatial units, y i and y j Let represent the attribute values ​​of the i-th and j-th spatial units respectively, y be the mean of the attribute values ​​of all spatial units, and w be the mean of the attribute values ​​of all spatial units. ij This represents the spatial weight value.

[0039] Preferably, the result analysis module includes a Moran index interpretation unit and an industry clustering and divergence analysis unit;

[0040] The Moran index interpretation unit interprets the Moran index as follows, based on the Moran index's value range of -1 to 1:

[0041] Spatial positive correlation: When Moran's I>0, it indicates spatial positive correlation. The larger the value, the more obvious the spatial correlation, indicating a more significant industrial agglomeration phenomenon.

[0042] Spatial negative correlation: When Moran's I < 0, it indicates spatial negative correlation. The smaller the value, the greater the spatial difference, indicating that the industrial divergence phenomenon is more obvious.

[0043] Spatial randomness: When Moran's I = 0, the space exhibits randomness, indicating that the distribution of attribute values ​​in space does not show a clear trend of clustering or dispersion;

[0044] The industrial agglomeration and divergence analysis unit is configured to combine Moran's index results to analyze the degree of technology concentration in specific industries and identify agglomeration and divergence regions; the industrial agglomeration and divergence analysis unit includes:

[0045] The cluster area identification sub-unit is configured to identify industrial cluster areas through the calculation of the global Moran index; a high global Moran index value indicates that the industry is highly clustered in the corresponding area.

[0046] The divergence region identification sub-unit is configured to use low or even negative global Moran indices to indicate that the industry is showing a divergence trend in the corresponding region.

[0047] The local Moran index analysis subunit is configured to use the local Moran index to identify agglomeration or divergence phenomena in a specific region, and to discover local hot spots and cold spots, i.e., local spatial patterns of industrial agglomeration and divergence.

[0048] The sub-unit for analyzing industrial technology concentration is configured to combine Moran's index results to analyze the degree of technology concentration in a specific industry; a high Moran's index value indicates that technology resources and innovation activities are highly concentrated in the corresponding region, while a low value indicates that technology resources are relatively dispersed.

[0049] The policy-making and industrial planning sub-unit is configured to provide a scientific basis for policy-making and industrial planning through the analysis results of the Moran Index.

[0050] The second aspect of this invention discloses a method for comprehensive evaluation of patent data on industrial agglomeration and divergence trends, the method comprising:

[0051] S10. Determine the patent data analysis dimensions of the target industry chain, construct retrieval strategies for the patent data of each dimension, and download them;

[0052] S20. Clean the downloaded data information, perform data indexing on the cleaned data, and determine the patent dataset of the target industry chain in different analytical dimensions.

[0053] S30. Based on patent dataset information, analyze the geographic spatial critical state of the target industry chain within the target region and construct a spatial weight matrix based on geographic proximity.

[0054] S40. Using the calculation methods of global Moran index and local Moran index, and based on the application data of the target professional chain using patent data, the spatial autocorrelation of the degree of industrial technology concentration is described.

[0055] S50. Based on the Moran Index results, analyze the degree of technology concentration in the target industry chain and identify areas of clustering and dispersion.

[0056] Based on the above-described inventive principles, the beneficial effects of this invention are as follows:

[0057] 1. Comprehensive patent data collection covering the entire industry chain

[0058] This invention's evaluation system and method comprehensively cover the patent data scope of a specific industry by determining the upstream, midstream, and downstream relationships of the industry chain, including multiple dimensions such as the number of patent applications and grants. This method can more accurately capture the full picture of industry technology concentration because it considers not only the quantity of patents but also their quality and other relevant indicators, thus providing a more comprehensive perspective.

[0059] 2. Efficient retrieval strategies and data downloading

[0060] The evaluation system and method of this invention construct an efficient retrieval strategy capable of accurately downloading relevant patent data, thereby improving the efficiency and accuracy of data processing. This method reduces data collection time and costs, while also minimizing analytical errors caused by inaccurate or incomplete data.

[0061] 3. Advanced data cleaning and indexing processes

[0062] The evaluation system and method of this invention employ advanced data cleaning and indexing processes to ensure data quality and consistency, laying a solid foundation for subsequent analysis. This method improves the reliability of analysis results and avoids misleading conclusions due to data quality issues by removing invalid, duplicate, or incomplete data and accurately indexing the cleaned data.

[0063] 4. In-depth analysis of critical states in geospatial space

[0064] The evaluation system and method of this invention take a spatial geography perspective, deeply analyzing the geospatial critical state of the study area, which helps to better understand the spatial distribution characteristics of industrial technology concentration. This method is more advanced than existing methods that may ignore spatial factors because it considers the impact of geographical proximity on industrial technology concentration, thus providing deeper insights.

[0065] 5. Combined use of global and local Moran indices

[0066] The evaluation system and method of this invention simultaneously utilize both the global and local Moran indices, which facilitates the analysis of industry technology concentration from both macro and micro perspectives. This method can reveal the overall trend and local characteristics of industry technology concentration, providing a more comprehensive analysis than existing methods that may only use a single Moran indice.

[0067] 6. Depth and breadth of results analysis

[0068] This invention's assessment system and method, combined with Moran's index results, provide an in-depth analysis of the technology concentration level in specific industries, identifying areas of clustering and dispersion. This helps provide a scientific basis for policy-making and industrial planning. By providing more in-depth analytical results, this method helps policymakers better understand the current status and trends of industrial technology concentration, thereby making more informed decisions.

[0069] The above-described technical solution of the present invention provides a more comprehensive, accurate and in-depth method for analyzing industrial technology concentration. This method not only improves the efficiency and accuracy of the analysis, but also enhances the reliability and practicality of the results. Attached Figure Description

[0070] Figure 1 This is a schematic diagram of the overall structure of the patent data comprehensive evaluation system for the industrial agglomeration and divergence trends of the present invention.

[0071] Figure 2 This is a flowchart illustrating the patent data comprehensive evaluation method for the industrial agglomeration and divergence trends of this invention. Detailed Implementation

[0072] The following embodiments further illustrate the content of the present invention, but should not be construed as limiting the present invention. Any modifications or substitutions made to the methods, steps, or conditions of the present invention without departing from the spirit and essence of the invention are within the scope of the present invention.

[0073] This invention discloses a comprehensive patent data evaluation system and method for assessing industrial agglomeration and divergence trends. It is used to analyze the degree of technology concentration in specific industries. The system mainly comprises five modules: patent data collection, data processing and indexing, construction of a spatial weight matrix, Moran's index calculation, and result analysis. The technical problem this invention aims to solve is primarily through analyzing the degree of industrial technology concentration using patent data and Moran's index from economic geography.

[0074] In a specific example, this invention provides a comprehensive evaluation system for patent data on industrial agglomeration and divergence trends. Figure 1 As shown, the system includes

[0075] The data collection module is configured to determine the patent data analysis dimensions of the target industry chain, construct search strategies for patent data in each dimension, and download them.

[0076] The data processing module is configured to clean the downloaded data, index the cleaned data, and determine the patent dataset of the target industry chain in different analytical dimensions.

[0077] The spatial weight matrix construction module is configured to analyze the geographic spatial critical state of the target industry chain within the target region based on patent dataset information, and construct a spatial weight matrix based on geographic proximity.

[0078] The Moran index calculation module is configured to use the calculation methods of global Moran index and local Moran index, and calculates the application data of the target professional chain based on patent data to describe the spatial autocorrelation of the degree of industrial technology concentration.

[0079] The results analysis module is configured to combine Moran's index results to analyze the degree of technology concentration in the target industry chain and identify areas of clustering and dispersion.

[0080] The data collection module includes a patent analysis dimension determination unit, a search strategy construction unit, and a data download unit.

[0081] The patent analysis dimension determination unit is configured to determine the patent data analysis dimensions of the target industry chain based on the upstream, midstream, and downstream relationships. These dimensions include patent application data, authorized data, valid invention patent data, patent asset index data, patent citation and cited data, patent litigation and transfer information, etc., to achieve comprehensive coverage of all aspects of the industry chain.

[0082] It should be noted that before executing the patent analysis dimension determination unit, relevant data on enterprises in the target industry's industrial chain and supply chain within the target region are collected; the supply and demand relationships between enterprises in the industrial chain and supply chain, as well as the branch relationships between headquarters and branch enterprises, are determined; the target industry is divided into upstream, midstream, and downstream sectors, and enterprises in the industrial chain and supply chain are further divided into upstream, midstream, and downstream sectors respectively. A specific example is the classification of patents in the pharmaceutical industry chain according to the relationship between upstream pharmaceutical raw materials, midstream drug research and development, and downstream drug application.

[0083] The search strategy construction unit is configured to build search strategies based on information such as keyword and classification number combinations, time range limits, applicant and inventor information, and patent legal status filtering, in order to ensure the comprehensiveness and accuracy of the data.

[0084] The data download unit is configured to execute a search strategy in the patent database and download the required patent data. The patent database includes official websites such as the European Patent Office (EPO) and the China National Intellectual Property Administration (CNIPA), or other data sources. In addition to patent databases, open data platforms, academic databases, and industry reports can also be used as data sources to increase the diversity and comprehensiveness of the data. The data download unit can also be configured to automatically collect patent data from the internet using web crawler technology, reducing manual intervention and improving the efficiency and real-time nature of data collection.

[0085] The data processing module includes a data cleaning unit and a data indexing unit. Data cleaning involves screening, filtering, correcting, and transforming the raw data to ensure its accuracy, integrity, and consistency. In the context of patent data, data cleaning includes the following key steps:

[0086] ① Format conversion: Convert patent data in different formats (such as txt, xml, csv, json, etc.) into a format suitable for analysis, usually csv or xls format;

[0087] ② Field extraction and format processing: Extract useful fields from patent data, such as extracting country information from publication number and extracting application year from applicant;

[0088] ③ Normalization process: Normalize the names of applicants or inventors to resolve issues caused by inconsistent spelling or different names of parent and subsidiary companies;

[0089] ④ Data deduplication: Delete duplicate patent records and retain unique data records to avoid bias during data analysis;

[0090] ⑤ Outlier handling: Identify and handle outlier data points, possibly by deleting, pruning, or replacing outlier values ​​with values ​​that are closer to other data points;

[0091] ⑥ Handling missing values: For missing data, methods such as deletion, interpolation, using default values, or extrapolation can be used to handle it.

[0092] It should be noted that, in another example, the data processing module uses other data cleaning techniques, such as machine learning methods like cluster analysis, to automatically identify and process outliers and noisy data, thereby improving the accuracy of data cleaning.

[0093] The data indexing unit indexes the cleaned data to identify each patent dataset, preparing it for subsequent analysis. Data indexing is the process of converting the natural language of patents into a retrieval language; it includes patent asset indexing and patent information indexing. The following are the key steps of data indexing:

[0094] ① Read the indexes one by one: If the number of patents is small or there is enough time, the patent information can be read one by one for precise indexing.

[0095] ② Batch indexing: For large amounts of data, tools can be used for batch indexing. Before this, the data needs to be cleaned to remove noise.

[0096] ③ Collaborative indexing: The indexing task is assigned to multiple people. The patent data is downloaded to an Excel spreadsheet, assigned to multiple people for indexing, and then imported online for analysis.

[0097] ④ Specialized classification indexing: Based on different patent databases, specialized classification indexing is implemented, such as F-TERM, ECLA, US patent classification indexing, etc.

[0098] ⑤ Subject indexing: Subject indexing of patents facilitates subject retrieval, such as in the CLAIMS patent database and the Derwent WPI database.

[0099] ⑥ Patent indexing for specialized information: Constructing corresponding databases for chemical structure retrieval and gene sequence retrieval, such as the databases constructed by CAS and Derwent Publications Ltd.

[0100] Furthermore, the data indexing can also employ other advanced data indexing technologies. In one example, natural language processing techniques, such as named entity recognition (NER), are used to automatically extract key information from patent data for indexing, improving the efficiency and accuracy of indexing.

[0101] The evaluation system of this invention has developed a comprehensive data cleaning and indexing process to remove invalid, duplicate or incomplete data and accurately index the cleaned data, providing a high-quality dataset for subsequent analysis.

[0102] The spatial weight matrix construction module includes a geospatial critical state analysis unit and a spatial weight matrix construction unit.

[0103] The geospatial critical state analysis unit analyzes the geospatial critical state of the study area from a spatial geography perspective. This process involves a deep understanding and analysis of geospatial data to identify critical points or critical regions in the geospatial space. Specifically, the geospatial critical state analysis unit includes:

[0104] The spatial data analysis subunit is configured to use GIS (Geographic Information System) and spatial analysis methods, such as overlay analysis, buffer analysis and network analysis, to identify and analyze key features and patterns in geospatial space.

[0105] The critical point identification subunit is configured to identify critical points by analyzing geospatial data, the critical points including places where significant changes occur in ecosystems, economic activities or social structures;

[0106] Spatial heterogeneity sub-units are configured to add geographic spatial heterogeneity data during the analysis process, i.e., the differences between different regions, which is crucial for understanding the critical state of geospatiality.

[0107] The Earth Boundary Concept Subunit is configured to analyze the impacts of human activities on the Earth system, and how these impacts approach or exceed the Earth's carrying capacity, with reference to the Earth Boundary Concept Framework.

[0108] In addition to being based on geographical proximity, based on the technical principles of this invention, a spatial weight matrix can also be constructed based on economic ties, transportation networks, etc., to reflect the actual connections between different spatial units.

[0109] The spatial weight matrix construction unit is an important tool for describing the geographical proximity relationships between different spatial units. It plays a core role in spatial statistical analysis, and its functions specifically include:

[0110] The adjacency definition sub-unit is configured to define the adjacency relationship between spatial units. The adjacency relationship can be based on boundary contact (Queen adjacency) or corner contact (Rook adjacency).

[0111] The weight allocation subunit is configured to assign weights to the adjacency relationships between spatial units; the weights are binary (0 or 1) to indicate whether an adjacency relationship exists. If two regions are adjacent, the weight is 1, and if they are not adjacent, the weight is 0; or the weights are continuous to indicate the strength or distance of the adjacency relationship.

[0112] GIS software uses sub-cells to create spatial weight matrices; for example, ArcGIS provides tools to generate and transform spatial weight matrices to make them suitable for different spatial statistical analyses.

[0113] The spatial statistics software processing subunit is configured to use spatial statistics software to process and analyze spatial weight matrices, such as the spdep package in the R language.

[0114] The data format conversion subunit is configured to convert the spatial weight matrix from one format to another to facilitate sharing and use across different software and tools; for example, converting the spatial weight matrix from a GAL file to a CSV file.

[0115] The weight matrix verification sub-unit is configured to verify the accuracy and rationality of the spatial weight matrix, ensuring that it correctly reflects the actual geographical relationships between spatial units.

[0116] In another example, the spatial weight matrix construction unit can use a distance matrix to construct a spatial weight matrix, specifically by using the centroid distance or administrative center distance between regions to construct a spatial weight matrix that reflects the spatial distance relationship between regions.

[0117] The Moran index calculation module includes a global Moran index calculation unit and a local Moran index calculation unit. This module introduces calculation methods for the global Moran index and the local Moran index to quantify the spatial autocorrelation of patent data to describe the degree of industrial technology concentration.

[0118] The global Moran's index calculation unit is configured to use the global Moran's index formula to calculate the average correlation between the entire region and surrounding areas:

[0119] The formula for calculating the global Moran index is as follows:

[0120]

[0121] in, n is the total number of spatial units, y i and y j Let represent the attribute values ​​of the i-th and j-th spatial units respectively, y be the mean of the attribute values ​​of all spatial units, and w be the mean of the attribute values ​​of all spatial units. ij This represents the spatial weight value;

[0122] The local Moran index calculation unit is configured to use the local Moran index formula to calculate the correlation between the attribute value of a specific spatial unit and the surrounding area;

[0123] The formula for calculating the local Moran index is as follows:

[0124]

[0125] Where n is the total number of spatial units, y i and y j Let represent the attribute values ​​of the i-th and j-th spatial units respectively, y be the mean of the attribute values ​​of all spatial units, and w be the mean of the attribute values ​​of all spatial units. ij This represents the spatial weight value.

[0126] For the calculation module of the Moran's index, in addition to using statistical software, one example is that it can also be calculated using Geographic Information System (GIS) software or specialized spatial statistical software, which provide more intuitive visualization and analysis capabilities. In yet another example, in Geoda software, a Monte Carlo simulation method is used to perform a randomization test on the Moran's index, improving the robustness of the results.

[0127] The spatial matrix of the Moran Index calculation module is calculated using geographical adjacency relationships, with adjacent elements having a weight of 1 and non-adjacent elements having a weight of 0. Then, based on patent data, such as the number of patent applications, patent grants, and patent validity, calculations are performed for upstream raw materials, midstream drug development, and downstream drug application data. Specifically, the results analysis module includes a Moran Index interpretation unit and an industry clustering and divergence analysis unit.

[0128] Moran's I is a statistic that measures the degree of clustering of data points, commonly used in spatial autocorrelation analysis. It reflects the spatial distribution pattern of a specific attribute value, indicating whether it tends to cluster or disperse spatially. The Moran's I ranges from -1 to 1, explained as follows:

[0129] ① Spatial positive correlation: When Moran's I > 0, it indicates spatial positive correlation, meaning that similar attribute values ​​tend to cluster spatially. The larger the value, the more obvious the spatial correlation, indicating a more significant industrial agglomeration phenomenon.

[0130] ② Spatial negative correlation: When Moran's I < 0, it indicates spatial negative correlation, meaning that similar attribute values ​​tend to be spatially dispersed. The smaller the value, the greater the spatial difference, indicating a more obvious industrial divergence.

[0131] ③ Spatial randomness: When Moran's I = 0, the space exhibits randomness, indicating that the attribute values ​​do not show a clear trend of clustering or dispersion in space. Industrial clustering and dispersion analysis: Combining Moran's index results, the degree of technological concentration in specific industries is analyzed to identify areas of clustering and dispersion.

[0132] The industrial clustering and divergence analysis unit, combined with the Moran index results, can analyze the degree of technology concentration in a specific industry and identify areas of clustering and divergence.

[0133] ① Cluster Identification: By calculating the global Moran's index, regions where industries cluster can be identified. A high global Moran's index value indicates a high degree of industrial clustering in the corresponding region.

[0134] ② Identification of divergent regions: Conversely, a low or even negative global Moran index may indicate that the industry is diverging in the corresponding region.

[0135] ③ Local Moran's I Analysis: The Local Moran's I can further identify clustering or dispersion phenomena within a specific region. This helps to discover local hotspots and coldspots, i.e., local spatial patterns of industrial clustering and dispersion.

[0136] ④ Industry Technology Concentration Analysis: Combining the Moran Index results, the degree of technology concentration in a specific industry can be analyzed. A high Moran Index value may indicate that technological resources and innovation activities are highly concentrated in the corresponding region, while a low value may indicate that technological resources are more dispersed.

[0137] ⑤ Policy Making and Industrial Planning: The Moran Index analysis results can provide a scientific basis for policy making and industrial planning. For example, policymakers can provide more support to industrial clusters to promote regional economic development, or take measures in industrial dispersion areas to promote balanced industrial development.

[0138] In addition to the functional solutions mentioned above, the results analysis module can also introduce more spatial statistical analysis methods. For example, it can combine other spatial statistical analysis methods such as geographic detectors to provide a more comprehensive analysis of industrial clustering and divergence. Alternatively, it can use data visualization tools such as Tableau or Power BI to combine Moran's index results with other geographical and economic data to provide more intuitive analysis results.

[0139] The second aspect of this invention provides a method for comprehensive evaluation of patent data on industrial agglomeration and divergence trends. Figure 2 As shown, the method includes

[0140] S10. Determine the patent data analysis dimensions of the target industry chain, construct retrieval strategies for the patent data of each dimension, and download them;

[0141] S20. Clean the downloaded data information, perform data indexing on the cleaned data, and determine the patent dataset of the target industry chain in different analytical dimensions.

[0142] S30. Based on patent dataset information, analyze the geographic spatial critical state of the target industry chain within the target region and construct a spatial weight matrix based on geographic proximity.

[0143] S40. Using the calculation methods of global Moran index and local Moran index, and based on the application data of the target professional chain using patent data, the spatial autocorrelation of the degree of industrial technology concentration is described.

[0144] S50. Based on the Moran Index results, analyze the degree of technology concentration in the target industry chain and identify areas of clustering and dispersion.

[0145] Specifically, step S10 includes the following steps:

[0146] S11. Based on the upstream, midstream, and downstream relationships of the target industry chain, determine the patent data analysis dimensions for that industry, including patent application data, authorized data, valid invention patent data, patent asset index data, patent citation and cited data, and patent litigation and transfer information.

[0147] S12. Construct a search strategy by combining keywords and classification numbers, limiting the time range, filtering information on applicants and inventors, and filtering information on patent legal status;

[0148] S13. Execute the search strategy in the patent database and download the required patent data.

[0149] The data cleaning in step S20 specifically includes operations such as format conversion, field extraction and format processing, normalization processing, data deduplication, outlier processing, and missing value processing; the data indexing in step S20 specifically includes one or more of the following: reading indexing based on specific data, batch indexing, collaborative indexing, specialized classification indexing, subject indexing, and patent indexing for special professional information.

[0150] The geospatial critical state analysis in step S30 specifically includes the following methods:

[0151] S31. Using geographic information systems and spatial analysis methods, identify and analyze key features and patterns in geospatial space;

[0152] S32. By analyzing geospatial data, identify critical points, which include places where significant changes occur in ecosystems, economic activities, or social structures;

[0153] S33. Consider the heterogeneity of geographical space, that is, the differences between different regions;

[0154] S34. Referring to the concept framework of Earth's limits, analyze the impact of human activities on the Earth system and how these impacts approach or exceed the Earth's carrying capacity.

[0155] The construction of the spatial weight matrix in step S30 specifically includes the following methods:

[0156] S35. Define the adjacency relationship between spatial units, which can be based on boundary contact or corner contact;

[0157] S36. Assign weights to the adjacency relationships between spatial units; the weights are binary, indicating whether an adjacency relationship exists; or the weights are continuous, indicating the strength or distance of the adjacency relationship.

[0158] S37. Create a spatial weight matrix using GIS software;

[0159] S38. Use spatial statistical software to process and analyze the spatial weight matrix;

[0160] S39. Convert the spatial weight matrix from the GAL file to a CSV file to verify the accuracy and rationality of the spatial weight matrix.

[0161] Step S40 specifically includes:

[0162] S41. Use the global Moran's index formula to calculate the average degree of correlation with surrounding areas across the entire region:

[0163] The formula for calculating the global Moran index is as follows:

[0164]

[0165] in, n is the total number of spatial units, y i and y j Let represent the attribute values ​​of the i-th and j-th spatial units respectively, y be the mean of the attribute values ​​of all spatial units, and w be the mean of the attribute values ​​of all spatial units. ij This represents the spatial weight value.

[0166] S42. Use the local Moran index formula to calculate the correlation between the attribute value of a specific spatial unit and the surrounding area;

[0167] The formula for calculating the local Moran index is as follows:

[0168]

[0169] Where n is the total number of spatial units, y i and y j Let represent the attribute values ​​of the i-th and j-th spatial units respectively, y be the mean of the attribute values ​​of all spatial units, and w be the mean of the attribute values ​​of all spatial units. ij This represents the spatial weight value.

[0170] The specific method for step S50 is as follows:

[0171] S51. Identify industrial clusters by calculating the global Moran index; a high global Moran index value indicates that industries are highly concentrated in the corresponding region.

[0172] S52. Low or even negative global Moran's Index values ​​indicate that industries are showing a divergent trend in the corresponding regions.

[0173] S53. The local Moran index identifies agglomeration or dispersion phenomena in specific areas, and discovers local hot spots and cold spots, that is, local spatial patterns of industrial agglomeration and dispersion.

[0174] S54. Based on the Moran Index results, analyze the degree of technology concentration in a specific industry; a high Moran Index value indicates that technology resources and innovation activities are highly concentrated in the corresponding region, while a low value indicates that technology resources are more dispersed.

[0175] S55. The analysis results of the Moran Index provide a scientific basis for policy making and industrial planning.

[0176] The following is a specific example illustrating the patent data comprehensive evaluation method based on the industrial agglomeration and divergence trends of this invention; evaluating the industrial concentration of global antibody drugs in major countries.

[0177] In terms of the geographical distribution of global antibody drug patent applicants, global antibody drug industry patents are mainly distributed across 92 countries / regions, with the United States, China, and Japan being the primary regions where antibody drug innovators have established their patent portfolios. The United States (denoted as y...) 11 The highest number of antibody drug patent applications was 4,310; China ranked second (denoted as y). 12 The number of antibody drug patent applications reached 4,086; the third is Japan (denoted as y). 13 The number of antibody drug patent applications reached 3,675; Australia and Canada (denoted as y) 14 and y 15 The number of antibody drug patent applications in the top 5 countries were 1790 and 1706, respectively. This example focuses on analyzing the industry concentration of antibody drugs in the top 5 countries globally. Based on the geographical proximity of the top 5 countries in global antibody drug patent applications, the first-order proximity spatial weight matrix W1 of the top 5 countries in global antibody drug patent applications is derived as follows:

[0178]

[0179] Calculated based on the global Moran index formula We obtain S0 = 4, where n = 5. Because y 11 =4310, y 12 =4086, y 13 =3675, y 14 =1790, y 15 =1706, so the average value of the 5 countries is y = 3113.4, and thus we get The formula can be used to calculate: Substituting the above calculation results into the global Moran's index formula, the global Moran's index for the global antibody drug industry is calculated as I1 = -0.0941. Since I1 < 0, this indicates that the overall industry concentration of the top 5 countries in the global antibody drug industry shows a trend of dispersion. To calculate the local Moran's index for each country, the formula for the local Moran's index is needed. Given the adjacent spatial weight matrix W1 and the average value of the 5 countries y = 3113.4, substituting these into the Moran's index formula, the I1 for the United States can be calculated separately. 11 China I 12 , Japan I 13 Australia I 14 and Canada I 15 Moran Index: I 11 =-0.1815、I 12 =0.9044, I 13 =1.7325, I 14 =0 and I 15 = -0.8509.

[0180] The analysis reveals spatial patterns in the distribution of antibody drug industry patents across different countries. China and Japan exhibit high concentration, while the United States and Canada show lower concentration. Australia, lacking proximity to the other four countries, shows no trend towards concentrated or dispersed patent distribution.

[0181] This invention utilizes patent data and the Moran's index from economic geography to analyze the degree of industrial technology concentration. Existing technologies for analyzing industrial technology concentration often lack comprehensive application of patent data, leading to potentially incomplete and superficial results. This invention aims to provide a more accurate method for analyzing industrial technology concentration by integrating and analyzing patent data. Although the Moran's index in economic geography is an effective tool for spatial autocorrelation analysis, its application in existing research is insufficient, failing to fully utilize its value in analyzing industrial technology concentration. This invention applies the Moran's index to patent data to reveal the spatial clustering characteristics and trends of industrial technology. Existing technologies often lack quantitative analytical methods for describing industrial technology concentration, making the measurement and comparison of concentration difficult. This invention provides a quantitative analytical approach by using patent data and the Moran's index to scientifically assess and compare the degree of technology concentration in different industries. Existing research lacks a deep understanding of the relationship between technological innovation and industrial concentration. This invention, through the combined analysis of patent data and the Moran's index, aims to reveal the complex relationship between the two, providing a scientific basis for policy formulation and industrial planning. With the development of the digital economy, its impact on industrial concentration is becoming increasingly significant. This invention will consider the moderating effect of the digital economy and analyze how it affects the relationship between industrial concentration and technological innovation in order to adapt to the new trends in economic development.

[0182] As can be seen from the above specific examples, the innovation and practicality of the technical solution of this invention in patent data analysis and industrial technology concentration assessment are mainly reflected in:

[0183] 1. Achieve precise determination of the scope of patent data in the industrial chain: Based on the upstream, midstream and downstream relationships of the industrial chain, systematically determine the scope of patent data for a specific industry, including the number of patent applications, the number of grants, the patent asset index, the number of valid invention patents, and the number of various patents in strategic emerging industries, so as to comprehensively cover all links of the industrial chain.

[0184] 2. Construct and download efficient search strategies: Construct efficient search strategies and accurately download relevant patent data from patent databases to ensure the comprehensiveness and accuracy of the data.

[0185] 3. Comprehensive data cleaning and indexing process: The system has developed a comprehensive data cleaning and indexing process to remove invalid, duplicate or incomplete data and accurately index the cleaned data, providing a high-quality dataset for subsequent analysis.

[0186] 4. Realize the analysis of critical state of geospatial space and the construction of spatial weight matrix: From the perspective of spatial geography, analyze the critical state of geospatial space of the study area and construct a spatial weight matrix of geographical adjacency to accurately describe the geographical proximity relationship between different spatial units.

[0187] 5. Calculation methods for global and local Moran indices: The calculation methods for global and local Moran indices are introduced, and the calculation is based on application data of the professional chain with patent data as the target, to describe the spatial autocorrelation of the degree of industrial technology concentration.

[0188] 6. Multidimensional interpretation and analysis of the Moran index: Based on the positive or negative sign and magnitude of the Moran index, multidimensional interpretations are provided to identify industrial clusters, random distributions, or divergences, and to analyze the degree of technological concentration in specific industries.

[0189] 7. Systematization of process steps and precise setting of technical parameters: The entire analysis process is broken down into clear process steps, and precise technical parameters are set for each step to ensure the repeatability of the analysis process and the reliability of the results.

[0190] 8. In-depth insights from the results analysis module: Combining Moran's index results, we conduct in-depth analysis of the technological concentration of specific industries, identify areas of clustering and dispersion, and provide a scientific basis for policy making and industrial planning.

[0191] The system described in some embodiments of this application also provides a controller, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the methods of the above embodiments, for example, performing all the steps of the methods described above.

[0192] The controller in this application embodiment includes one or more processors and a memory. The processor and the memory can be connected via a bus or other means. The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device.

[0193] In some implementations, the memory may optionally include memory remotely located relative to the processor, which can be connected to the controller via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0194] In some embodiments, when the processor executes a computer program, it executes the enterprise intellectual property capability profiling and analysis method and system of any of the above embodiments at preset intervals.

[0195] This application also provides a computer-readable storage medium storing computer-executable instructions for executing the above-described method for establishing and analyzing enterprise intellectual property capability profiles. For example, the computer-executable instructions can cause one or more processors to execute the method for establishing and analyzing enterprise intellectual property capability profiles in the above-described method embodiments, such as executing all the steps of the method described above.

[0196] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0197] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0198] 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 variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included 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 comprehensive patent data evaluation system for industrial agglomeration and divergence trends, characterized in that, The system includes The data collection module is configured to determine the patent data analysis dimensions of the target industry chain, construct search strategies for patent data in each dimension, and download them. The data processing module is configured to clean the downloaded data, index the cleaned data, and determine the patent dataset of the target industry chain in different analytical dimensions. The spatial weight matrix construction module is configured to analyze the geographic spatial critical state of the target industry chain within the target region based on patent dataset information, and construct a spatial weight matrix based on geographic proximity. The Moran index calculation module is configured to use the calculation methods of global Moran index and local Moran index, and calculates the spatial autocorrelation of the degree of industrial technology concentration based on the application data of the target professional chain using patent data. The results analysis module is configured to combine Moran's index results to analyze the degree of technology concentration in the target industry chain and identify areas of clustering and dispersion.

2. The patent data comprehensive evaluation method for industrial agglomeration and divergence trends as described in claim 1, characterized in that, The data collection module includes: The patent analysis dimension determination unit is configured to determine the patent data analysis dimensions of the target industry based on the upstream, midstream, and downstream relationships of the industry chain. The patent data includes patent application data, authorization data, valid invention patent data, patent asset index data, patent citation and cited data, and patent litigation and transfer information. The search strategy construction unit is configured to construct search strategies by combining keywords and classification numbers, limiting time ranges, applicant and inventor information, and filtering information based on patent legal status. The data download unit is configured to execute a search strategy in the patent database and download the required patent data.

3. The patent data comprehensive evaluation method for industrial agglomeration and divergence trends as described in claim 1, characterized in that, The data processing module includes a data cleaning unit and a data indexing unit; The data cleaning unit is configured to perform operations such as format conversion, field extraction and format processing, normalization, data deduplication, outlier handling, and missing value handling on the data information. The data indexing unit specifically includes the following sub-units: Read the indexing units one by one. If the number is small or there is enough time, read the patent information one by one and perform precise indexing. Batch indexing sub-units: For large amounts of data, tools are used for batch indexing. The collaborative indexing subunit assigns indexing tasks to multiple people, downloads patent data to an Excel spreadsheet, assigns it to multiple people for indexing, and then imports it online for analysis. Specialized classification indexing sub-units enable specialized classification indexing based on different patent databases; Subject indexing sub-units are used to index patents by subject, facilitating subject retrieval. Patent indexing units for specialized professional information are used to construct corresponding databases for chemical structure retrieval and gene sequence retrieval.

4. The patent data comprehensive evaluation method for industrial agglomeration and divergence trends as described in claim 1, characterized in that, The spatial weight matrix construction module includes a geospatial critical state analysis unit and a spatial weight matrix construction unit; The geospatial critical state analysis unit includes: The spatial data analysis subunit is configured to use geographic information systems and spatial analysis methods to identify and analyze key features and patterns in geospatial space. The critical point identification subunit is configured to identify critical points by analyzing geospatial data, the critical points including places where significant changes occur in ecosystems, economic activities or social structures; Spatial heterogeneity sub-units are configured to add geographic spatial heterogeneity data during the analysis process, i.e., the differences between different regions; The Earth Boundary Concept Subunit is configured to analyze the impact of human activities on the Earth system and how these impacts approach or exceed the Earth's carrying capacity, with reference to the Earth Boundary Concept Framework. The spatial weight matrix construction unit specifically includes: The adjacency definition sub-unit is configured to define the adjacency relationship between spatial units, which can be based on boundary contact or corner contact; The weighting subunit is configured to assign weights to the adjacency relationships between spatial units; the weights are binary, indicating whether an adjacency relationship exists; or the weights are continuous, indicating the strength or distance of the adjacency relationship. GIS software uses sub-cells to create spatial weight matrices; The spatial statistics software processing subunit is configured to use spatial statistics software to process and analyze the spatial weight matrix; The data format conversion subunit is configured to convert the spatial weight matrix from a GAL file to a CSV file. The weight matrix verification subunit is configured to verify the accuracy and rationality of the spatial weight matrix.

5. The patent data comprehensive evaluation method for industrial agglomeration and divergence trends as described in claim 1, characterized in that, The Moran index calculation module includes a global Moran index calculation unit and a local Moran index calculation unit; The global Moran's index calculation unit is configured to use the global Moran's index formula to calculate the average correlation between the entire region and surrounding areas: The formula for calculating the global Moran index is as follows: in, n is the total number of spatial units, y i and y j Let represent the attribute values ​​of the i-th and j-th spatial units respectively, y be the mean of the attribute values ​​of all spatial units, and w be the mean of the attribute values ​​of all spatial units. ij This represents the spatial weight value; The local Moran index calculation unit is configured to use the local Moran index formula to calculate the correlation between the attribute value of a specific spatial unit and the surrounding area; The formula for calculating the local Moran index is as follows: Where n is the total number of spatial units, y i and y j Let represent the attribute values ​​of the i-th and j-th spatial units respectively, y be the mean of the attribute values ​​of all spatial units, and w be the mean of the attribute values ​​of all spatial units. ij This represents the spatial weight value.

6. The patent data comprehensive evaluation method for industrial agglomeration and divergence trends as described in claim 1, characterized in that, The results analysis module includes a Moran index interpretation unit and an industrial clustering and divergence analysis unit. The Moran index interpretation unit interprets the Moran index as follows, based on the Moran index's value range of -1 to 1: Spatial positive correlation: When Moran's I > 0, it indicates spatial positive correlation. The larger the value, the more obvious the spatial correlation, indicating a more significant industrial agglomeration phenomenon. Spatial negative correlation: When Moran's I < 0, it indicates spatial negative correlation. The smaller the value, the greater the spatial difference, indicating that the industrial divergence phenomenon is more obvious. Spatial randomness: When Moran's I = 0, the space exhibits randomness, indicating that the distribution of attribute values ​​in space does not show a clear trend of clustering or dispersion; The industrial clustering and divergence analysis unit is configured to combine Moran's index results to analyze the degree of technological concentration in a specific industry and identify areas of clustering and divergence. The industrial clustering and divergence analysis unit includes: The cluster area identification sub-unit is configured to identify industrial cluster areas through the calculation of the global Moran index; a high global Moran index value indicates that the industry is highly clustered in the corresponding area. The divergence region identification sub-unit is configured to use low or even negative global Moran indices to indicate that the industry is showing a divergence trend in the corresponding region. The local Moran index analysis subunit is configured to use the local Moran index to identify agglomeration or divergence phenomena in a specific region, and to discover local hot spots and cold spots, i.e., local spatial patterns of industrial agglomeration and divergence. The sub-unit for analyzing industrial technology concentration is configured to combine Moran's index results to analyze the degree of technology concentration in a specific industry; a high Moran's index value indicates that technology resources and innovation activities are highly concentrated in the corresponding region, while a low value indicates that technology resources are more dispersed. The policy formulation and industrial planning sub-unit is configured to provide a scientific basis for policy formulation and industrial planning through the analysis results of the Moran Index.

7. A comprehensive evaluation method for patent data on industrial agglomeration and divergence trends, characterized in that, The method includes S10. Determine the patent data analysis dimensions of the target industry chain, construct retrieval strategies for the patent data of each dimension, and download them; S20. Clean the downloaded data information, perform data indexing on the cleaned data, and determine the patent dataset of the target industry chain in different analytical dimensions. S30. Based on patent dataset information, analyze the geographic spatial critical state of the target industry chain within the target region and construct a spatial weight matrix based on geographic proximity. S40. Using the calculation methods of global Moran index and local Moran index, and based on the application data of the target professional chain using patent data, the spatial autocorrelation of the degree of industrial technology concentration is described. S50. Based on the Moran Index results, analyze the degree of technology concentration in the target industry chain and identify areas of clustering and dispersion.

8. The patent data comprehensive evaluation method for industrial agglomeration and divergence trends as described in claim 7, characterized in that, Step S10 specifically includes the following steps: S11. Based on the upstream, midstream, and downstream relationships of the target industry chain, determine the patent data analysis dimensions for that industry, including patent application data, authorized data, valid invention patent data, patent asset index data, patent citation and cited data, and patent litigation and transfer information. S12. Construct a search strategy by combining keywords and classification numbers, limiting the time range, filtering information on applicants and inventors, and filtering information on patent legal status; S13. Execute the search strategy in the patent database and download the required patent data; The data cleaning in step S20 specifically includes operations such as format conversion, field extraction and format processing, normalization, data deduplication, outlier handling, and missing value handling; The data indexing in step S20 specifically includes one or more of the following: reading indexing based on specific data, batch indexing, collaborative indexing, specialized classification indexing, subject indexing, and patent indexing for special professional information.

9. The patent data comprehensive evaluation method for industrial agglomeration and divergence trends as described in claim 7, characterized in that, The geospatial critical state analysis in step S30 specifically includes the following methods: S31. Using geographic information systems and spatial analysis methods, identify and analyze key features and patterns in geospatial space; S32. By analyzing geospatial data, identify critical points, which include places where significant changes occur in ecosystems, economic activities, or social structures; S33. Consider the heterogeneity of geographical space, that is, the differences between different regions; S34. Referring to the concept framework of Earth's limits, analyze the impact of human activities on the Earth system and how these impacts approach or exceed the Earth's carrying capacity. The construction of the spatial weight matrix in step S30 specifically includes the following methods: S35. Define the adjacency relationship between spatial units, which can be based on boundary contact or corner contact; S36. Assign weights to the adjacency relationships between spatial units; The weight is binary, indicating whether an adjacency relationship exists; or the weight is continuous, indicating the strength or distance of the adjacency relationship. S37. Use GIS software to create a spatial weight matrix; S38. Use spatial statistical software to process and analyze the spatial weight matrix; S39. Convert the spatial weight matrix from the GAL file to a CSV file to verify the accuracy and rationality of the spatial weight matrix.

10. The patent data comprehensive evaluation method for industrial agglomeration and divergence trends as described in claim 7, characterized in that, Step S40 specifically includes: S41. Use the global Moran's index formula to calculate the average degree of correlation with surrounding areas across the entire region; The formula for calculating the global Moran index is as follows: in, n is the total number of spatial units, y i and y j Let represent the attribute values ​​of the i-th and j-th spatial units respectively, y be the mean of the attribute values ​​of all spatial units, and w be the mean of the attribute values ​​of all spatial units. ij This represents the spatial weight value; S42. Use the local Moran index formula to calculate the correlation between the attribute value of a specific spatial unit and the surrounding area; The formula for calculating the local Moran index is as follows: Where n is the total number of spatial units, y i and y j Let represent the attribute values ​​of the i-th and j-th spatial units respectively, y be the mean of the attribute values ​​of all spatial units, and w be the mean of the attribute values ​​of all spatial units. ij This represents the spatial weight value; The specific method for step S50 is as follows: S51. Identify industrial clusters by calculating the global Moran index; a high global Moran index value indicates that industries are highly concentrated in the corresponding region. S52. Low or even negative global Moran's Index values ​​indicate that industries are showing a divergent trend in the corresponding regions. S53. The local Moran index identifies agglomeration or dispersion phenomena in specific areas, discovering local hot spots and cold spots, i.e., local spatial patterns of industrial agglomeration and dispersion. S54. Based on the Moran Index results, analyze the degree of technology concentration in a specific industry; a high Moran Index value indicates that technology resources and innovation activities are highly concentrated in the corresponding region, while a low value indicates that technology resources are more dispersed. S55. The analysis results of the Moran Index provide a scientific basis for policy making and industrial planning.

Images