A pragmatic approach towards automatic story generation and reasoning

Abstract

Automatic story generation is one of the most interesting applications of natural language generation in Artificial intelligence that has received significant attention, especially during the past few years. The main reasons are the wide range of commercial and entertainment applications and the availability of feasible technologies after thirty years of research. In this work, two approaches are proposed for theme conception in order to decide the flow of the story. The static conception method allows the user to tailor the list of events to conceive the theme for story generation. The random conception method allows the system to collect and organize the list of events to conceive the theme for story generation. Both the approaches help to lead the flow of the story. The existing story generation algorithm uses a built-in theme for story construction, as well as a canned sequence of text to represent the story. In this thesis, five different story generation systems are developed using both the theme conception procedures. In this work, the existing Purdom s sentence generation algorithm is revised based on ontology to construct the sentences. By applying cognitive intelligence, human beings can analyze the stories easily whereas it is difficult for the system to perform. Hence, a new systematic approach for reasoning the stories has been introduced to overcome the problems of automatic story generation. This approach utilized the benefits of ontology and first order logic (FOL). Automatic story generation is one of the essential features in the entertainment world to generate automatic script writing, report generation, etc. Hence, it requires a strong force to generate and reason the stories, in order to improve the quality of the same. In this work, five different systems for story generation are developed, the existing language generation algorithm is revised, and a new system for the semantic reasoner is constructed to cater to the need for automatic story generation. newline newline newline