Research Of Association Rule Mining And Risk Management And Controlsystemfor Student Accidental Injury Factors Of Yunnan Colleges

The emergency accidental injury of college students occurs frequently and there is a tendency that the occurrence of the emergency accidental injury will increase dramatically. It really hits the sensitive nerve of colleges, students, parents, and even the education management departments. Outwardly these injury accidents are unexpected and independent. Indeed, it is the externalization of inherent risk of education sector and college management system. Academia has already paid close attention to the issue. However, the existing research perspectives are generally restricted to special cases or individual factors and lack of integrity andrelevancy.This research designed a questionnaire that contains 43 risk factors by selected Yunnan college students as samples in accordance with the situation of student accidental injury and practical management experiences in the past years. The questionnaire survey sampled 8000 students from different levels in six types of colleges. The survey collected 7243 valid questionnaires and the result showed that 12.29% of the sampled students experienced accidental injury.The research proposed the "Algorithm of Mining Frequent Item from Student Injury Data" (AMFISI) based on the improvement of Apriori frequent mining algorithm. The AMFISI optimizes the time complexity, storage space complexity and operating efficiency. It shows three association rules cluster which are high-incidence time, frequent accidents places and accident-prone groups via data mining of 2125 candidate frequent items in 125 sub-items. Then the time "code" like "midnight injury", "heartbreak moment", "grieved season" and "traffic accidents tend to occur in spring to summer alternation and autumn to winter alternation" are obtained. The sequence of places that student frequently take accidental injury such as internship and trainee places, playgrounds, dormitories, surrounding area of colleges, on the way to commute to school and so forth are detected. Furthermore, the characteristics of vulnerable groups are disclosed, for instance, "higher percentage of injured by acquaintance in student accidental injury", "negative correlation between maternal educational level and college student injury" and "no significant correlation between whether the student is the only child in the family or not and student injury" and so on.On account of the consequences of the data mining and referring the risk management theory, the research specifically presents internal control strategies about identification of college student accidental injury risk source, construction of monitoring platform, risk treatment, risk control effectiveness evaluation and improvementand so on. On this basis, the college student accidental injury risk internal control model is established.