Research And Application On Data Mining Of Association Rules In Process Industry

When the time comes into the information age of 21st century, modern science and technology make great progress and the scope of enterprises expand increasingly. More and more enterpises begin to use computer to manage. The high tech's use bring people great convenient for their daily operations, at the same time, bring them huge amounts of data, while existing tools and methods can only implement the data storage, retrieval and the database transaction processing, they can't find interesting patterns hidden in large data. All these stated above result in the emergence of the phenomenon of "data blast but knowledge deficiency". So, there is an imminent need for turning such data into useful information and knowledge. KDD (Knowledge discovery in database) and data mining emerges as the times require. In fact, data mining can be viewed as a result of natural evolution of information technology.KDD and Data mining has attracted a great deal of attention in the information industry in recent years. The research on it involved fields of statistics, Artifical Intelligence, machine learning and neural network etc.In this paper, we put emphasis on the research of one of the most important method, mining of association rules, make deep research on the mining algorithm, and give detailed descriptions on the problem ofassociation rules' mining. Then we make deep analysis on the implementation method, data structure and use of memory of the Apriori algorithm and AprioriTid algorithm, and propose the improved method according to the several problems in mining algorithm, which influence the mining efficiency. The new algorithm's use greatly reduces the record number read in every time, and improves the efficiency of mining.The mining of association rules comes of the basket transaction data, but it is not only used in the transaction database. In this paper, we attempt to apply the mining of association rules into process industry, and according to the characteristic of mining in process industry, mine the general association rules and trend association rules in industry. Based on all of these, we use large amount of simulating data and two actual problem of yeild process in chemical industry to validate the algorithm improved, then make data analysis on the rules generated.In the end, integrating with the scour product quality management information system of LuoYang MingHua Group, we designed the visualized throughput data mining system (VTPDM), which is mainly used to help the decision maker to find the factors that influence the throughput.