An Abstractive Text Summarization Using Recurrent Neural Network

Dipanwita Debnath, Partha Pakray, Ranjita Das, Alexander Gelbukh

Producción científica: Capítulo del libro/informe/acta de congresoContribución a la conferenciarevisión exhaustiva

Resumen

With the accelerated advancement of technology and massive content surging over the Internet, it has become an arduous task to abstract the information efficiently. However, automatic text summarization provides an acceptable means for fast procurement of such information in the form of a summary through compression and refinement. Abstractive text summarization, in particular, builds an internal semantic representation of the text and uses natural language generation techniques to create summaries closer to human-generated summaries. This paper uses Long Short Term Memory (LSTM) based Recurrent Neural Network to generate comprehensive abstractive summaries. To train LSTM based model requires a corpus having a significant number of instances containing parallel running article and summary pairs. For this purpose, we have used various news corpus, namely DUC 2003, DUC 2004 and Gigaword corpus, after eliminating the noise and other irrelevant data. Experiments and analyses of this work are performed on a subset of these whole corpora and evaluated using ROUGE evaluation. The experimental result verifies the accuracy and validity of the proposed system.

Idioma originalInglés
Título de la publicación alojadaComputational Linguistics and Intelligent Text Processing - 19th International Conference, CICLing 2018, Revised Selected Papers
EditoresAlexander Gelbukh
EditorialSpringer Science and Business Media Deutschland GmbH
Páginas364-378
Número de páginas15
ISBN (versión impresa)9783031238031
DOI
EstadoPublicada - 2023
Evento19th International Conference on Computational Linguistics and Intelligent Text Processing, CICLing 2018 - Hanoi, Vietnam
Duración: 18 mar. 201824 mar. 2018

Serie de la publicación

NombreLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen13397 LNCS
ISSN (versión impresa)0302-9743
ISSN (versión digital)1611-3349

Conferencia

Conferencia19th International Conference on Computational Linguistics and Intelligent Text Processing, CICLing 2018
País/TerritorioVietnam
CiudadHanoi
Período18/03/1824/03/18

Huella

Profundice en los temas de investigación de 'An Abstractive Text Summarization Using Recurrent Neural Network'. En conjunto forman una huella única.

Citar esto