CN-121980038-A - Knowledge graph construction method and system for energy field containing technical permeability

Abstract

The application provides a method and a system for constructing an energy domain knowledge graph with technical permeability, wherein the method comprises the steps of acquiring multi-source unstructured text data in the energy domain, constructing a standard technical node system by adopting a mode of combining semantic clustering and large language model naming, and calculating a time sequence fusion score of each technical node; and according to the time sequence fusion score, a preset permeability calculation model is adopted, and a Logistic main trend fitting algorithm is combined, so that the dynamic permeability of the technical nodes is calculated, the inflection points of the development stage are identified, and then the energy domain knowledge graph containing the technical permeability is generated. The technical scheme provided by the application can realize dynamic quantification and visualization of the technical development situation in the energy field, and provides data support for the energy technical evolution analysis.

Inventors

• ZHANG FAN
• ZHANG JINGKAI
• HU HAN

Assignees

• 国能经济技术研究院有限责任公司

Dates

Publication Date

20260505

Application Date

20251211

Claims (10)

1. 1. The method for constructing the knowledge graph in the energy field comprises the following steps of: S1, acquiring multi-source unstructured text data in the energy field, constructing a standard technical node system by adopting a mode of combining semantic clustering and large language model naming, and calculating a time sequence fusion score of each technical node; S2, according to the time sequence fusion score, a preset permeability calculation model is adopted, and a Logistic main trend fitting algorithm is combined, so that the dynamic permeability of the technical nodes is calculated, the inflection points of the development stage are identified, and then the energy domain knowledge graph containing the technical permeability is generated.
2. 2. The method according to claim 1, wherein the standard technology node architecture construction step specifically comprises: Cleaning and field standardization are carried out on the multi-source unstructured text data; merging the achievements describing the same technical subject into the same cluster by using a semantic clustering algorithm; and (5) carrying out standardized naming on the clustered technical units by using a large language model to form independent technical nodes.
3. 3. The method of claim 1, wherein the fusion score Obtained by the following formula: ; Wherein, the The time is represented by the time period of the day, The type of data source is indicated and, Representing the weight of the corresponding data source, Representation of The number of this type of achievements at the moment.
4. 4. The method according to claim 1, wherein the predetermined permeability calculation model specifically comprises: ; Wherein, the In order for the permeability to be a function of, For the current fusion score, Is the industry average value of technical nodes in the same field, For the larger of the historical peak and the predicted peak, Is an adjustable weight coefficient.
5. 5. The method of claim 4, wherein the greater of the historical peak and the predicted peak is obtained by: ; Wherein, the As a result of the history of the peaks, A predicted peak value within a preset time period in the future.
6. 6. The method of claim 1, wherein the Logistic main trend fitting algorithm specifically comprises: in a normalized permeability sequence The upper fitting S-shaped curve has the formula: ; Wherein, the Is a saturation value, and is set to be a saturation value, In order to increase the rate of the increase, Is the potential burst midpoint; when the actual permeability deviates from the fitting curve and exceeds a preset threshold, starting a sliding window to carry out smooth correction on the current time point and the subsequent time points on average.
7. 7. The method of claim 4, wherein the adjustable weight coefficients are dynamically adjusted based on a lifecycle stage in which a technology node is located.
8. 8. The method of claim 1, wherein identifying the stage of development inflection point comprises comparing the calculated dynamic permeability to a predetermined plurality of stage thresholds to determine the stage of development in which the technology node is located.
9. 9. The method of claim 1, wherein the data sources of the multi-source unstructured text data include academic paper databases, patent databases, scientific projects databases, and industry standard databases.
10. 10. The utility model provides an energy field knowledge graph construction system that contains technical permeability which characterized in that includes: the technical node construction and quantization module is used for acquiring multi-source unstructured text data in the energy field, constructing a standard technical node system by adopting a mode of combining semantic clustering with large language model naming, and calculating a time sequence fusion score of each technical node; And the permeability calculation and trend prediction module is used for calculating the dynamic permeability of the technical nodes and identifying inflection points of the development stage by adopting a preset permeability calculation model and combining a Logistic main trend fitting algorithm according to the time sequence fusion score so as to generate an energy domain knowledge graph containing the technical permeability.

Description

Knowledge graph construction method and system for energy field containing technical permeability Technical Field The document relates to the technical field of knowledge graphs in the energy field, in particular to a method and a system for constructing the knowledge graphs in the energy field, wherein the knowledge graphs comprise technical permeability. Background Driven by the dual-carbon strategy and the development of high quality, energy technology is showing the characteristics of multi-source concurrency, cycle shortening and cross-boundary fusion. In the face of increasingly academic papers, patents, projects, standards and other multi-source data, how to systematically study and judge technical trends and identify development key periods has become an important challenge for decision making in the energy field. The current knowledge-graph-based technical analysis can realize knowledge correlation, but is difficult to dynamically quantify the technical life cycle and predict the evolution of the technical life cycle. The existing method mainly has the following problems that the method is excessively dependent on expert experience, is difficult to realize automatic and large-scale analysis, is high in subjectivity and low in reproducibility, lacks a unified quantization mechanism for fusing multi-source data, cannot comprehensively evaluate the comprehensive development degree of the technology, lacks dynamic prediction capability based on time sequences, is difficult to identify key inflection points of technology penetration, and cannot effectively support prospective decision. Therefore, a knowledge graph method capable of fusing multi-source data, quantifying technical permeability and realizing dynamic prediction needs to be constructed so as to improve objectivity, comprehensiveness and prospective of research and judgment. Disclosure of Invention The invention provides a method and a system for constructing an energy domain knowledge graph with technical permeability, and aims to solve the problems. According to an embodiment of the invention, there is provided an energy domain knowledge graph construction method including technical permeability, including: S1, acquiring multi-source unstructured text data in the energy field, constructing a standard technical node system by adopting a mode of combining semantic clustering and large language model naming, and calculating a time sequence fusion score of each technical node; S2, according to the time sequence fusion score, a preset permeability calculation model is adopted, and a Logistic main trend fitting algorithm is combined, so that the dynamic permeability of the technical nodes is calculated, the inflection points of the development stage are identified, and then the energy domain knowledge graph containing the technical permeability is generated. According to an embodiment of the present invention, there is provided an energy domain knowledge graph construction system including technical permeability, including: the technical node construction and quantization module is used for acquiring multi-source unstructured text data in the energy field, constructing a standard technical node system by adopting a mode of combining semantic clustering with large language model naming, and calculating a time sequence fusion score of each technical node; And the permeability calculation and trend prediction module is used for calculating the dynamic permeability of the technical nodes and identifying inflection points of the development stage by adopting a preset permeability calculation model and combining a Logistic main trend fitting algorithm according to the time sequence fusion score so as to generate an energy domain knowledge graph containing the technical permeability. The embodiment of the invention has the following beneficial effects: Aiming at the weaknesses of strong subjectivity, single index and the like in the research and judgment of the energy technology, the invention creatively provides a quantitative evaluation system based on multi-source data fusion. By introducing the predicted peak value as a normalization standard, the early touch false judgment caused by only referencing the historical data is effectively avoided, the prospective of trend research judgment is improved, and meanwhile, the longitudinal growth of the technology is inspected and the transverse industry competitiveness is considered by the combination normalization of the double standard and the industry average value. In addition, by combining Logistic fitting and sliding window correction, the detonation and saturation inflection points of the technology can be accurately identified, and scientific and objective data support is provided for strategic decisions, fund orientation and productivity layout in the energy field. Drawings For a clearer description of one or more embodiments of the present description or of the solutions of the prior art, the drawi