Approaches to Building Semantic Networks for Machine Translation

Usmanova Kamola
Bulletin of TUIT: Management and Communication Technologies
№ 3s(65)2024 DOI:
Open article file

Abstract

This research delves into approaches for identifying semantic links in natural language by analyzing how symbols, concepts, and patterns are close to each other. Two computational techniques were applied: cosine similarity and TF-IDF. Cosine similarity, using a text-word matrix, assessed how often terms appear together within the corpus, while TF-IDF highlighted terms that are both frequent and significant in context. The outcomes were visualized as semantic networks, indicating weighted connections among words. To illustrate, in the sports domain, sportsmen (0.59), badminton (0.49), culture (0.48), and physical education (0.48) adduced the strongest associations with the main concept “sport.” These networks grouped terms into categories such as participants, sporting disciplines, related fields, cultural components, and mobility activities. The study ensures that integrating cosine similarity and TF-IDF effectively captures contextual meaning and supports the development of semantic models useful for machine translation, information retrieval, and natural language processing.

References

[1] J. Slocum, A survey of machine translation: its history, current status, and future prospects, Computational linguistics 11 (1) (1985) 1–17.
[2] P. Koehn, StatisticalMachine Translation, Cambridge University Press,2010.
[3] D. Jurafsky, Speech and language processing: An introduction to natural language processing, Prentice Hall, 2000.
[4] Bafna, P., Pramod, D., and Vaidya, A. (2016). "Document clustering: TF-IDF approach," International Conference on Electrical, Electronics, and Optimization Techniques (ICEEOT), Chennai, 2016, pp. 61-66
[5] Trstenjak, B., Mikac, S., & Donko, D. (2014). “KNN with TF-IDF based framework for text categorization” In Procedia Engineering. Vol. 69, pp. 1356–1364. Elsevier Ltd
[6] R. Feldman and J. Sanger, “The text mining handbook: advanced approaches in analyzing unstructured data,” Choice Rev. Online, 2007, doi: 10.5860/choice.44-5684.
[7] R. Ahuja, A. Chug, S. Kohli, S. Gupta, and P. Ahuja, “The impact of features extraction on the sentiment analysis,” Procedia Computer Science, vol. 152, pp. 341-348, 2019, doi: 10.1016/j.procs.2019.05.008.
[8]B. B. Elov, Z. Khusainova, and N. Khudayberganov, “O‘zbek tili korpusi matnlari uchun TF-IDF statistik ko‘rsatkichni hisoblash,” Article, Aug. 2023.
[9] W. Zhu, W. Zhang, G. Z. Li, C. He, and L. Zhang, “A study of damp-heat syndrome classification using Word2vec and TF-IDF,” 2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2016, pp. 1415-1420, doi: 10.1109/BIBM.2016.7822730.