Proposal of an Automatic Single Document Text Summarization
The past few years have witnessed a development of numerous applications of summarization in many fields such as: news, email threads, scientific articles, voicemail, broadcast news, video and meeting recording. These applications, even though they are not perfectly accurate, have already been shown to help users and to enhance other automatic applications and interfaces.
In this paper, we describe the proposed work for the automatic text summarization using a single scientific article which is based on creating an extractive summary. It is obtained by selecting the salient sentences, ranking them and finally presenting as a summary with taking in consideration the size of the summary that should be equal one on three from the original text. In purpose to prove the efficiency of our proposed summarization method, we have compared it to some existing summarization systems specially INTELLEXER, DIGEST and COPERNIC.
Keywords - Automatic Text Summarization; Abstractive Approach; Extractive Approach; Natural Language Processing, Single Document Summarization.