444853 Big Data Analysis: A Text Mining-Based Chemical Incident Case Study

Tuesday, April 12, 2016: 11:21 AM
335A (Hilton Americas - Houston)
Whapeoung Kim and Seungho Jung, Department of Environmental and Safety Engineering, Ajou University, Suwon, South Korea

Chemical incidents can be mitigated and prevented by lessons learned from past incidents. Safety Taxonomy provides a useful, consistent, structured definition of safety terms, which allows better communication and knowledge sharing. Therefore, over the incident to develop an efficient Safety Taxonomy model and understanding them is very important. This paper describe safety concepts using advanced text mining methodology to conduct a comprehensive analysis, from September 1998 to October 2015 74 cases of the US Chemical Safety and Hazard Investigation Board (CSB) report and from November 1979 to April 2015 851 cases of major accident reporting system (eMARS). Text mining also allows topic clustering and analyses of Common factors and differences of each of the data. Based on the research, developing of safety Taxonomy concepts modeling and describe the topic keywords by the analysis.

Keywords : Text mining, Topic clustering, Data mining

Extended Abstract: File Not Uploaded
See more of this Session: Big Data Analytics and Smart Manufacturing II
See more of this Group/Topical: Topical A: 2nd Big Data Analytics