An industry park industry chain breakpoint identification method and system based on a patent technology association network

Abstract

The application discloses an industrial park industrial chain breakpoint identification method and system based on a patent technology association network. The application realizes full-process automation of industrial chain breakpoint identification, does not need manual intervention, and significantly improves identification efficiency and objectivity. By fusing multi-dimensional information such as technology similarity, technology complementation and patent citation dependence, the technical dependence relationship between enterprises can be accurately described, and the accuracy and interpretability of breakpoint identification are improved. The quantitative method is used to comprehensively evaluate the technology supply degree, demand degree and substitution degree, accurately positioning the weak link of the industrial chain, and distinguishing the missing type, monopoly type and weak connection type breakpoint, providing a clear basis for chain supplement and chain strengthening decision. At the same time, data dynamic updating and result visualization output are supported, and good scalability and practicability are achieved.

Description

Technical Field

[0001] This invention relates to the fields of industrial big data analysis, knowledge engineering, graph network analysis and intelligent decision-making technology, specifically to a method and system for identifying industrial park supply chain breakpoints based on patented technology-based association networks. Background Technology

[0002] Industrial parks (including high-tech zones, economic development zones, and industrial parks) are important spatial units for enterprise clustering. The security and resilience of the industrial chain within these parks are receiving increasing attention, especially when key links are subject to external control or when supply chain bottlenecks or breaks exist. Park management departments need to quickly identify weak links and formulate strategies to strengthen and repair the chain. Current technologies for identifying supply chain breakpoints commonly include: chain analysis based on industry classification, input-output tables, or expert experience; manual screening based on enterprise lists and upstream / downstream interviews; and identification of key enterprises / industries based on simple statistics. These methods have significant shortcomings: coarse data granularity, strong subjectivity, non-reproducibility, slow updates, and difficulty in quantifying the "breakpoint index."

[0003] With the increased availability of data such as patent databases, enterprise technical documents, and park enterprise directories, it is feasible to use patent technology information to construct technology association networks and apply them to supply chain diagnosis. However, existing publicly available solutions often remain at the level of "patent quantity statistics" and "technology similarity calculation," lacking a system implementation that addresses the quantifiable modeling of technology demand-supply in the context of the industrial park supply chain, the directional construction of technology dependencies, the interpretable identification of breakpoint types (missing / monopolistic / weakly connected), and dynamic updates and visualization output. Therefore, it is necessary to propose a method and system for identifying industrial park supply chain breakpoints based on technology associations, achieving full automation from data collection, technology extraction, network construction to breakpoint index calculation and breakpoint type identification. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a method for identifying supply chain breakpoints in industrial parks based on patent technology association networks, comprising the following steps: S1. Based on the set of enterprises in the park, collect the patent set corresponding to each enterprise, and extract the technology items based on the patent set to obtain the technology item set of each enterprise; S2. Based on the set of technology items for each enterprise, construct an enterprise-technology bipartite graph and determine the technology supply intensity, as well as construct technology similarity associations, technology complementarity associations, and technology reference dependency associations between enterprises; S3. Based on the technological similarity, technological complementarity, and technological reliance dependencies among enterprises, construct a weighted directed network and identify key technology units based on the enterprise-technology bipartite graph; S4. Based on the key technology units, calculate the technology supply, technology demand, and technology substitution of the key technology units, and calculate the breakpoint index based on the technology supply, technology demand, and technology substitution to determine the breakpoint. S5. Based on the key technical units identified as breakpoints, determine the breakpoint type and output the breakpoint information.

[0005] Preferably, in step S2, the steps of constructing technology similarity associations, technology complementarity associations, and technology reference dependency associations between enterprises include: Based on each company's set of technologies, calculate the technology similarity between companies to construct technology similarity associations; The complementary supply degree among enterprises is calculated based on the park's reference technology demand vector to construct complementary technology relationships. The strength of citation dependencies is calculated based on the patent citation relationships between enterprises to construct technology citation dependency associations.

[0006] Preferably, in step S3, the step of identifying key technology units includes: constructing the supply intensity of technology items based on the enterprise-technology bipartite diagram to identify key technology units at the technology node layer; and further includes: clustering the technology item vectors to identify key technology units at the technology cluster layer.

[0007] Preferably, in step S4, the steps of calculating the technology supply, technology demand, and technology substitutability of key technology units include: The degree of technology supply is calculated based on the sum of the number of enterprises that possess key technology units or the intensity of technology supply. The degree of technology demand is calculated based on the aggregate value of the dependence intensity of key technology units on enterprises in the park. The degree of technology substitutability is calculated based on the average technological similarity among suppliers that possess key technological units.

[0008] Preferably, in step S5, the step of determining the breakpoint type includes: When the technology supply is zero, the breakpoint type is determined to be a missing breakpoint; When the number of suppliers that possess key technology units is 1 and the technology demand exceeds a predetermined threshold, the breakpoint type is determined to be a monopoly breakpoint. When the degree of technological substitution is lower than a predetermined threshold and the dependence intensity related to the key technology unit exceeds a predetermined threshold, the breakpoint type is determined to be a weak connection breakpoint.

[0009] This invention also provides an industrial park supply chain breakpoint identification system based on a patent technology association network. The system is used to implement the above method and includes: The data collection module is used to collect the patent set corresponding to each enterprise in the park, and extract the technology items from the patent set to obtain the technology item set of each enterprise. The construction module is used to build a company-technology bipartite graph based on each company's set of technology items, determine the intensity of technology supply, and build technology similarity associations, technology complementarity associations, and technology reference dependency associations between companies; The identification module is used to construct a weighted directed network based on the technological similarity, technological complementarity, and technological reference dependency relationships between enterprises, and to identify key technology units based on the enterprise-technology bipartite graph. The judgment unit is used to calculate the technology supply, technology demand and technology substitution of key technology units, and to calculate the breakpoint index based on the technology supply, technology demand and technology substitution to determine the breakpoint. The output unit is used to determine the breakpoint type based on the key technology unit that is identified as a breakpoint, and output the breakpoint information.

[0010] Preferably, the workflow of building modules includes: Based on each company's set of technologies, calculate the technology similarity between companies to construct technology similarity associations; The complementary supply degree among enterprises is calculated based on the park's reference technology demand vector to construct complementary technology relationships. The strength of citation dependencies is calculated based on the patent citation relationships between enterprises to construct technology citation dependency associations.

[0011] Preferably, the workload of the identification module includes: constructing the supply intensity of technology items based on the enterprise-technology bipartite diagram to identify key technology units at the technology node layer; and also includes: clustering technology item vectors to identify key technology units at the technology cluster layer.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention automates the entire process of identifying supply chain breakpoints without human intervention, significantly improving identification efficiency and objectivity. By integrating multi-dimensional information such as technological similarity, technological complementarity, and patent citation dependence, it accurately depicts the technological dependencies between enterprises, improving the accuracy and interpretability of breakpoint identification. Employing a quantitative approach to comprehensively assess technology supply, demand, and substitutability enables precise location of weak links in the supply chain and distinguishes between missing, monopolistic, and weakly connected breakpoints, providing a clear basis for supply chain strengthening and supplementation decisions. Furthermore, it supports dynamic data updates and visualized results output, exhibiting good scalability and practicality. Attached Figure Description

[0013] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0015] 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 embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] Example 1 This embodiment provides a method for identifying supply chain breakpoints in industrial parks based on patent technology association networks, the steps of which include: S1. Based on the set of enterprises in the park, collect the patent set corresponding to each enterprise, and extract the technology items based on the patent set to obtain the technology item set of each enterprise.

[0018] Let the set of enterprises in the park be: in, i This represents the enterprise index (i=1,2,...,N).

[0019] Collect the patent collection corresponding to each company: in, For enterprises The number of patents (or the number of valid patents retrieved). For each patent At a minimum, the following fields should be collected: title, abstract, claims (optional), IPC / CPC classification number, application date / publication date, and citations (cited / quoted).

[0020] To avoid noise, this embodiment may limit the time window. Only patents published / granted during that period will be retained.

[0021] After data collection, the data is cleaned.

[0022] (1) Text cleaning.

[0023] Perform Chinese word segmentation, stop word removal, synonym unification, word form standardization, and entity disambiguation on the patent title / abstract (or claims).

[0024] Patent The technical terms used after cleaning are: in, For patent The number of technical terms w Technical terms.

[0025] (2) Definition of technical item (technical node).

[0026] The "technical items" of this invention may be defined in any of the following ways (in parallel): Method A: Use IPC subclasses / combinations as technical items, such as using IPC four-digit or finer-grained IPC as technical items. .

[0027] Method B: Keywords / terms are treated as technical items, for The terms in the text are filtered using TF-IDF or extracted using TextRank to form a set of technical items.

[0028] Method C: Technology clusters are treated as technology items. First, technology vectors are constructed, and then clusters are formed to create technology clusters. The technology cluster is treated as a technology item. To balance interpretability and stability, the preferred representation in this embodiment is a hybrid of "IPC + keyword": in, It is a collection of all technology items, with a size of , Represents a set of IPCs. This represents a set of keywords.

[0029] (3) Enterprise technology collection.

[0030] enterprise The set of technical items is defined as: in, Indicates from patent The set of extracted technical items.

[0031] S2. Based on the set of technology items for each enterprise, construct an enterprise-technology bipartite graph and determine the technology supply intensity, as well as construct technology similarity associations, technology complementarity associations, and technology citation dependency associations between enterprises.

[0032] Construct a bipartite diagram or matrix of enterprise and technology : in, Indicates enterprise In technical items The "technical supply intensity" in the context of technology.

[0033] In this embodiment, the technical supply intensity can be optionally defined as follows (at least one): Definition 1: Number of occurrences in, This is an indicator function.

[0034] Definition 2: TF-IDF weighted in, For enterprise i's technology items k frequency, For those with technical items k The number of companies.

[0035] Definition 3: Time decay weighted (emphasizing recent capability) in, For patent The time difference from the current point in time (e.g., years). This is the attenuation coefficient.

[0036] Each enterprise is represented as a technology vector: It is subsequently used for calculations such as similarity, complementarity, and network construction.

[0037] Subsequently, this embodiment categorizes "technical association" into three types: similarity, complementarity, and citation dependency.

[0038] (1) Technological similarity association: Technology-similar related enterprises and The technical similarity is: in, To ensure that the number is a very small positive number, and to prevent division by zero, Indicates the first j The technology patent vector of a patented enterprise.

[0039] Jaccard similarity can also be used (more interpretable): when Establish similar related edges at the same time. Indicates enterprise i All the technologies they possess Indicates enterprise j All the technologies we possess.

[0040] (2) Technological complementarity: Complementarity is used to characterize potential collaborations or supply-demand relationships. Define the firm gap vector (an approximation of the demand vector): in, This serves as a reference vector for the technology requirements of the park or industry.

[0041] The construction methods include: ① Obtained from the technology profile / industry technology dictionary of the park's leading industries; ② Obtained from the mean / quantile of the technology vectors of all enterprises in the park: in, To represent quantiles, in this embodiment, we take... This indicates a "high-level reference requirement".

[0042] enterprise i For enterprises Complementary supply degree (direction): Meaning: Enterprise Technology supply and enterprises i The degree of matching of the technology gap. When At that time, establish from point to i complementary edges (representing) Can supplement i (The gap).

[0043] (3) Technology reference dependency association Establish a company i patent collection Patents exist in China The company was cited patents Then it is believed i right There is a technological dependency.

[0044] Define the number of references as: in, Patent Citing patents .

[0045] Define reference dependency strength (normalized): in, N This represents all companies that may be cited; Indicates enterprise i All cited companies involved in the patent citation relationship, m⊆N.

[0046] The above formula represents the enterprise i What percentage of the reference dependencies depend on enterprises? .when Establish directed dependency edges .

[0047] S3. Based on the technological similarity, technological complementarity, and technological reliance dependencies among enterprises, construct a weighted directed network and identify key technology units based on the enterprise-technology bipartite graph.

[0048] A weighted directed network is obtained by combining similarity, complementarity, and reference dependency relationships. Edge weights can be weighted and merged: in, ,and ; This indicates strong directional dependence (the most "industrial chain"); Indicates the direction of supply and demand complementarity; This indicates that similar technologies are used (to supplement connections, but are prone to homogenization); when Establish directed edges .

[0049] From the above formula, we can obtain the technology dependence network at the enterprise level in the park, and its direction can be interpreted as "technology dependence / upstream support".

[0050] Next, key technology units (technology nodes or technology clusters) are identified. Supply chain disruptions often occur at "key technology links." This embodiment provides two identification paths: (1) Technology node layer (based on technology items).

[0051] Supply intensity of construction technology items: in, Indicates enterprise i In the kThe intensity of technology supply in terms of technology.

[0052] This formula represents the mastery of technical items. Number of companies. Threshold. Used to filter out weak supplies that "occasionally appear".

[0053] (2) Technology cluster layer (more robust).

[0054] Clustering of technology term vectors or enterprise vectors (e.g., K-means, spectral clustering, etc.) forms a set of technology clusters: in, cM Indicates the first M A cluster of technologies.

[0055] And define the supply intensity of a company to a technology cluster: .

[0056] S4. Based on the key technology units, calculate the technology supply, technology demand, and technology substitution of the key technology units, and calculate the breakpoint index based on the technology supply, technology demand, and technology substitution to determine the breakpoint.

[0057] In this embodiment, a "breakpoint" is defined as: a point where the demand for technology is high but the supply is low, and this manifests in the network structure as high dependence, low substitutability, or fragile connectivity.

[0058] The breakpoint index consists of three parts: demand, supply, and substitutability (or resilience).

[0059] The following is in technical units (This could be a technical item) or technology cluster For example, ( ).

[0060] (1) Technology supply Supply level reflects the scale of the technology's capabilities within the industrial park. It can be defined as the number of enterprises or the sum of their strengths.

[0061] Quantity of enterprises: in, Indicates enterprise i In the technical unit z The technical supply intensity, wherein the technical unit can be a single technical item or a cluster of technologies.

[0062] Total Strength Type: For easier comparison, normalization is performed: .

[0063] (2) Degree of technical requirements Demand reflects the degree to which the park relies on the technology. The core idea is to statistically analyze the size of "enterprises or dependent edges that rely on this technology."

[0064] An implementable definition: Aggregate the dependency strengths related to technology units in the enterprise-level directed dependency network.

[0065] First define the enterprise i Technology demand weights (gap or dependency): in, For reference demand vectors, technical units The level of demand.

[0066] Redefine demand as aggregated across the entire park: And normalize: Optional alternative definitions (network-dependent, more "industry chain"): Explanation: If i The technology is lacking, while its dependency edges point to those who possess the technology. Then the dependency is contributed to the technical unit. Demand level.

[0067] (3) Technological Substitutability / Resilience Substitutability characterizes "whether there are alternative suppliers when a supplier fails". The more suppliers there are and the stronger the technological similarity and substitutability, the higher the resilience.

[0068] Assuming mastery of technology The list of suppliers: Substitutability is defined as the average similarity (or substitutability) between supplying firms: in, Indicates mastery of technology z The set of companies (the set of companies supplying the technology); if ,but .

[0069] Normalization: .

[0070] (4) Breakpoint index The breakpoint index integrates three aspects: "high demand, low supply, and low substitution." in, and .

[0071] in, The bigger it is, the more "indispensable" it becomes; The smaller the industrial park, the greater the supply shortage; The smaller the size, the less substitutable it is, and the more vulnerable it is; therefore The larger the value, the more likely it is to be a breakpoint.

[0072] Breakpoint detection rules: in, The threshold can be a fixed value (e.g., 0.7) or a quantile (e.g., top 5%). in, This means taking the first value from a set of values. q The value of a quantile.

[0073] S5. Based on the key technical units identified as breakpoints, determine the breakpoint type and output the breakpoint information.

[0074] For technical units identified as breakpoints Further, the type is given: (1) Missing breakpoints like: If the condition is missing, it indicates that no company in the park possesses the technology.

[0075] (2) Monopoly-type breakpoint like: Then it is a monopoly-type breakpoint: only one company controls it and the demand is high.

[0076] (3) Weakly connected breakpoints (structurally fragile) Define the strength of the set of dependency edges associated with this technology: If the number of supplying firms is not small (e.g., ≥2) but the substitutability is low and the dependence is strong, then it belongs to the weakly connected type: .

[0077] (4) Output explanation information To enhance interpretability, this embodiment outputs the following: ① Breakpoint technology unit name (technology item / technology cluster keyword / IPC); ② Breakpoint Index With three components; ③ Top-K companies that generate the most demand (those with the largest gap or the greatest dependence); ④ A list of suppliers (if any) and their supply intensity; ⑤ Explanation of substitution relationships (reasons for low similarity / substitutability between suppliers, such as large technological gaps).

[0078] Finally, this embodiment also provides a dynamic update and storage visualization of the method.

[0079] (5) Incremental update When a new company or patent is added, only the affected portion will be updated: ① Update the enterprise patent collection ; ② Update the mapping matrix The corresponding row; ③ Update similarity and dependency edge weights (Partial updates are possible); ④ Recalculate the affected technical units .

[0080] (6) Storage Store the results as: ① Breakpoint List (Technical Unit, Index, Type, Evidence); ② Enterprise Technology Profile (Enterprise - Technology Intensity); ③ Network edge table (directed edges, weights, source components); ④ It can also be stored in a graph database to support visualization.

[0081] The method flow of this embodiment is as follows: Figure 1 As shown.

[0082] Example 2 This embodiment also provides an industrial park supply chain breakpoint identification system based on a patent technology association network, including: a data acquisition module, used to collect the patent set corresponding to each enterprise based on the set of enterprises in the park, and extract technology items based on the patent set to obtain the technology item set of each enterprise; a construction module, used to construct an enterprise-technology bipartite graph based on the technology item set of each enterprise and determine the technology supply intensity, and construct technology similarity association, technology complementarity association and technology reference dependency association between enterprises; an identification module, used to construct a weighted directed network based on the technology similarity association, technology complementarity association and technology reference dependency association between enterprises, and identify key technology units based on the enterprise-technology bipartite graph; a judgment unit, used to calculate the technology supply degree, technology demand degree and technology substitution degree of the key technology units, and calculate the breakpoint index based on the technology supply degree, technology demand degree and technology substitution degree to determine the breakpoint; and an output unit, used to determine the breakpoint type based on the key technology units determined to be breakpoints, and output the breakpoint information.

[0083] The workflow of the construction module includes: calculating the technology similarity between enterprises based on the technology item set of each enterprise to build technology similarity associations; calculating the complementary supply degree between enterprises based on the park's reference technology demand vector to build technology complementary associations; and calculating the citation dependency strength based on the patent citation relationship between enterprises to build technology citation dependency associations.

[0084] The workload of the identification module includes: constructing the supply intensity of technology items based on the enterprise-technology bipartite diagram to identify key technology units at the technology node layer; and clustering technology item vectors to identify key technology units at the technology cluster layer.

[0085] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for identifying supply chain breakpoints in industrial parks based on patent technology association networks, characterized in that, Includes the following steps: S1. Based on the set of enterprises in the park, collect the patent set corresponding to each enterprise, and extract the technology items based on the patent set to obtain the technology item set of each enterprise; S2. Based on the set of technology items for each enterprise, construct an enterprise-technology bipartite graph and determine the technology supply intensity, as well as construct technology similarity associations, technology complementarity associations, and technology reference dependency associations between enterprises; S3. Based on the technological similarity, technological complementarity, and technological reliance dependencies among enterprises, construct a weighted directed network and identify key technology units based on the enterprise-technology bipartite graph; S4. Based on the key technology units, calculate the technology supply, technology demand, and technology substitution of the key technology units, and calculate the breakpoint index based on the technology supply, technology demand, and technology substitution to determine the breakpoint. S5. Based on the key technical units identified as breakpoints, determine the breakpoint type and output the breakpoint information.

2. The method for identifying supply chain breakpoints in industrial parks based on patent technology association networks according to claim 1, characterized in that, In step S2, the steps for constructing technology similarity associations, technology complementarity associations, and technology reference dependency associations between enterprises include: Based on each company's set of technologies, calculate the technology similarity between companies to construct technology similarity associations; The complementary supply degree among enterprises is calculated based on the park's reference technology demand vector to construct complementary technology relationships. The strength of citation dependencies is calculated based on the patent citation relationships between enterprises to construct technology citation dependency associations.

3. The method for identifying supply chain breakpoints in industrial parks based on patent technology association networks according to claim 1, characterized in that, In S3, the step of identifying key technology units includes: constructing the supply intensity of technology items based on the enterprise-technology bipartite diagram to identify key technology units at the technology node layer; and also includes: clustering technology item vectors to identify key technology units at the technology cluster layer.

4. The method for identifying supply chain breakpoints in industrial parks based on patent technology association networks according to claim 1, characterized in that, In step S4, the steps for calculating the technology supply, technology demand, and technology substitutability of key technology units include: The degree of technology supply is calculated based on the sum of the number of enterprises that possess key technology units or the intensity of technology supply. The degree of technology demand is calculated based on the aggregate value of the dependence intensity of key technology units on enterprises in the park. The degree of technology substitutability is calculated based on the average technological similarity among suppliers that possess key technological units.

5. The method for identifying supply chain breakpoints in industrial parks based on patent technology association networks according to claim 1, characterized in that, In step S5, the step of determining the breakpoint type includes: When the technology supply is zero, the breakpoint type is determined to be a missing breakpoint; When the number of suppliers that possess key technology units is 1 and the technology demand exceeds a predetermined threshold, the breakpoint type is determined to be a monopoly breakpoint. When the degree of technological substitution is lower than a predetermined threshold and the dependence intensity related to the key technology unit exceeds a predetermined threshold, the breakpoint type is determined to be a weak connection breakpoint.

6. A system for identifying supply chain breakpoints in industrial parks based on a patent technology association network, the system being used to implement the method described in any one of claims 1-5, characterized in that, include: The data collection module is used to collect the patent set corresponding to each enterprise in the park, and extract the technology items from the patent set to obtain the technology item set of each enterprise. The construction module is used to build a company-technology bipartite graph based on each company's set of technology items, determine the intensity of technology supply, and build technology similarity associations, technology complementarity associations, and technology reference dependency associations between companies; The identification module is used to construct a weighted directed network based on the technological similarity, technological complementarity, and technological reference dependency relationships between enterprises, and to identify key technology units based on the enterprise-technology bipartite graph. The judgment unit is used to calculate the technology supply, technology demand, and technology substitution of key technology units, and to calculate the breakpoint index based on the technology supply, technology demand, and technology substitution to determine the breakpoint. The output unit is used to determine the breakpoint type based on the key technology unit that is identified as a breakpoint, and output the breakpoint information.

7. The industrial park supply chain breakpoint identification system based on patent technology association network according to claim 6, characterized in that, The workflow of building modules includes: Based on each company's set of technologies, calculate the technology similarity between companies to construct technology similarity associations; The complementary supply degree among enterprises is calculated based on the park's reference technology demand vector to construct complementary technology relationships. The strength of citation dependencies is calculated based on the patent citation relationships between enterprises to construct technology citation dependency associations.

8. The industrial park supply chain breakpoint identification system based on patent technology association network according to claim 6, characterized in that, The workload of the identification module includes: constructing the supply intensity of technology items based on the enterprise-technology bipartite diagram to identify key technology units at the technology node layer; and clustering technology item vectors to identify key technology units at the technology cluster layer.

Images