.png){kind=icon}
Return
Triplet trustworthiness validation with knowledge graph reasoning
.png){kind=icon}
PRE
AI
DOI：10.1016/j.engappai.2024.109813.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Knowledge graph is widely used in intelligent analysis and graph structure applications, which uses triplets to describe the relations and facts in the real world. In the process of incorporating external triplets into existing knowledge graph, it is vital to verify the trustworthiness level of triplet knowledge for building a comprehensive knowledge graph. In this paper, we establish a model for evaluating knowledge graph trustworthiness based on a triple strategy. The model quantifies the meaning of entities and relations expressed in the knowledge graph and obtains the quantification of triple trustworthiness measurement. And it integrates the internal semantic information of triplets and the global information of the knowledge graph, including entity-level, relation-level, and graph-level trustworthiness measurement, and finally uses multi-layer perceptron fusion to obtain the final score. This paper analyzes the effectiveness of the model output trustworthiness values and conducts error detection experiments in five real-world knowledge graph datasets. Experimental results show that compared with other models, our model has achieved significant effects.
Keywords:
Multiple-hop nodes reasoning
Triplets validation
Path graph inference
Graph reasoning
Journal
IF:
8
Papers:
5.8K
Citations:
3.5W
Organization
Cited Papers
.png){kind=icon}
ShareSave
ShareSaveShareSaveShareSaveShareSave no more
no more