18-21
Event Data Analysis Using Data Mining
Authors: - IJCT A.Arokia Marshal Roach , G.Raja Raja Cholan
Number of views: 273
There is a need to understand how industry as a whole is performing from a safety
standpoint. To date, no one can really answer this question with certainty. People do a decent job
collecting data on Events, but few take the analyses of the data past basic trending. Having the
capability to collect enormous amounts of data is a feat in and of itself; however, it begs the
question, “So what?” With the amount of resources spent to collect data, it seems logical to look
at the data under extreme scrutiny to obtain as much knowledge about the data as possible. Data
in a database is just that, data. By analyzing and understanding what is in the database yields
knowledge. Passing this knowledge on to others can improve the understanding of what went
wrong with Events from the past thereby greatly enabling the prevention of future Events.
Trending analyses do provide useful comparisons in the data, however, going beyond
comparisons by using data mining techniques can enable one to build predictive models, unveil
relationships within the data that are not necessarily intuitive, and perhaps answer the question,
“How is industry’s safety performance doing?” Marketers have successfully harnessed the
power of data mining to build predictive models to increase profit by, for example, determining
customer buying habits based on advertisement campaigns.