Search

CN-121998046-A - Knowledge graph construction method, device, equipment, storage medium and computer product

CN121998046ACN 121998046 ACN121998046 ACN 121998046ACN-121998046-A

Abstract

The application discloses a knowledge graph construction method, a device, equipment, a storage medium and a computer product, which relate to the technical field of knowledge graphs and disclose the knowledge graph construction method, wherein the knowledge graph construction method comprises the steps of constructing and obtaining a plurality of single-mode knowledge graphs according to multi-mode information to be arranged; linking each node extracted from each single-mode knowledge graph to an external knowledge graph for alignment to obtain a multi-mode synthetic graph, and carrying out information complementation on the multi-mode synthetic graph according to the external knowledge graph. By the method, the nodes extracted from the information of multiple modes are linked to the same external knowledge graph by introducing the external knowledge graph, the constructed graph is expanded and supplemented by utilizing the information in the external knowledge graph, complex and diverse data sources in the open source information field are effectively utilized, and the graph construction is expanded by utilizing the external knowledge.

Inventors

  • YU JUNHUI

Assignees

  • 北京奇虎科技有限公司

Dates

Publication Date
20260508
Application Date
20241101

Claims (10)

  1. 1. The knowledge graph construction method is characterized by comprising the following steps: Constructing and obtaining a plurality of single-mode knowledge maps according to the multi-mode information to be tidied; linking each node extracted from each single-mode knowledge graph to an external knowledge graph for alignment to obtain a multi-mode synthetic graph; And carrying out information complementation on the multi-mode synthetic map according to the external knowledge map.
  2. 2. The method of claim 1, wherein the step of constructing a plurality of single-mode knowledge-maps from the multi-mode information to be consolidated comprises: acquiring multi-modal information to be tidied, and acquiring a plurality of pieces of single-modal information according to the multi-modal information; and constructing and obtaining a plurality of single-mode knowledge maps according to the single-mode information.
  3. 3. The method of claim 2, wherein when the single-modality information is text-modality information; the step of constructing and obtaining a plurality of single-mode knowledge maps according to the single-mode information comprises the following steps: Naming the entity according to the text modal information to obtain entity list information; extracting the relation according to the entity list information to obtain entity relation list information; extracting attributes according to the entity relation list information to obtain attribute list information; And constructing a text modal knowledge graph according to the attribute list information.
  4. 4. The method of claim 2, wherein when the single mode information is picture mode information; the step of constructing and obtaining a plurality of single-mode knowledge maps according to the single-mode information comprises the following steps: determining an image recognition entity according to the picture mode information; determining co-occurrence relationship information according to the image recognition entity; And constructing a picture mode knowledge graph according to the image recognition entity and the co-occurrence relation information.
  5. 5. The method of claim 2, wherein when the single mode information is video mode information; the step of constructing and obtaining a plurality of single-mode knowledge maps according to the single-mode information comprises the following steps: determining a video identification entity according to the video modality information; Performing behavior detection and video event detection on the video identification entity to obtain video entity relation information; and constructing a video mode knowledge graph according to the video identification entity and the video entity relation information.
  6. 6. The method of claim 2, wherein when the single mode information is audio mode information; the step of constructing and obtaining a plurality of single-mode knowledge maps according to the single-mode information comprises the following steps: Converting the audio mode information to obtain identification text information; And constructing a text modal knowledge graph according to the identification text information.
  7. 7. A knowledge graph construction apparatus, characterized in that the apparatus comprises: The map construction module is used for constructing and obtaining a plurality of single-mode knowledge maps according to the multi-mode information to be tidied; The map synthesis module is used for linking each node extracted from each single-mode knowledge map to an external knowledge map for alignment to obtain a multi-mode synthesis map; and the information completion module is used for carrying out information completion on the multi-mode synthetic map according to the external knowledge map.
  8. 8. A knowledge graph construction apparatus comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the knowledge graph construction method according to any one of claims 1 to 6.
  9. 9. A storage medium, characterized in that the storage medium is a computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, realizes the steps of the knowledge graph construction method according to any one of claims 1 to 6.
  10. 10. A computer program product, characterized in that the computer program product comprises a computer program which, when executed by a processor, implements the steps of the knowledge-graph construction method according to any one of claims 1 to 6.

Description

Knowledge graph construction method, device, equipment, storage medium and computer product Technical Field The present application relates to the technical field of knowledge graphs, and in particular, to a knowledge graph construction method, apparatus, device, storage medium, and computer product. Background In the field of open source information, the source of information is complex, including information of multiple modes such as text, video, audio and tables, and the node extraction technology is adopted to construct nodes and the node relation extraction technology is adopted to construct edges between the nodes when a knowledge graph is constructed. However, the constructed atlas is very limited only by a single-mode extraction technology, and the information of each mode is not fully utilized and verified. The foregoing is provided merely for the purpose of facilitating understanding of the technical solutions of the present application and is not intended to represent an admission that the foregoing is prior art. Disclosure of Invention The application mainly aims to provide a knowledge graph construction method, a device, equipment, a storage medium and a computer product, which aim to solve the technical problems that graph information constructed by a single-mode extraction technology is limited, and information of each mode is not fully utilized and verified. In order to achieve the above object, the present application provides a knowledge graph construction method, which includes: Constructing and obtaining a plurality of single-mode knowledge maps according to the multi-mode information to be tidied; linking each node extracted from each single-mode knowledge graph to an external knowledge graph for alignment to obtain a multi-mode synthetic graph; And carrying out information complementation on the multi-mode synthetic map according to the external knowledge map. Optionally, the step of constructing a plurality of single-mode knowledge patterns according to the multi-mode information to be consolidated includes: acquiring multi-modal information to be tidied, and acquiring a plurality of pieces of single-modal information according to the multi-modal information; and constructing and obtaining a plurality of single-mode knowledge maps according to the single-mode information. Optionally, when the single-mode information is text-mode information; the step of constructing and obtaining a plurality of single-mode knowledge maps according to the single-mode information comprises the following steps: Naming the entity according to the text modal information to obtain entity list information; extracting the relation according to the entity list information to obtain entity relation list information; extracting attributes according to the entity relation list information to obtain attribute list information; And constructing a text modal knowledge graph according to the attribute list information. Optionally, when the single-mode information is picture mode information; the step of constructing and obtaining a plurality of single-mode knowledge maps according to the single-mode information comprises the following steps: determining an image recognition entity according to the picture mode information; determining co-occurrence relationship information according to the image recognition entity; And constructing a picture mode knowledge graph according to the image recognition entity and the co-occurrence relation information. Optionally, when the single-mode information is video-mode information; the step of constructing and obtaining a plurality of single-mode knowledge maps according to the single-mode information comprises the following steps: determining a video identification entity according to the video modality information; Performing behavior detection and video event detection on the video identification entity to obtain video entity relation information; and constructing a video mode knowledge graph according to the video identification entity and the video entity relation information. Optionally, when the single-mode information is audio-mode information; the step of constructing and obtaining a plurality of single-mode knowledge maps according to the single-mode information comprises the following steps: Converting the audio mode information to obtain identification text information; And constructing a text modal knowledge graph according to the identification text information. Optionally, the step of linking each node extracted from each single-mode knowledge graph to an external knowledge graph to align, and obtaining a multi-mode synthetic graph includes: and (3) linking each node obtained by extracting the text mode knowledge graph and/or the audio mode knowledge graph in the single mode knowledge graph to an external knowledge graph for alignment through a preset text processing strategy, and/or linking each node obtained by extracting the picture mode knowledge graph and/or the