Framework for Story Generation using RNN
In story generation the problem that occurs mostly is selection of events that satisfies the criteria of story creation. It is knowledge-intensive. Conventional story generation frameworks depend on prior well-defined domain paradigms about fictional-world, including characters, scenes and events which should be included. In this paper, we prospect a public-domain story generation framework that constructs stories while input given is storyline. A framework is proposed that first frames a story based on storyline with help of hierarchical neural networks. A technique is suggested for automated generation of story through which the framework decomposes a storyline into the generation of sequential events and the generation of spoken-written language sentences from events. The algorithm is used to extract stories from text and dataset further to compute optimal story. The framework produces an example of computational artistry of resourcefulness that can be comparable against generous writing.