Automatic Rule-based Identification of Characters and Their Relations in Fairy Tales

DOI: 10.5176/2251-2195_CSEIT17.62

Authors: Owen Noel Newton Fernando, Anupam Chattopadhyay, Lie Thomas Alison, Shelina Lusandro

Abstract: We propose a rule-driven and data-driven approach designed with the aim to automatically acquire knowledge in this paper, about the characters in fairy tales and their relationship to each other. The system has been designed in an iterative process using a dataset of stories to create a general rule that caters to the needs of a large variety of fairy tales. This paper is part of a bigger objective of converting a textual story into animation. For the conversion, there are two major steps. Firstly, the system needs to construct the storyboard, which contains information extracted from the text. Next is to create animations based on the output of the first step. We primarily focus on the first phase in this paper.

Keywords: Data Mining, Natural Language Processing, Text Analysis