We propose a topic-dependent attention model for sentiment classifification and topic extraction. Our model assumes that a global topic embedding is shared across documents and employs an attention mechanism to derive local topic embedding for words and sentences. These are subsequently incorporated in a modifified Gated Recurrent Unit (GRU) for sentiment classifification and extraction of topics bearing difffferent sentiment polarities. Those topics emerge from the words’ local topic embeddings learned by the internal attention of the GRU cells in the context of a multi-task learning framework. In this paper, we present the hierarchical architecture, the new GRU unit and the experiments conducted on users’ reviews which demonstrate classifification performance on a par with the state-of-the-art methodologies for sentiment classifification and topic coherence outperforming the current approaches for supervised topic extraction. In addition, our model is able to extract coherent aspect-sentiment clusters despite using no aspect-level annotations for training.