Nuclear power information intelligent retrieval method and device and related storage medium and equipment

Abstract

The invention belongs to the technical field of nuclear power related data mining, and particularly relates to a nuclear power information intelligent retrieval method and device, a related storage medium and equipment. Comprising the following steps: S1, obtaining nuclear power corpora, and constructing a nuclear power knowledge graph; s2, detecting the relationship of each piece of nuclear power information in the nuclear power database in the nuclear power knowledge graph to obtain a first entity relationship set list corresponding to each piece of nuclear power information; s3, acquiring an input search text, and detecting the relationship of the search text in the nuclear power knowledge graph to obtain a second entity relationship set list; s4, calculating intersections of the entity relationship sets in the first entity relationship set list and the entity relationship sets in the second entity relationship set list to obtain a plurality of intersections; s5, calculating a weight coefficient between the search text and each piece of nuclear power information according to the plurality of intersections; and S6, arranging each piece of nuclear power information according to the weight coefficient to obtain a nuclear power information retrieval result. The method has the beneficial effects that the retrieval efficiency, recall ratio and precision ratio of nuclear power information are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of nuclear power related data mining, and specifically relates to a method and device for intelligent retrieval of nuclear power information and related storage media and equipment. Background Art

[0002] A knowledge graph is a knowledge base used for knowledge management. It describes various entities or concepts and their relationships in the real world, forming a semantic network graph in which nodes represent entities or concepts and edges consist of attributes or relationships. Knowledge graphs facilitate the collection, organization and extraction of knowledge in specific fields, and can improve the recall and precision of various search engines and search systems.

[0003] In some special fields such as nuclear power, commonly used search information such as keywords are less common and some search information is not well understood by ordinary users due to the high level of technical expertise. Therefore, it is necessary to build a knowledge graph directly based on common and easily available nuclear power information, so as to improve the recall and precision of nuclear power information. Summary of the invention

[0004] The purpose of the present invention is to provide a method, device and related storage medium and equipment for intelligent retrieval of nuclear power information, which can quickly retrieve nuclear power information with a larger weight, that is, more important, improve the retrieval efficiency of nuclear power information, and also improve the completeness and precision of nuclear power information.

[0005] The technical solution of the present invention is as follows: A method for intelligent retrieval of nuclear power information comprises the following steps:

[0006] S1: Obtain nuclear power corpus and build a nuclear power knowledge graph;

[0007] S2: Detect the relationship between each piece of nuclear power information in the nuclear power database and the nuclear power knowledge graph, and obtain a first entity relationship set list corresponding to each piece of nuclear power information;

[0008] S3: Obtain the input search text, detect the relationship of the search text in the nuclear power knowledge graph, and obtain a second entity relationship set list;

[0009] S4: Calculate the intersection of the entity relationship set in the first entity relationship set list and the entity relationship set in the second entity relationship set list to obtain multiple intersections;

[0010] S5: Calculate the weight coefficient between the search text and each piece of nuclear power information according to the multiple intersections;

[0011] S6: Arrange each piece of nuclear power information according to the weight coefficient to obtain the nuclear power information retrieval result.

[0012] Further, step S1 includes:

[0013] S11: Obtain nuclear power text as nuclear power corpus;

[0014] S12: preprocess the nuclear power text to obtain an entity word library;

[0015] S13: performing vectorization processing on the nuclear power text to obtain a text vector;

[0016] S14: Using the document vector as a training set, training a preset word vector model to obtain a trained word vector model;

[0017] S15: Traverse each entity in the entity vocabulary, calculate the entities with similar relationships with it through the trained word vector model, form a triple structure of "entity-approximate value-entity", and obtain the nuclear power knowledge graph.

[0018] Further, step S12 includes:

[0019] S121: performing word segmentation processing on the nuclear power text to obtain the first word segmentation;

[0020] S122: Tagging the first participle for part of speech;

[0021] S123: extracting the first participle according to the preset part-of-speech requirement to obtain the second participle;

[0022] S124: filtering the second participle to obtain a filtered second participle;

[0023] S125: The filtered second word segment is used as an entity to form an entity word library.

[0024] Further, step S2 includes:

[0025] S21: performing word segmentation processing on each piece of nuclear power information in the nuclear power database to obtain a third word segmentation;

[0026] S22: perform part-of-speech tagging on the third participle in sequence: extract and filter to obtain the fourth participle;

[0027] S23: Detect the relationship of the fourth participle in the nuclear power knowledge graph, obtain the entity relationship set corresponding to the fourth participle, and then obtain the first entity relationship set list.

[0028] Further, step S3 includes:

[0029] S31: Perform word segmentation processing on the search text to obtain the fifth word segmentation;

[0030] S32: performing part-of-speech tagging on the fifth participle in sequence: extracting and filtering to obtain the sixth participle;

[0031] S33: Detect the relationship of the sixth participle in the nuclear power knowledge graph, obtain the entity relationship set corresponding to the sixth participle, and then obtain the second entity relationship set list.

[0032] Further, step S5 includes:

[0033] S51: Detecting the maximum approximate value in each intersection according to the multiple intersections;

[0034] S52: Calculate the weight coefficient between the search text and each piece of nuclear power information according to the largest approximate value in each intersection.

[0035] Further, step S52 includes:

[0036] The weight coefficient between the search text and each piece of nuclear power information is calculated by formula (1):

[0037] W=e[(Rn- R1)] / 2+[(Rn+ R1) *(R1+R2+R3+…+Rn)] / n (1)

[0038] Among them, W represents the weight coefficient, e represents the natural constant, R1, R2, ..., Rn represent the largest approximate values ​​in the 1st to nth intersections respectively, and n is the number of intersections.

[0039] The nuclear power information intelligent retrieval device includes:

[0040] Construction module, used to obtain nuclear power corpus and build nuclear power knowledge graph;

[0041] The first detection module is used to detect the relationship between each piece of nuclear power information in the nuclear power database and the nuclear power knowledge graph, and obtain a first entity relationship set list corresponding to each piece of nuclear power information;

[0042] The first detection module is used to obtain the input search text, detect the relationship of the search text in the nuclear power knowledge graph, and obtain a second entity relationship set list;

[0043] A first calculation module, used to calculate the intersection of the entity relationship set in the first entity relationship set list and the entity relationship set in the second entity relationship set list to obtain multiple intersections;

[0044] A second calculation module is used to calculate the weight coefficient between the search text and each piece of nuclear power information according to the multiple intersections;

[0045] The arrangement module is used to arrange each piece of nuclear power information according to the weight coefficient to obtain the nuclear power information retrieval result.

[0046] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements any of the aforementioned methods for intelligent retrieval of nuclear power information.

[0047] A computer device comprises a memory, a processor and a computer program stored in the memory, wherein the computer program, when executed by the processor, implements any of the aforementioned methods for intelligent retrieval of nuclear power information.

[0048] The beneficial effect of the present invention lies in: calculating the weight coefficient between the search text and each piece of nuclear power information respectively through a specific empirical formula, and then arranging each piece of nuclear power information according to the weight coefficient to obtain the nuclear power information retrieval result, so that the nuclear power information intelligent retrieval method provided by the present invention can quickly retrieve nuclear power information with a larger weight, that is, more important, thereby improving the retrieval efficiency of nuclear power information and also improving the recall rate and precision rate of nuclear power information. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 The figure is a flow chart of a method for intelligent retrieval of nuclear power information according to an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0051] The present invention provides a method for intelligent retrieval of nuclear power information, comprising the following steps:

[0052] S1: Obtain nuclear power corpus and build a nuclear power knowledge graph;

[0053] Step S1 includes:

[0054] S11: Obtain nuclear power text as nuclear power corpus;

[0055] S12: preprocess the nuclear power text to obtain an entity word library;

[0056] Step S12 includes:

[0057] S121: performing word segmentation processing on the nuclear power text to obtain the first word segmentation;

[0058] S122: Tagging the first participle for part of speech;

[0059] S123: extracting the first participle according to the preset part-of-speech requirement to obtain the second participle;

[0060] S124, filtering the second participle to obtain a filtered second participle;

[0061] S125: The filtered second word segment is used as an entity to form an entity word library.

[0062] S13: performing vectorization processing on the nuclear power text to obtain a text vector;

[0063] S14: Using the document vector as a training set, training a preset word vector model to obtain a trained word vector model;

[0064] S15: Traverse each entity in the entity vocabulary, calculate the entities with similar relationships with it through the trained word vector model, form a triple structure of "entity-approximate value-entity", and obtain the nuclear power knowledge graph.

[0065] S2: Detect the relationship between each piece of nuclear power information in the nuclear power database and the nuclear power knowledge graph, and obtain a first entity relationship set list corresponding to each piece of nuclear power information;

[0066] Step S2 includes:

[0067] S21: performing word segmentation processing on each piece of nuclear power information in the nuclear power database to obtain a third word segmentation;

[0068] S22: performing part-of-speech tagging, extraction, and filtering on the third participle in sequence to obtain a fourth participle;

[0069] S23: Detect the relationship of the fourth participle in the nuclear power knowledge graph, obtain the entity relationship set corresponding to the fourth participle, and then obtain the first entity relationship set list.

[0070] S3: Obtain the input search text, detect the relationship of the search text in the nuclear power knowledge graph, and obtain a second entity relationship set list;

[0071] Step S3 includes:

[0072] S31: Perform word segmentation processing on the search text to obtain the fifth word segmentation;

[0073] S32: performing part-of-speech tagging, extraction, and filtering on the fifth participle in sequence to obtain the sixth participle;

[0074] S33: Detect the relationship of the sixth participle in the nuclear power knowledge graph, obtain the entity relationship set corresponding to the sixth participle, and then obtain the second entity relationship set list.

[0075] S4: Calculate the intersection of the entity relationship set in the first entity relationship set list and the entity relationship set in the second entity relationship set list to obtain multiple intersections;

[0076] S5: Calculate the weight coefficient between the search text and each piece of nuclear power information according to the multiple intersections;

[0077] Step S5 includes:

[0078] S51: Detecting the maximum approximate value in each intersection according to the multiple intersections;

[0079] S52: Calculate the weight coefficient between the search text and each piece of nuclear power information according to the largest approximate value in each intersection.

[0080] Step S52 includes:

[0081] The weight coefficient between the search text and each piece of nuclear power information is calculated by formula (1):

[0082] W=e[(Rn- R1)] / 2+[(Rn+ R1) *(R1+R2+R3+…+Rn)] / n (1)

[0083] Among them, W represents the weight coefficient, e represents the natural constant, R1, R2, ..., Rn represent the largest approximate values ​​in the 1st to nth intersections respectively, and n is the number of intersections.

[0084] S6: Arrange each piece of nuclear power information according to the weight coefficient to obtain the nuclear power information retrieval result.

[0085] The present invention also provides a nuclear power information intelligent retrieval device, which comprises:

[0086] Construction module, used to obtain nuclear power corpus and build nuclear power knowledge graph;

[0087] The first detection module is used to detect the relationship between each piece of nuclear power information in the nuclear power database and the nuclear power knowledge graph, and obtain a first entity relationship set list corresponding to each piece of nuclear power information;

[0088] The first detection module is used to obtain the input search text, detect the relationship of the search text in the nuclear power knowledge graph, and obtain a second entity relationship set list;

[0089] A first calculation module, used to calculate the intersection of the entity relationship set in the first entity relationship set list and the entity relationship set in the second entity relationship set list to obtain multiple intersections;

[0090] A second calculation module is used to calculate the weight coefficient between the search text and each piece of nuclear power information according to the multiple intersections;

[0091] The arrangement module is used to arrange each piece of nuclear power information according to the weight coefficient to obtain the nuclear power information retrieval result.

[0092] The present invention further provides a computer-readable storage medium storing a computer program, which implements any of the aforementioned methods for intelligent retrieval of nuclear power information when executed by a processor.

[0093] The present invention also provides a computer device, which includes a memory, a processor and a computer program stored in the memory, and when the computer program is executed by the processor, it implements any of the aforementioned nuclear power information intelligent retrieval methods.

Claims

1. A method for intelligent retrieval of nuclear power information, characterized in that: The following steps are involved: S1: Obtain nuclear power corpus and build a nuclear power knowledge graph; S2: Detect the relationship between each piece of nuclear power information in the nuclear power database and the nuclear power knowledge graph, and obtain a first entity relationship set list corresponding to each piece of nuclear power information; S3: Obtain the input search text, detect the relationship of the search text in the nuclear power knowledge graph, and obtain a second entity relationship set list; S4: Calculate the intersection of the entity relationship set in the first entity relationship set list and the entity relationship set in the second entity relationship set list to obtain multiple intersections; S5: Calculate the weight coefficient between the search text and each piece of nuclear power information according to the multiple intersections; S6: Arrange each piece of nuclear power information according to the weight coefficient to obtain the nuclear power information retrieval result.

2. The method for intelligent retrieval of nuclear power information according to claim 1, characterized in that: The step S1 comprises: S11: Obtain nuclear power text as nuclear power corpus; S12: preprocess the nuclear power text to obtain an entity word library; S13: performing vectorization processing on the nuclear power text to obtain a text vector; S14: Using the document vector as a training set, training a preset word vector model to obtain a trained word vector model; S15: Traverse each entity in the entity vocabulary, calculate the entities with similar relationships with it through the trained word vector model, form a triple structure of "entity-approximate value-entity", and obtain the nuclear power knowledge graph.

3. The method for intelligent retrieval of nuclear power information according to claim 2, characterized in that: The step S12 comprises: S121: performing word segmentation processing on the nuclear power text to obtain the first word segmentation; S122: Tagging the first participle for part of speech; S123: extracting the first participle according to the preset part-of-speech requirement to obtain the second participle; S124: filtering the second participle to obtain a filtered second participle; S125: The filtered second word segment is used as an entity to form an entity word library.

4. The method for intelligent retrieval of nuclear power information according to claim 1, characterized in that: The step S2 comprises: S21: performing word segmentation processing on each piece of nuclear power information in the nuclear power database to obtain a third word segmentation; S22: perform part-of-speech tagging on the third participle in sequence: extract and filter to obtain the fourth participle; S23: Detect the relationship of the fourth participle in the nuclear power knowledge graph, obtain the entity relationship set corresponding to the fourth participle, and then obtain the first entity relationship set list.

5. The method for intelligent retrieval of nuclear power information according to claim 1, characterized in that: The step S3 comprises: S31: Perform word segmentation processing on the search text to obtain the fifth word segmentation; S32: performing part-of-speech tagging on the fifth participle in sequence: extracting and filtering to obtain the sixth participle; S33: Detect the relationship of the sixth participle in the nuclear power knowledge graph, obtain the entity relationship set corresponding to the sixth participle, and then obtain the second entity relationship set list.

6. The method for intelligent retrieval of nuclear power information according to claim 1, characterized in that: The step S5 comprises: S51: Detecting the maximum approximate value in each intersection according to the multiple intersections; S52: Calculate the weight coefficient between the search text and each piece of nuclear power information according to the largest approximate value in each intersection.

7. The method for intelligent retrieval of nuclear power information according to claim 6, characterized in that: The step S52 comprises: The weight coefficient between the search text and each piece of nuclear power information is calculated by formula (1): W=e[(Rn- R1)] / 2+[(Rn+ R1) *(R1+R2+R3+…+Rn)] / n (1) Among them, W represents the weight coefficient, e represents the natural constant, R1:R2:…:Rn represent the largest approximate values ​​in the 1st to nth intersections respectively, and n is the number of intersections.

8. The intelligent retrieval device for nuclear power information is characterized in that: include: Construction module, used to obtain nuclear power corpus and build nuclear power knowledge graph; The first detection module is used to detect the relationship between each piece of nuclear power information in the nuclear power database and the nuclear power knowledge graph, and obtain a first entity relationship set list corresponding to each piece of nuclear power information; The first detection module is used to obtain the input search text, detect the relationship of the search text in the nuclear power knowledge graph, and obtain a second entity relationship set list; A first calculation module, used to calculate the intersection of the entity relationship set in the first entity relationship set list and the entity relationship set in the second entity relationship set list to obtain multiple intersections; The second calculation module is used to calculate the weight coefficient between the search text and each piece of nuclear power information according to the multiple intersections; The arrangement module is used to arrange each piece of nuclear power information according to the weight coefficient to obtain the nuclear power information retrieval result.

9. A computer-readable storage medium, characterized in that: It stores a computer program, which, when executed by a processor, implements the intelligent retrieval method for nuclear power information of any one of the preceding claims 1-8.

10. A computer device comprising a memory: a processor and a computer program stored in the memory, wherein the computer program, when executed by the processor, implements the intelligent retrieval method for nuclear power information of any one of claims 1 to 8.

Images