Hello, I've been working on automatic extractive text summarization where the program picks the most important sentences of a text and forms its summary. What I would like to do now is more challenging. I would like to generate summaries in natural language, and not just pick important sentences. The computer should generate abstracts that are grammatically correct and content wise coherent. How do I go about this problem? What machine learning algorithms would help? What kind of features am I supposed to identify? Sorry for the bad formatting, I'm on my phone. Thanks
[link][comment]