Inproceedings
Book title :
International Conference on INnovations in Intelligent SysTems and Applications (INISTA 2021)
Address :
Kocaeli, Turkey
ISBN : 978-1-6654-1162-2
Publisher :
Information
Period : August 2021
Pages : 1-6
Details
A Discrete Differential Evolution Algorithm for Extractive Text Summarization
Khadidja Chettah Amer Draa
With the exponential growth of textual data on the web, automatic text summarization has become a key task in many natural language processing applications. Many works have been proposed to solve it, but the problem is still far from being totally solved. In this work, we propose a new extractive text summarizer that uses a discrete differential evolution algorithm. It looks for the best set of sentences to be put in the summary, in addition to the best order to be respected. To evaluate the performances of our approach, we used as benchmarks the DUC2001 and DUC2002 collections and opted for the ROUGE evaluation toolkit. The proposed approach is compared against eleven state-of-the-art methods. The obtained results show that our approach is very competitive. It outperforms many state-of-the-art methods; it ranks first out of 12 methods.
Key words :
Discrete Differential Evolution Extractive Text Summarization Natural Language Processing
Ref. laboratory citation :
misc-lab-374
DOI :
10.1109/INISTA52262.2021.9548632
Link :
Texte intégral
ACM :
K. Chettah and A. Draa. 2021. A Discrete Differential Evolution Algorithm for Extractive Text Summarization. In Proceedings of the International Conference on INnovations in Intelligent SysTems and Applications (INISTA 2021), Kocaeli, Turkey (August 2021), IEEE, 1-6. DOI: https://doi.org/10.1109/INISTA52262.2021.9548632.
APA :
Chettah, K. & Draa, A. (2021, August). A Discrete Differential Evolution Algorithm for Extractive Text Summarization. In Proceedings of the International Conference on INnovations in Intelligent SysTems and Applications (INISTA 2021), Kocaeli, Turkey, IEEE, 1-6. DOI: https://doi.org/10.1109/INISTA52262.2021.9548632
IEEE :
K. Chettah and A. Draa, "A Discrete Differential Evolution Algorithm for Extractive Text Summarization". In Proceedings of the International Conference on INnovations in Intelligent SysTems and Applications (INISTA 2021), Kocaeli, Turkey, IEEE, pp. 1-6, August, 2021. DOI: https://doi.org/10.1109/INISTA52262.2021.9548632.
BibTeX :
@inproceedings{misc-lab-374,
author = {Chettah, Khadidja and Draa, Amer},
title = {A Discrete Differential Evolution Algorithm for Extractive Text Summarization},
booktitle = {International Conference on INnovations in Intelligent SysTems and Applications (INISTA 2021)},
location = {Kocaeli, Turkey},
isbn = {978-1-6654-1162-2},
pages = {1--6},
publisher = {IEEE},
year = {2021},
month = {August},
doi = {10.1109/INISTA52262.2021.9548632},
url = {https://doi.org/10.1109/INISTA52262.2021.9548632},
keywords = {Discrete Differential Evolution, Extractive Text Summarization, Natural Language Processing}
}
RIS :
TY  - CONF
TI - A Discrete Differential Evolution Algorithm for Extractive Text Summarization
AU - K. Chettah
AU - A. Draa
PY - 2021
BT - International Conference on INnovations in Intelligent SysTems and Applications (INISTA 2021), Kocaeli, Turkey
SN - 978-1-6654-1162-2
SP - 1
EP - 6
PB - IEEE
AB - With the exponential growth of textual data on the web, automatic text summarization has become a key task in many natural language processing applications. Many works have been proposed to solve it, but the problem is still far from being totally solved. In this work, we propose a new extractive text summarizer that uses a discrete differential evolution algorithm. It looks for the best set of sentences to be put in the summary, in addition to the best order to be respected. To evaluate the performances of our approach, we used as benchmarks the DUC2001 and DUC2002 collections and opted for the ROUGE evaluation toolkit. The proposed approach is compared against eleven state-of-the-art methods. The obtained results show that our approach is very competitive. It outperforms many state-of-the-art methods; it ranks first out of 12 methods.
KW - Discrete Differential Evolution
KW - Extractive Text Summarization
KW - Natural Language Processing
DO - 10.1109/INISTA52262.2021.9548632
UR - https://doi.org/10.1109/INISTA52262.2021.9548632
ID - misc-lab-374
ER -