Identifying the Underlying Relationship Between Water Quality Parameters of the Groundwater Samples using Association and Clustering Algorithms in Coimbatore District

Water is a highly complex environmental system; itsprotection cannot be met by traditional methods. As a part of theprocess, it is mandatory to evaluate the parameters of groundwater so as to pursue suitable treatment. These days’ datamining algorithms have been developed to handle variousdata-rich environmental problems. In data mining, severaltechniques such as complex non-linear science, soft computingtechniques, clustering and association have been applied in thedomain of ground water quality assessment and evaluation inand around Coimbatore District. In this work, the statisticalcluster analysis methods and association rule mining techniqueswere used to identify the spatial distribution of different clusterof wells having similar characteristics and determine therelationship between different water quality variables. The waterquality assessment in Coimbatore was done using 13 parameters,namely NO3-, TDS, Mg2+, Ca2+, Na+, Cl-, F-, SO42-, EC, pH andHardness including location in different sites. The mainobjective of the present study is to assess the performance ofvarious clustering algorithms of WEKA and identify the mostsuitable algorithm for clustering water quality samples. K-Meanalgorithm and centroid method of Hierarchical clusteringperformed in the similar manner in clustering. In addition tothat, this study focused on identifying the water qualityparameters exceeding permissible limits that occur together(TDS, Mg2+, SO42-, EC, hardness) in the given samples usingAssociation Algorithms. The performance and efficiency ofdifferent association algorithms like Apriori and FrequentPattern Growth algorithm was evaluated by factors like support,confidence, lift, leverage and conviction values