Research And Application Of Weighted Association Rules In Safety Early Warning Of Drilling Operations

Association rule mining,as one of the important research contents of data mining,mainly studies the interesting and hidden rules among the large number of data items in transaction database,relational database and other information storage.The concept of weighted association rules by using the weight association rules based on improved,all rules are regarded to produce a large number of invalid rules shortcomings as equally important,compared with the traditional algorithm,weighted association rules is more conducive to solve the uneven distribution of the database project and the importance of the different problems.In this paper,the classic algorithm Apriori of association rules is studied.Through the thought of Apriori algorithm and the process of mining rules,the deficiency of Apriori algorithm is known.Then the optimization idea of algorithm is put forward.Apriori algorithm can not dig out small weight items,so we add weights to association rules.This paper focuses on the weighted association rules algorithm.Firstly,several algorithms of association rules are introduced,and the advantages and disadvantages of these algorithms are analyzed.Then the New-Apriori algorithm research,MINWAL(O)algorithm,and weighting normalization based on probability weighted association mining algorithm,and analyzes the advantages and disadvantages of these algorithms,and the lack of these algorithm optimization.Then put forward an improved weighted association rule algorithm,the algorithm of the data stored in the 0-1 matrix,improve the support speed by using matrix operations,and only need to scan the database once in the whole data mining process,reduce the storage space;keep the downward closure property of frequent itemsets,join the candidate set the pruning operation,the algorithm decreases the middle term production,enhance the running speed of the algorithm;adding interest threshold based on traditional support and confidence,improve the interesting rules production.The improved algorithm is applied to the safety warning of drilling operation,and the data of safety operation record of oil company is adopted.The hidden danger of safety early warning of drilling operation can be divided into three categories:human's violation of regulations,hidden danger of material and environmental factor.In the process of excavation,different weights of these three dangerous sources are set up to carry out experiments.According to the experimental results of the hidden association rules mining weighted association rules can be applied in the field of oil and gas drilling operation safety warning,improved weighted association rules mining based on the improved algorithm than the previous algorithm has significant efficiency,better efficiency in time and space.