Preprocessing the Groundwater Quality data by LSR and QDR techniques

Abstract

Data mining is the process of identifying patterns and their relationships to solve problems through data analysis. Data mining is utilized to haul out working information from a colossal dataset of any crude information. Environmental mining is one of the wide areas to find impact on environment. Data mining encourages the usage of essential strategies and finds noteworthy information from gigantic measure of environmental information. Data preprocessing techniques are very essential in data mining, which uses various techniques to convert the raw data into a meaningful data to further research work. In this research work, Logical Similarity Replacement (LSR) and Quantity based Discrepancy Replacement (QDR) algorithms are proposed to ascertain the quality of groundwater. The numerical information are preprocessed by the statistical techniques Mean, Median methods and non-numeric information are preprocessed by the proposed LSR and QDR methods to satisfy the fragmented and conflicting information in the dataset. The conflicting and the missing information are corrected by the picked strategies for preprocessing. In the wake of applying these preprocessing systems connected in the dataset, the nature of the informational index is improved.