Knowledge-Grounded Attention-Based Neural Machine Translation Model

Neural machine translation (NMT) model processes sentences in isolation and ignores additional contextual or side information beyond sentences. The input text alone often provides limited knowledge to generate contextually correct and meaningful translation. Relying solely on the input text could yi...

Full description

Saved in:
Bibliographic Details
Main Authors: Huma Israr, Safdar Abbas Khan, Muhammad Ali Tahir, Muhammad Khuram Shahzad, Muneer Ahmad, Jasni Mohamad Zain
Format: Article
Language:English
Published: Wiley 2025-01-01
Series:Applied Computational Intelligence and Soft Computing
Online Access:http://dx.doi.org/10.1155/acis/6234949
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850196454908362752
author Huma Israr
Safdar Abbas Khan
Muhammad Ali Tahir
Muhammad Khuram Shahzad
Muneer Ahmad
Jasni Mohamad Zain
author_facet Huma Israr
Safdar Abbas Khan
Muhammad Ali Tahir
Muhammad Khuram Shahzad
Muneer Ahmad
Jasni Mohamad Zain
author_sort Huma Israr
collection DOAJ
description Neural machine translation (NMT) model processes sentences in isolation and ignores additional contextual or side information beyond sentences. The input text alone often provides limited knowledge to generate contextually correct and meaningful translation. Relying solely on the input text could yield translations that lack accuracy. Side information related to either source or target side is helpful in the context of NMT. In this study, we empirically show that training an NMT model with target-side additional information used as knowledge can significantly improve the translation quality. The acquired knowledge is leveraged in the encoder-/decoder-based model utilizing multiencoder framework. The additional encoder converts knowledge into dense semantic representation called attention. These attentions from the input sentence and additional knowledge are then combined into a unified attention. The decoder generates the translation by conditioning on both the input text and acquired knowledge. Evaluation of translation from Urdu to English with a low-resource setting yields promising results in terms of both perplexity reduction and improved BLEU scores. The proposed models in the respective group outperform in LSTM and GRU with attention mechanism by +3.1 and +2.9 BLEU score, respectively. Extensive analysis confirms our claim that the translations influenced by additional information may occasionally contain rare low-frequency words and faithful translation. Experimental results on a different language pair DE-EN demonstrate that our suggested method is more efficient and general.
format Article
id doaj-art-86c1f1dd131a4d979c672fefa2cd5fc5
institution OA Journals
issn 1687-9732
language English
publishDate 2025-01-01
publisher Wiley
record_format Article
series Applied Computational Intelligence and Soft Computing
spelling doaj-art-86c1f1dd131a4d979c672fefa2cd5fc52025-08-20T02:13:27ZengWileyApplied Computational Intelligence and Soft Computing1687-97322025-01-01202510.1155/acis/6234949Knowledge-Grounded Attention-Based Neural Machine Translation ModelHuma Israr0Safdar Abbas Khan1Muhammad Ali Tahir2Muhammad Khuram Shahzad3Muneer Ahmad4Jasni Mohamad Zain5Department of Computer ScienceDepartment of Computer SciencesDepartment of Computer ScienceDepartment of Computer ScienceDepartment of Computer ScienceInstitute for Big Data Analytics and Artificial Intelligence (IBDAAI)Neural machine translation (NMT) model processes sentences in isolation and ignores additional contextual or side information beyond sentences. The input text alone often provides limited knowledge to generate contextually correct and meaningful translation. Relying solely on the input text could yield translations that lack accuracy. Side information related to either source or target side is helpful in the context of NMT. In this study, we empirically show that training an NMT model with target-side additional information used as knowledge can significantly improve the translation quality. The acquired knowledge is leveraged in the encoder-/decoder-based model utilizing multiencoder framework. The additional encoder converts knowledge into dense semantic representation called attention. These attentions from the input sentence and additional knowledge are then combined into a unified attention. The decoder generates the translation by conditioning on both the input text and acquired knowledge. Evaluation of translation from Urdu to English with a low-resource setting yields promising results in terms of both perplexity reduction and improved BLEU scores. The proposed models in the respective group outperform in LSTM and GRU with attention mechanism by +3.1 and +2.9 BLEU score, respectively. Extensive analysis confirms our claim that the translations influenced by additional information may occasionally contain rare low-frequency words and faithful translation. Experimental results on a different language pair DE-EN demonstrate that our suggested method is more efficient and general.http://dx.doi.org/10.1155/acis/6234949
spellingShingle Huma Israr
Safdar Abbas Khan
Muhammad Ali Tahir
Muhammad Khuram Shahzad
Muneer Ahmad
Jasni Mohamad Zain
Knowledge-Grounded Attention-Based Neural Machine Translation Model
Applied Computational Intelligence and Soft Computing
title Knowledge-Grounded Attention-Based Neural Machine Translation Model
title_full Knowledge-Grounded Attention-Based Neural Machine Translation Model
title_fullStr Knowledge-Grounded Attention-Based Neural Machine Translation Model
title_full_unstemmed Knowledge-Grounded Attention-Based Neural Machine Translation Model
title_short Knowledge-Grounded Attention-Based Neural Machine Translation Model
title_sort knowledge grounded attention based neural machine translation model
url http://dx.doi.org/10.1155/acis/6234949
work_keys_str_mv AT humaisrar knowledgegroundedattentionbasedneuralmachinetranslationmodel
AT safdarabbaskhan knowledgegroundedattentionbasedneuralmachinetranslationmodel
AT muhammadalitahir knowledgegroundedattentionbasedneuralmachinetranslationmodel
AT muhammadkhuramshahzad knowledgegroundedattentionbasedneuralmachinetranslationmodel
AT muneerahmad knowledgegroundedattentionbasedneuralmachinetranslationmodel
AT jasnimohamadzain knowledgegroundedattentionbasedneuralmachinetranslationmodel