Citation

Fan B, Shao C, Liu Y, Li J. J. Transp. Saf. Secur. 2023; 15(10): 1057-1085.

Copyright

(Copyright © 2023, Southeastern Transportation Center, and Beijing Jiaotong University, Publisher Informa - Taylor and Francis Group)

DOI

10.1080/19439962.2022.2147613

PMID

unavailable

Abstract

Urban rail transit emergencies in China's large cities are frequent occurrences but currently, operation managers lack effective analysis tools that can help in reducing them. In this study we present a knowledge graph tool, developed using historical emergency text information from Beijing's urban rail transit fault logs from which an information model is developed enabling key information to be mined and subsequently analyzed so that interrelationships within the text can be determined. The knowledge graph tool assists urban rail transit operation managers to analyze more effectively, through knowledge query and semantic search, the relations and attributes of emergencies enabling more insight into their root causes. Compared with traditional first and second order text parsing algorithms, the extended high order parsing algorithm proposed in this paper has better performance in the extraction of both phrases and inter-phrase relations, with an extraction accuracy of more than 85%. Furthermore, compared with traditional failure mode effect analysis methods, the extended method proposed in this paper can also calculate phrase attributes and therefore provide a reference for quantitative risk calculations.

Language: en

Keywords

dependency syntactic analysis model; failure mode effect analysis; knowledge graph; prospect theory; urban rail transit emergency