Study Of Sequential Association Mining Algorithm And Application For Water Environment Security

Data mining is a new technique which gets the useful information from the vast, incomplete, noised, fuzzy and random data. Association rules analysis is one of the most important research contents. The aim of the association rules mining is to extract associations from vast data or objects. Data mining and association rules analyzing has been applying at more and more different fields, gaining improvement and development constantly.Along with the advance of information level and improvement of measure instrument, there are a mass of measure data accumulated in the database of water-environment security information of three-gorge reservoir district. It's a important focus that how to use these data to service for the security of three-gorge reservoir district. There are latent association knowledge contained in water-environment security, which are fit for us to study by using association rules analyzing methods. And water-environment security data has its particularity in the character of style. It is significative that using advanced data analysis technique to do groping research and mine latent and valuable pattern, whether in meanings of theory research or practice guidance.Aim at the questions above, a Constraint-based Sequential Pattern Mining algorithm (CSPM) is presented in this paper. The main contents and production of this paper includes three aspects:â‘ This paper first analyzes the basic theory and main method of data mining technique and Association rules analyzing. Sum up with their research actuality and direction. Then introduced the basic concept and main process of Apriori algorithm which is the classical method of assacition rules analyzing and analyze the character of this algorithm.â‘¡Analyze and discuss the questions of association rules analyzing with multidimensional data and consecutive data. Methods of defining and depicting sequence patterns have been introduced. A Constraint-based Sequential Pattern Mining algorithm and its process and pivotal questions have been presented. Besides, this algorithm introduces the concept of attribute constrain and time-limit constrain to find sequential patterns to enhances the performance of algorithm and sequential pattern's describing ability on sequential knowledge. â‘¢We use CSPM algorithm on the database of water-environment information of three-gorge reservoir district. The flow of mining and data pretreatments has been introduced. Then we mine the time sequential patterns of water-environment security information of three-gorge reservoir district by using CSPM algorithm. and analyzed the results of mining and validate the validity of the algorithm using knowledge in water-environment information of three-gorge reservoir district areas.At last, application of sequential pattern mining in early warning and decision-making for water environment security information of three-gorge reservoir district has been discussed.