This type of summarization can produce output summaries containing words or phrases that are not in the original text but preserving the original intent of the input document. This can result in novel phrases and thereby natural summaries. We will first look at an overview of abstractive text summarization using deep learning approaches. In text summarization, as the input and output, are both sequences of text, the deep learning model commonly used in practice is the sequence-to-sequence model. We will briefly describe this approach to text summarization next.