Due to the tremendous increase in the activity of social media, there has been a considerable drift in the propagation of rumours causing damage to several organizations or personal. Since last few years, a significant exploration has been done due to hazardous effects of rumours on society. In this paper, we propose a deep neural network (RUEvaL20) for effective rumour detection. Experiments have been conducted with real-world dataset: PHEME. We have achieved remarkable results, and our proposed model outperformed with most of the existing state-of-the-art methods.