Water is an asset that influences every facet of existence. Intentional or incidental contamination of water could have an enormous effect on the welfare of the living beings and on the yield of the crops. Swift detection of the contamination event can diminish their consequences and lessen possible damage. Therefore, water quality monitoring has become a foremost concern. For this purpose, we propose an effective approach. Water sensors can be positioned in the drinking and agricultural water supply network to collect data of all the water parameters like pH, free chlorine, turbidity, temperature etc. The water supply network, is designed as an interconnection network where the vertices are the sensor placement positions. By the edge-forcing property, which is a variation of vertex forcing in graph theory, we determine the edge-forcing sets (EFS) and connect them together to model the virtual nodes which form the backbone of the wireless sensor network (WSN). Sensed data is then transmitted to the EFS. Once data is obtained through a single output from the EFS, the contamination event is detected based on the calculation of the water quality index (WQI). We can predict the quality of water using the Linear Regression, which is a supervised machine learning algorithm and view it on a graphical scale. Thus, the proposed model has a higher efficiency concerning anomaly detection and detection delay involving less human intervention.