Evaluation of Water Security by Data Mining Techniques Based on Rough Set

As the rapid development of economy, increased population and lagged water conservancy, a series of civic ecological environment problems have arisen, for example, water shortage, water pollution, ecological deteriorate, which influences the sustainable development of economy. This paper studied on evaluation of water security. An improved classification algorithm by attribute importance is provided. Attributes are reduced by rough set theory, redundant attributes are removed and the core attributes are gained. When building the decision tree through the improved algorithm, the current node was chosen from the core attributes of the simplified decision table and decision tree splitting is according to the importance degree of attribute so as to reduce computation and gain relative simple classification rules. An example of water security evaluation is given to validate the improved algorithm. The results show that the method is effective. The research lays a foundation for further study on water security evaluation.