The Research And Implementation Of Association Rules Mining On Human Resource Website

Association Rule Mining (ARM) is a method which is used to process huge amount of data to discover valuable relationships among data items. Such method can find implicit, unknown, and potential valuable knowledge which user might be interested in. ARM is one of most popular research directions in the domain of data mining, which has great research value and broad application prospects.Through analyzing existing domestic human resource websites and weak points of researching on statistic data of these websites, this paper has proposed ARM to analyze statistic data of website tendency. Meanwhile, based on the fact that using association rules can effectively find out valuable relationship among data, the thesis introduces ARM into the procedure of analyzing statistic data of human resource website, and improves the "Apriori" algorithm from the aspects of lowering the number of candidate set and time of scanning database. Then, the improved "Apriori" algorithm is implemented on the test platform of website to count human resource website trend and realize corresponding functions. The major work of the author includes:1) Improving the "Apriori" algorithm which adds data pre-processing procedure and new event group to solve problems such as too-large candidate set and multiple-database-scanning, and making comparison between the original algorithm and the improved algorithm. Compared with the original algorithm, the improved "Apriori" algorithm has saved computing time and memory space which are used by the original algorithm to proceeds part of candidate sets. The improved algorithm also increases mining efficiency of association relationship rules, and solves major problems of the original "Apriori" algorithm.2) Studying human resource website model which is based on users, positions, and interview questions, which utilizes highly scalable LAMP (Linux+Apache+Mysql+PHP) framework and Discuz system.The website model is proposing Nemo module which separates the art program from the application design, and has finally realized "interview encyclopedia website".3) Realizing the human resource website trend prediction which is based on the improved "Apriori" algorithm and website application environment, and taking comparison experiment for testing algorithm’s performance when used by human resource trend prediction. The test result shows that the improved algorithm can greatly improve the efficiency of website, which could provide important reference for future website design.