Enterprise technical capability assessment method based on document content analysis

Abstract

The invention discloses an enterprise technical capability assessment method based on document content analysis. The method comprises the following steps: S1, acquiring multi-source technical document data of a to-be-assessed enterprise and preprocessing the multi-source technical document data; s2, carrying out semantic analysis on the preprocessed document data, extracting technical concepts, and mapping the technical concepts to a preset technical field knowledge graph to form structured representation of enterprise technical capabilities; the invention relates to the technical field of enterprise technical capability evaluation. According to the enterprise technical capability assessment method based on document content analysis, a semantic analysis model is adopted to process a standardized document set, technical concept entities are extracted, the semantic similarity between technical concepts and knowledge graph nodes is calculated, the technical concepts meeting the preset conditions are mapped to the corresponding nodes of the knowledge graph, and the technical capability assessment efficiency is improved. The enterprise exclusive technology knowledge graph is formed, key information in the enterprise technology document can be accurately captured, and more reliable technology capability representation is provided for evaluation.

Description

Technical Field

[0001] This invention relates to the field of enterprise technical capability assessment, and more specifically, to an enterprise technical capability assessment method based on document content analysis. Background Technology

[0002] With the continuous advancement of technology and the intensification of industry competition, a company's technological capabilities have become a crucial indicator for measuring its competitiveness and innovation. Traditional methods for assessing technological capabilities often rely on expert review or external scoring. These methods are not only highly subjective but also suffer from problems such as lengthy assessment times, inconsistent evaluation standards, and difficulty in quantification. Therefore, how to scientifically, objectively, and systematically assess a company's technological capabilities has become a pressing technical challenge that needs to be addressed.

[0003] Existing technology assessment methods often focus on single dimensions such as technological depth, breadth, and number of patents, neglecting the interconnectedness of technologies and the dynamic nature of technological progress. Furthermore, traditional methods often suffer from low processing efficiency and an inability to effectively extract valuable technical information when dealing with large amounts of technical documents. Therefore, there is an urgent need for an automated and comprehensive technology capability assessment method based on document content analysis, capable of overcoming the shortcomings of traditional assessment methods and offering higher assessment accuracy and flexibility. Summary of the Invention

[0004] The purpose of this invention is to provide a method for evaluating enterprise technical capabilities based on document content analysis. This method addresses the problem that existing technical evaluation methods often focus on a single dimension such as technical depth, technical breadth, and number of patents, neglecting the relevance of technologies and the dynamic nature of technological progress. Furthermore, traditional methods often suffer from low processing efficiency and an inability to effectively extract valuable technical information when dealing with a large number of technical documents.

[0005] This invention achieves the above objectives through the following technical solution: a method for assessing enterprise technical capabilities based on document content analysis, comprising the following steps: S1. Obtain multi-source technical document data of the company to be evaluated and preprocess it; S2. Perform semantic analysis on the preprocessed document data, extract technical concepts and map them to a pre-defined technical knowledge graph to form a structured representation of the enterprise's technical capabilities; S3. Use graph fusion technology to perform fusion operations on multi-source knowledge graphs, and generate enterprise technical capability characteristics by calculating the semantic relationships between nodes and edges in the graph; S4. Based on the aforementioned technical capability characteristics, construct a multi-dimensional evaluation index system, and output the enterprise's technical capability evaluation results by means of index weight allocation and comprehensive calculation.

[0006] Furthermore, step S1 involves acquiring and preprocessing multi-source technical document data of the company to be evaluated, specifically including: Obtain a multi-source technical document data set of the company to be evaluated, which contains various types of technical documents; The acquired document data is cleaned to remove redundant, invalid information and formatting symbols, while retaining the core technology-related content. The cleaned document data is standardized, the encoding format is unified, and specific terms are standardized and converted. The standardized document data is segmented and stop word removed. The domain-related dictionary is used for segmentation, and irrelevant words are filtered out to obtain a standardized document set.

[0007] Furthermore, in step S2, semantic analysis is performed on the preprocessed document data to extract technical concepts and map them to a predefined technical field knowledge graph, forming a structured representation of the enterprise's technical capabilities, specifically including: Construct a technical knowledge graph that includes technical concept nodes, inter-node connections, and types of relationships. A semantic analysis model is used to process the standardized document set, extract technical concept entities, and obtain a set of technical concepts. Semantic similarity between computing technology concepts and knowledge graph nodes; Technical concepts that meet preset conditions in terms of semantic similarity are mapped to corresponding nodes in the knowledge graph to form a unique technical knowledge graph for the enterprise, serving as a structured representation of the enterprise's technical capabilities.

[0008] Furthermore, when calculating the semantic similarity between technical concepts and knowledge graph nodes, a cosine similarity calculation method is used, which calculates the semantic similarity between the technical concept vector and the knowledge graph node vector.

[0009] Furthermore, the preset semantic similarity condition is that the semantic similarity is greater than a preset threshold, and the preset threshold can be adjusted and set according to actual evaluation needs.

[0010] Furthermore, step S3 employs graph fusion technology to fuse multi-source knowledge graphs, generating enterprise technical capability features by calculating the semantic relationships between nodes and edges in the graph. Specifically, this includes: Acquire multi-source knowledge graph data, which includes at least enterprise-specific technology knowledge graphs, industry-wide public technology graphs, and authoritative domain graphs. A semantic mapping-based graph fusion strategy is adopted to perform entity alignment on nodes in multi-source graphs and calculate node alignment confidence. The entity-aligned graph is fused to retain consistent relationships, while conflicting relationships are filtered according to preset rules to generate a fused enterprise technology knowledge graph. Calculate the strength of nodes and the semantic relationship strength of edges in the fused graph separately; Based on node strength and the semantic relationship strength of edges, a feature vector of enterprise technical capabilities is constructed as the feature of enterprise technical capabilities.

[0011] Furthermore, when calculating node alignment confidence, the alignment confidence is calculated by combining node semantic similarity and node attribute matching degree with a preset weight ratio.

[0012] Furthermore, when calculating node strength, the weights of the associated edges between the node and other nodes in the graph are summed. When calculating the semantic relationship strength of an edge, the weight of the associated edge and the weight of the corresponding relationship type are combined.

[0013] Furthermore, in step S4, a multi-dimensional evaluation index system is constructed based on the aforementioned technical capability characteristics. Through index weight allocation and comprehensive calculation, the enterprise's technical capability evaluation results are output, specifically including: Construct a multi-dimensional evaluation indicator system that includes multiple primary indicators and corresponding secondary indicators. The primary indicators should at least cover technological depth, technological breadth, technological relevance, and technological advancement. Weights are assigned to each indicator using a pre-defined weighting method, resulting in an indicator weight vector. Standardize the raw values ​​of each indicator; The comprehensive score of a company’s technological capabilities is calculated by combining the weights of the indicators with the standardized values ​​of the indicators. The company's technical capability level is determined based on its comprehensive score, and an evaluation report containing the score, level, and related analysis of technical capabilities is generated.

[0014] Furthermore, the analytic hierarchy process (AHP) is used to determine the weights of the indicators; and the min-max standardization method is used to standardize the original values ​​of the indicators.

[0015] The beneficial effects of this invention are as follows: 1. By acquiring multi-source technical document data from the company to be evaluated, covering various types of technical documents, the evaluation bias caused by a single data source is avoided, and the company's technical situation can be reflected more comprehensively.

[0016] 2. The acquired document data underwent a series of preprocessing operations, including cleaning, standardization, word segmentation, and stop word removal. This removed redundant and invalid information, standardized the encoding format, standardized specific terms, and filtered out irrelevant words, providing a high-quality data foundation for subsequent semantic analysis and improving the accuracy of the evaluation.

[0017] 3. A semantic analysis model is used to process the standardized document set, extract technical concept entities, and calculate the semantic similarity between technical concepts and knowledge graph nodes. Technical concepts that meet the preset conditions are mapped to the corresponding nodes in the knowledge graph to form an enterprise-specific technical knowledge graph. This graph can accurately capture key information in the enterprise's technical documents and represent them in a structured way, providing a more reliable representation of technical capabilities for evaluation.

[0018] 4. By adopting a semantic similarity-based calculation method and node strength analysis method, the enterprise's technical capabilities can be transformed into quantifiable technical capability characteristics, which facilitates subsequent analysis and decision support. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of the overall method of the present invention; Figure 2 This is the semantic analysis and knowledge graph mapping diagram of the present invention; Figure 3 This is a diagram illustrating the construction of the evaluation index system for this invention. Detailed Implementation

[0020] 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 a part of the embodiments of the present invention, and not all of them. 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.

[0021] Example 1: Please see Figure 1-3 This invention provides a technical solution: a method for assessing enterprise technical capabilities based on document content analysis, the method comprising: S1. Obtain multi-source technical document data of the company to be evaluated and preprocess it; Multi-source technical document data refers to technical documents of different types and from different channels. These channels may include internal enterprise systems, external public resources, etc.; preprocessing involves a series of processing operations on the acquired raw technical document data to make the data more suitable for subsequent analysis and processing. S2. Perform semantic analysis on the preprocessed document data, extract technical concepts and map them to a pre-defined technical knowledge graph to form a structured representation of the enterprise's technical capabilities; Semantic analysis involves a deep understanding of text content, extracting semantic information expressed by words, sentences, and paragraphs. Technical concepts are words or phrases with specific meanings and references in the technical field, representing key elements and knowledge points in technical activities. A pre-defined technical field knowledge graph is a knowledge representation model that graphically displays various concepts and their interrelationships in the technical field. The structured representation of enterprise technical capabilities presents enterprise technical capabilities in an organized and regular form, usually using structured data structures such as tables and graphs. By mapping the extracted technical concepts to the technical field knowledge graph, it is possible to clearly show the enterprise's layout in different technical fields, the relationships between technologies, and the hierarchical structure of technical capabilities. S3. Employ graph fusion technology to fuse multi-source knowledge graphs, and generate enterprise technical capability characteristics by calculating the semantic relationships between nodes and edges in the graph; Among them, graph fusion technology is a technical method used to integrate knowledge graphs from different sources with different structures and characteristics into a unified knowledge graph. Its purpose is to integrate information from multiple knowledge graphs, eliminate data redundancy and conflicts, and improve the completeness and accuracy of knowledge. Multi-source knowledge graphs are composed of knowledge graphs constructed from multiple different sources. The semantic relationship between nodes and edges in the knowledge graph is that nodes represent technical concepts, and edges represent semantic connections between concepts. Enterprise technical capability characteristics are attributes or indicators that reflect the essence and characteristics of enterprise technical capabilities, extracted through the analysis and calculation of the semantic relationship between nodes and edges in the fused knowledge graph. S4. Construct a multi-dimensional evaluation index system based on technical capability characteristics, and output the evaluation results of enterprise technical capabilities through index weight allocation and comprehensive calculation; The multi-dimensional evaluation index system is a set of indicators constructed from multiple different angles or dimensions to evaluate a company's technological capabilities. These dimensions may include technological innovation, technological maturity, technological applicability, and technical team capabilities. Each dimension contains several specific evaluation indicators, which together comprehensively evaluate the company's technological capabilities. The indicator weighting is based on the importance of each evaluation indicator to the company's technological capabilities, assigning corresponding weight values. The weight reflects the relative importance of the indicator in the overall evaluation; the larger the weight, the more significant the impact of the indicator on the evaluation result. The comprehensive calculation involves weighting and summing the scores of each evaluation indicator according to their corresponding weights or performing other mathematical operations to obtain a comprehensive evaluation score. This comprehensive score can more comprehensively and accurately reflect the overall level of the company's technological capabilities. The evaluation result of the company's technological capabilities is the final evaluation of the company's technological capabilities obtained through comprehensive calculation, and is usually presented in the form of quantitative scores, grades, or detailed textual descriptions.

[0022] It should be noted that during use, multi-source technical document data acquisition and preprocessing comprehensively collect enterprise technical information, providing rich materials for assessment. Preprocessing ensures data quality and facilitates subsequent analysis. Semantic analysis and knowledge graph mapping transform document content into a structured representation, clearly presenting the layout and correlation of the enterprise's technical capabilities, enhancing the intuitiveness of the assessment. Graph fusion technology integrates multi-source knowledge graphs to comprehensively mine technical capability characteristics, avoiding information bias. Based on these characteristics, a multi-dimensional assessment indicator system is constructed, and weights are allocated for comprehensive calculation, making the assessment results more scientific, objective, and comprehensive. This accurately reflects the enterprise's technical capability level, providing a strong basis for enterprises to formulate technology strategies and optimize resource allocation, thereby enhancing the enterprise's technical competitiveness.

[0023] In one embodiment, acquiring and preprocessing multi-source technical documentation data of the enterprise to be evaluated includes: Obtain multi-source technical documentation data from the company to be evaluated. , ,in This indicates the number of document types, including patents, academic papers, industry standards, etc. Indicates the number of samples for a single document type; The document data is cleaned to remove redundant text, formatting symbols, and invalid information, while retaining core technology-related content. Text standardization processing is adopted to convert document data into UTF-8 encoding format, English terms are translated uniformly, and professional abbreviations are replaced with their full names; By employing text segmentation and stop word removal, and using a domain-specific dictionary to segment documents, filtering out general stop words and domain-irrelevant terms, a standardized document set is obtained. .

[0024] This design clearly defines the document data structure, covers various document types and samples, ensures information comprehensiveness, cleans and removes redundant and invalid information, focuses on core technologies, improves data quality, standardizes text with unified encoding, translates terminology and replaces abbreviations, eliminates data discrepancies, facilitates subsequent analysis, and uses word segmentation and stop word removal with the help of a domain dictionary for accurate word segmentation and filtering of irrelevant words to obtain a standardized document set. This provides a high-quality and standardized data foundation for semantic analysis, making subsequent technical capability assessments more accurate and reliable, and reducing assessment bias caused by data issues.

[0025] In one embodiment, semantic analysis is performed on the preprocessed document data to extract technical concepts and map them to a preset technical field knowledge graph, forming a structured representation of the enterprise's technical capabilities, including: Building knowledge graphs in the technology field ,in Represents a technical concept node. Indicates the edges connecting nodes. Indicates the type of association; The BERT-BiLSTM-CRF semantic analysis model was used to analyze the standardized document set. Technical concept extraction is performed by capturing semantic features of text through a pre-trained domain language model and combining them with conditional random fields to output technical concept entities, thus obtaining a set of technical concepts. ; The semantic similarity between computational technology concepts and knowledge graph nodes is calculated using the cosine similarity formula: in For technical concept vectors, For knowledge graph node vectors; Technical concepts with semantic similarity greater than a preset threshold (set to 0.85) are mapped to corresponding knowledge graph nodes to form a company-specific technical knowledge graph. As a structured representation of a company's technological capabilities, among which , .

[0026] This design constructs a knowledge graph that clearly defines technical concepts, associations, and relationship types, providing a clear framework for mapping. The BERT-BiLSTM-CRF model, combined with a domain-pre-trained model, accurately captures semantic features to extract technical concepts, calculates semantic similarity, and sets threshold mappings. This accurately maps technical concepts to graph nodes, forming a unique technical knowledge graph for the enterprise. The structured representation intuitively presents the enterprise's technical capability layout and associations, facilitating in-depth analysis of technical strengths and weaknesses. It provides a clear basis for formulating enterprise technology strategies, enhancing the intuitiveness and practicality of the assessment.

[0027] In one embodiment, graph fusion technology is used to fuse multi-source knowledge graphs, and enterprise technical capability features are generated by calculating the semantic relationships between nodes and edges in the graph, including: Acquire multi-source knowledge graph data, including data built from internal enterprise technical documents. Industry Public Technology Map and authoritative maps of the field ; A semantic mapping-based graph fusion strategy is adopted. Nodes sharing the same technical concept in multi-source graphs are matched through entity alignment. The alignment confidence is jointly calculated using entity attributes and semantic features, as shown in the formula: in, For attribute matching degree; The aligned graphs are then fused to preserve consistent relationships across the multiple sources. Conflicting relationships are filtered using a domain expert rule base to generate the fused enterprise technology knowledge graph. ; Calculate the semantic relationship strength between nodes and edges in the graph. Node strength: in, The weight of the associated edge is determined by the frequency and credibility of the relationship. The semantic relation strength of an edge: in Weights for relation types; Relationship type weight settings: Based on the impact of technological relationships on a company's technological capabilities, relationship types are categorized into different levels and assigned corresponding weights: The relationship of "core technology support" specifically means that technology A is the core foundation of technology B: weight. =1.2, such a relationship is crucial to the stability of the technical system; A "regular association" relationship specifically refers to a collaborative application between technology A and technology B: weighting. =1.0, which is a common association in the technical system; An "indirect association" relationship specifically refers to a connection between technology A and technology B via a third-party technology: weighting. =0.6, which has a relatively weak impact on the technology system; The weight values ​​are set based on industry technical logic and expert consensus to ensure that the differences in importance of different relationship types are reasonably quantified. Based on node strength and the semantic relationship strength of edges, construct a feature vector of enterprise technical capabilities: This design acquires multi-source knowledge graphs, comprehensively integrates internal and external technical information of the enterprise, and uses a graph fusion strategy based on semantic mapping. Through entity alignment and confidence calculation, it accurately matches nodes with the same technical concept, retains consistent relationships and filters conflicting relationships, generates a fused graph, calculates the semantic relationship strength between nodes and edges, and comprehensively considers the weight of associated edges and the weight of relationship types. This accurately measures the importance of technical elements, constructs a feature vector of the enterprise's technical capabilities, comprehensively reflects the characteristics of the enterprise's technical capabilities, and provides rich and accurate feature data for subsequent evaluation, making the evaluation more scientific and comprehensive.

[0028] In one embodiment, a multi-dimensional evaluation index system is constructed based on technical capability characteristics. Through index weight allocation and comprehensive calculation, the enterprise's technical capability evaluation results are output, including: A multi-dimensional evaluation index system is constructed, comprising four primary indicators: technological depth, technological breadth, technological relevance, and technological advancement. Each primary indicator has 3-4 secondary indicators, specifically: Technical depth: average strength of core technology nodes, proportion of high-weight related edges, and degree of specialization of technical concepts; Technology breadth: number of fields covered by technology concept nodes, number of cross-domain technology associations, and proportion of emerging technology nodes; Technology relevance: density of association between core nodes and supporting nodes, degree of cross-domain technology integration, and clustering coefficient of technology relationship network; Technological advancement: percentage of patented technology nodes, number of emerging technology nodes in the past 3 years, and degree of matching with leading nodes in authoritative field maps; The analytic hierarchy process (AHP) was used to determine the indicator weights. Domain experts were invited to conduct pairwise comparisons of indicator importance to construct a judgment matrix and calculate the weight vector. in The total number of secondary indicators, and satisfying the following conditions. ; Each secondary indicator was standardized using the min-max standardization formula: in These are the original index values. The value is the standardized value; The formula for calculating the overall score of a company's technological capabilities is as follows: Based on the scores, companies' technological capabilities are divided into 5 levels: excellent: ≥0.8; Good: 0.6≤ <0.8; Medium: 0.4≤ <0.6; Generally: 0.2≤ <0.4; weak: <0.2; The output includes an assessment report with scores, ratings, and areas of strengths / weaknesses.

[0029] The solution is designed as follows: The indicator system is comprehensive and covers the core dimensions of technical capabilities: it starts from four primary indicators: technical depth, breadth, relevance, and advancement. It includes dimensions such as "core technology node strength" that reflect the vertical depth of technical cultivation, as well as indicators such as "number of cross-domain technical connections" that reflect the horizontal expansion of technology. It also captures the technical synergy and timeliness through "technical relationship network clustering coefficient" and "frontier node matching degree", avoiding the one-sidedness of traditional single-dimensional evaluation and presenting a complete picture of the company's technical capabilities.

[0030] The weighting and standardization methods are scientific, ensuring the objectivity of the assessment: The analytic hierarchy process (AHP) is used to determine weights, and pairwise comparisons of indicator importance are made based on domain expert experience, avoiding subjective assumptions; min-max standardization eliminates differences in the dimensions of different indicators, ensuring a reasonable weighting of each indicator in the overall calculation. The combination of these two methods makes the overall score calculation logically rigorous, reduces human error, and makes the assessment results more credible.

[0031] The grading is clear and the application scenarios of the results are well-defined: the five capability levels (from excellent to weak) correspond to clear score ranges, allowing companies to quickly identify their own technical level; the assessment report simultaneously outputs the areas of strength and weakness, which can directly guide practice without additional interpretation, providing precise direction for companies to adjust R&D investment and make up for technical shortcomings.

[0032] The calculation logic is traceable and the methods are reusable and scalable: there are clear rules for everything from indicator definition and weight calculation to comprehensive score formula, and it does not rely on specific industry-specific data. It can be applied to enterprise technology assessment in different industries such as new energy and electronic information simply by replacing the technical documents and knowledge graphs of the corresponding fields, and has strong versatility and scalability.

[0033] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0034] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for assessing enterprise technical capabilities based on document content analysis, characterized in that, Includes the following steps: S1. Obtain multi-source technical document data of the company to be evaluated and preprocess it; S2. Perform semantic analysis on the preprocessed document data, extract technical concepts and map them to a pre-defined technical knowledge graph to form a structured representation of the enterprise's technical capabilities; S3. Use graph fusion technology to perform fusion operations on multi-source knowledge graphs, and generate enterprise technical capability characteristics by calculating the semantic relationships between nodes and edges in the graph; S4. Based on the aforementioned technical capability characteristics, construct a multi-dimensional evaluation index system, and output the enterprise's technical capability evaluation results by means of index weight allocation and comprehensive calculation.

2. The enterprise technical capability assessment method based on document content analysis according to claim 1, characterized in that, Step S1 involves acquiring and preprocessing multi-source technical documentation data of the company to be evaluated, specifically including: Obtain a multi-source technical document data set of the company to be evaluated, which contains various types of technical documents; The acquired document data is cleaned to remove redundant, invalid information and formatting symbols, while retaining the core technology-related content. The cleaned document data is standardized, the encoding format is unified, and specific terms are standardized and converted. The standardized document data is segmented and stop word removed. The domain-related dictionary is used for segmentation, and irrelevant words are filtered out to obtain a standardized document set.

3. The enterprise technical capability assessment method based on document content analysis according to claim 1, characterized in that, Step S2 involves semantic analysis of the preprocessed document data, extracting technical concepts and mapping them to a predefined technical field knowledge graph to form a structured representation of the enterprise's technical capabilities. Specifically, this includes: Construct a technical knowledge graph that includes technical concept nodes, inter-node connections, and types of relationships. A semantic analysis model is used to process the standardized document set, extract technical concept entities, and obtain a set of technical concepts. Semantic similarity between computing technology concepts and knowledge graph nodes; Technical concepts that meet preset conditions in terms of semantic similarity are mapped to corresponding nodes in the knowledge graph to form a unique technical knowledge graph for the enterprise, serving as a structured representation of the enterprise's technical capabilities.

4. The enterprise technical capability assessment method based on document content analysis according to claim 3, characterized in that: When calculating the semantic similarity between a technology concept and a knowledge graph node, a cosine similarity calculation method is used, which calculates the semantic similarity between the technology concept vector and the knowledge graph node vector.

5. The enterprise technical capability assessment method based on document content analysis according to claim 3, characterized in that: The preset semantic similarity condition is that the semantic similarity is greater than a preset threshold, which can be adjusted according to actual evaluation needs.

6. The enterprise technical capability assessment method based on document content analysis according to claim 1, characterized in that, Step S3 employs graph fusion technology to fuse multi-source knowledge graphs, generating enterprise technical capability features by calculating the semantic relationships between nodes and edges in the graph. Specifically, this includes: Acquire multi-source knowledge graph data, which includes at least enterprise-specific technology knowledge graphs, industry-wide public technology graphs, and authoritative domain graphs. A semantic mapping-based graph fusion strategy is adopted to perform entity alignment on nodes in multi-source graphs and calculate node alignment confidence. The entity-aligned graph is fused to retain consistent relationships, while conflicting relationships are filtered according to preset rules to generate a fused enterprise technology knowledge graph. Calculate the strength of nodes and the semantic relationship strength of edges in the fused graph separately; Based on node strength and the semantic relationship strength of edges, a feature vector of enterprise technical capabilities is constructed as the feature of enterprise technical capabilities.

7. The enterprise technical capability assessment method based on document content analysis according to claim 6, characterized in that: When calculating node alignment confidence, the alignment confidence is obtained by combining node semantic similarity and node attribute matching degree with a preset weight ratio.

8. The enterprise technical capability assessment method based on document content analysis according to claim 6, characterized in that: When calculating node strength, the weights of the edges connecting the node to other nodes in the graph are summed. When calculating the semantic relationship strength of an edge, the weight of the associated edge and the weight of the corresponding relationship type are combined.

9. The enterprise technical capability assessment method based on document content analysis according to claim 1, characterized in that, Step S4 involves constructing a multi-dimensional evaluation index system based on the aforementioned technical capability characteristics. Through index weight allocation and comprehensive calculation, the enterprise's technical capability evaluation results are output, specifically including: Construct a multi-dimensional evaluation indicator system that includes multiple primary indicators and corresponding secondary indicators. The primary indicators should at least cover technological depth, technological breadth, technological relevance, and technological advancement. Weights are assigned to each indicator using a pre-defined weighting method, resulting in an indicator weight vector. Standardize the raw values ​​of each indicator; The comprehensive score of a company’s technological capabilities is calculated by combining the weights of the indicators with the standardized values ​​of the indicators. The company's technical capability level is determined based on its comprehensive score, and an evaluation report is generated that includes the score, level, and related analysis of technical capabilities.

10. The enterprise technical capability assessment method based on document content analysis according to claim 9, characterized in that: The weights of the indicators were determined using the analytic hierarchy process (AHP); the original values ​​of the indicators were standardized using min-max standardization.

Images