Importance Degree Research of Safety Risk Management Processes of Urban Rail Transit Based on Text Mining Method

Abstract

China's urban rail transit (URT) construction is coming into the stage of rapid development under the guidance of national policies. However, the URT construction projects belong to high-risk projects and construction safety accidents occur frequently. Presently, safety risk management is in continuous development. Unfortunately, due to risk data deficiencies and lack of relationship between participants and safety risk factors, most of the research results cannot be well applied to URT projects. To overcome the limits, this paper has applied the text mining method into safety risk analysis. Through word frequency analysis and cluster analysis, 15 safety risk factors and 3 participants are identified from 156 accident reports. In addition, the accident descriptive model has been established, which is composed of indirect safety risk factors (management defects), direct safety risk factors and participants. In this model, each accident is the standardized description of the corresponding accident information. This is useful for risk data accumulation and analysis. Then the network structure analysis and risk assessment methods are utilized to make clear 63 relationships among participants, management defects and direct safety risk factors. Subsequently, the risk value of each relationship is evaluated. These safety risk information is integrated into the accident descriptive model by using accident points. Finally, ABC analysis which is a popular and effective method used to classify items into specific categories that can be managed and controlled separately is used to analyze the safety risk management's core process(A), important process(B) and general process(C) in the accident descriptive model. The research results show that the constructor should pay attention to construction coordination, safety specifications, safety measures and personnel education, the supervisor should attach importance to timely communication, the monitoring unit should pay attention to advanced forecast and dynamic control. The main research contributions are as follows: (1) A method of obtaining risk data from unstructured content has been provided; (2) The accident descriptive model could be utilized for risk data continuous accumulation; (3) The emphases of URT construction safety risk management are made clear.

Eco-innovation performance in Slovakia: Assessment based on ABC analysis of eco-innovation indicators

“…The use of ABC analysis to rank items depends on the goals a company's management needs to accomplish (Coyle 2009). According to Li et al (2018), ABC analysis is a popular and effective method used to classify items into specific categories that can be managed and controlled separately. The three predefined and ordered categories are: 1.…”

Construction delay risk taxonomy, associations and regional contexts

“…2.2.2 Word frequency analysis. Text mining including word frequency analysis is a common tool being used to identify topics described by diverse terminology in a database search (Stansfield et al, 2017;Li et al, 2018). Word frequency analysis can be used in exploring classifications implied in textual data (Zhou and Mi, 2017), as well as in taxonomy development (Pina et al, 2013).…”

Analysis of the Risk Management Process on the Development of the Public Sector Information Technology Master Plan

“…According to [8], there is no real risk or objective. The implication is that risk is not something that is waiting to be measured independently of our minds, cultures, policies or views of the world-it is inherently subjective.…”

“…The implication is that risk is not something that is waiting to be measured independently of our minds, cultures, policies or views of the world-it is inherently subjective. To protect oneself against all risks is impossible because any opportunity invariably entails risks [8].…”