Bengali Abstractive News Summarization (BANS): A Neural Attention Approach

Abstract

ive summarization is the process of generating novel sentences based on the information extracted from the original text document while retaining the context. Due to abstractive summarization's underlying complexities, most of the past research work has been done on the extractive summarization approach. Nevertheless, with the triumph of the sequence-to-sequence (seq2seq) model, abstractive summarization becomes more viable. Although a significant number of notable research has been done in the English language based on abstractive summarization, only a couple of works have been done on Bengali abstractive news summarization (BANS). In this article, we presented a seq2seq based Long Short-Term Memory (LSTM) network model with attention at encoder-decoder. Our proposed system deploys a local attention-based model that produces a long sequence of words with lucid and human-like generated sentences with noteworthy information of the original document. We also prepared a dataset of more than 19k articles and corresponding human-written summaries collected from bangla.bdnews24.com 1 which is till now the most extensive dataset for Bengali news document summarization and publicly published in Kaggle 2 . We evaluated our model qualitatively and quantitatively and compared it with other published results. It showed significant improvement in terms of human evaluation scores with state-of-the-art approaches for BANS.

Natural Language Processing For Automatic text summarization [Datasets] - Survey

“…Although there has been a substantial amount of critical study on abstractive summary in English, just a few works have been done on Bengali abstractive news summarization. Paper [36] described a seq2seq-based Long Short-Term Memory (LSTM) network architecture focused on the encoder-decoder. The suggested system uses the attention-based model to construct a long sequence of words.…”

BanglaNLG and BanglaT5: Benchmarks and Resources for Evaluating Low-Resource Natural Language Generation in Bangla

“…The Bangla Natural Language Generation (BanglaNLG) BenchmarkThere have been sporadic works on Bangla NLG, mostly catered to machine translation(Hasan et al, 2020; Mumin et al, 2019a,b) and text summarization(Bhattacharjee et al, 2021b;Dhar et al, 2021). However, Bangla NLG lacks a unified study comprising diverse and challenging tasks.…”