Text Summarization For Storytelling: Formal Document Case

https://doi.org/10.1016/j.procs.2018.08.053Get rights and content
Under a Creative Commons license
open access

Abstract

Text summarization (TS) is a significant part of the human text understanding process. A large number of summarization methods has been developed in the last decade. TS is important because human have limited cognitive abilities, which makes impossible to them to deal with a large number of text documents. There are two fundamental approaches to text summarization: extractive and abstractive one. The proposed abstractive method learns an internal language representation to generate more human-like summaries in form of comic book, paraphrasing the intention of the original text. Our idea is to automatically generate 3 images (comic strip). First the interesting concept word will be presented, secondly the tag clouds related to this concept will be visualized, and the last picture will present the attitude to this concept in documents (positive or negative). These three pictures create histories summarizing the concept representation given in the set of documents. The machine learning techniques ware used to carry out the analysis. The research analysis the Communication concerning the position of the Council published by the European Union office.

Keywords

Text Summarisation
Machine Learning
Storytelling

Cited by (0)