Sentiment Analysis aims to analyse a large amount of data to infer and deduce the different opinions expressed therein. It would be interesting to link several linguistic resources and tools to combine their advantages and obtain a result as close as possible to reality. Moreover, creating a new lexicon either manually or automatically is time-consuming and needs considerable efforts. This paper proposes a new approach to build a polar lexicon by using the existing resources to generate a rich and time-saving one. Nevertheless, not all linguistic resources have the same structure. It is then challenging to use them together at once. We propose to map the resources into the graph structure, where the nodes represent the words, and the edges represent the relations between the nodes, thus the words. The obtained results are promising, F1-measure was 0.82 for the English language, and accuracy ranged from 0.76 to 0.98 for all other languages used in the experiment.
Key words :
Sentiment analysis
Linguistic resources
Relationship
Multilingual
Lexicon-based approach
Graph-based
M. R. Kanfoud and A. Bouramoul. 2021. Linking the Linguistic Resources Using Graph Structure for Multilingual Sentiment Analysis. In Proceedings of the Proceedings of Sixth International Congress on Information and Communication Technology, Singapore. Lecture Notes in Networks and Systems (LNNS), Xin-She Yang; Simon Sherratt; Nilanjan Dey; Amit Joshi, Ed., 235 (February 2021), Springer Singapore, 973-982. DOI: https://doi.org/https://doi.org/10.1007/978-981-16-2377-6_88.
APA :
Kanfoud, M. R. & Bouramoul, A. (2021, February). Linking the Linguistic Resources Using Graph Structure for Multilingual Sentiment Analysis. In Proceedings of the Proceedings of Sixth International Congress on Information and Communication Technology, Singapore. In Xin-She Yang; Simon Sherratt; Nilanjan Dey; Amit Joshi (Ed.). Lecture Notes in Networks and Systems (LNNS), 235, Springer Singapore, 973-982. DOI: https://doi.org/https://doi.org/10.1007/978-981-16-2377-6_88
IEEE :
M. R. Kanfoud and A. Bouramoul, "Linking the Linguistic Resources Using Graph Structure for Multilingual Sentiment Analysis". In Proceedings of the Proceedings of Sixth International Congress on Information and Communication Technology, Singapore. Lecture Notes in Networks and Systems (LNNS), vol. 235, Springer Singapore, pp. 973-982, February, 2021. DOI: https://doi.org/https://doi.org/10.1007/978-981-16-2377-6_88.
BibTeX :
@inproceedings{misc-lab-362, author = {Kanfoud, Mohamed Raouf and Bouramoul, Abdelkrim}, title = {Linking the Linguistic Resources Using Graph Structure for Multilingual Sentiment Analysis}, volume = {235}, booktitle = {Proceedings of Sixth International Congress on Information and Communication Technology}, series = {Lecture Notes in Networks and Systems (LNNS)}, location = {Singapore}, isbn = {978-981-16-2377-6}, pages = {973--982}, editor = {Yang, Xin-She and Sherratt, Simon and Dey, Nilanjan and Joshi, Amit}, publisher = {Springer Singapore}, year = {2021}, month = {February}, doi = {https://doi.org/10.1007/978-981-16-2377-6\_88}, url = {https://link.springer.com/chapter/10.1007/978-981-16-2377-6\_88}, keywords = {Sentiment analysis, Linguistic resources, Relationship, Multilingual, Lexicon-based approach, Graph-based} }
RIS :
TY - CONF TI - Linking the Linguistic Resources Using Graph Structure for Multilingual Sentiment Analysis AU - M. R. Kanfoud AU - A. Bouramoul PY - 2021 VL - 235 BT - Proceedings of Sixth International Congress on Information and Communication Technology, Singapore SN - 978-981-16-2377-6 SP - 973 EP - 982 ED - X.-S. Yang ED - S. Sherratt ED - N. Dey ED - A. Joshi PB - Springer Singapore AB - Sentiment Analysis aims to analyse a large amount of data to infer and deduce the different opinions expressed therein. It would be interesting to link several linguistic resources and tools to combine their advantages and obtain a result as close as possible to reality. Moreover, creating a new lexicon either manually or automatically is time-consuming and needs considerable efforts. This paper proposes a new approach to build a polar lexicon by using the existing resources to generate a rich and time-saving one. Nevertheless, not all linguistic resources have the same structure. It is then challenging to use them together at once. We propose to map the resources into the graph structure, where the nodes represent the words, and the edges represent the relations between the nodes, thus the words. The obtained results are promising, F1-measure was 0.82 for the English language, and accuracy ranged from 0.76 to 0.98 for all other languages used in the experiment. KW - Sentiment analysis KW - Linguistic resources KW - Relationship KW - Multilingual KW - Lexicon-based approach KW - Graph-based DO - https://doi.org/10.1007/978-981-16-2377-6_88 UR - https://link.springer.com/chapter/10.1007/978-981-16-2377-6_88 ID - misc-lab-362 ER -